IMPROVING PARTICIPATION IN CANCER SCREENING PROGRAMS:
A REVIEW OF SOCIAL COGNITIVE MODELS, FACTORS AFFECTING
PARTICIPATION, AND STRATEGIES TO IMPROVE PARTICIPATION
Prepared for the Victorian Cytology Service by
Dr Susan Day
Pauline van Dort
Kiu-siang Tay-Teo
Centre for Health Policy, Programs and Economics, University of Melbourne
Victorian Cytology Service PO Box 161 Carlton South VIC 3053 Website: www.vcs.org.au © Victorian Cytology Service 2010 IMPROVING PARTICIPATION IN CANCER SCREENING PROGRAMS: A REVIEW OF SOCIAL COGNITIVE MODELS, FACTORS AFFECTING PARTICIPATION, AND STRATEGIES TO IMPROVE PARTICIPATION BACKGROUND The Victorian Cancer Action Plan (2008‐2011) identifies increasing participation rates in population‐based cancer screening programs as a Priority Area. In order to inform the VCAP activities aimed at improving participation in cancer screening, a review of the literature was commissioned by the Department of Health and prepared by the Centre for Health Policy, Programs and Economics at The University of Melbourne for the Victorian Cytology Service (VCS). The work was overseen by a Steering Committee with input from the Under‐screened Working Group convened by the Victorian Department of Health. The review includes both international and Australian literature (peer reviewed and the grey literature), and focuses on key areas that inform thinking, knowledge and approaches to engaging individuals, groups and communities to participate in population‐based cancer screening programs. Due to the large amount of literature to be reviewed, the work was undertaken in three phases and is contained in three stand‐alone volumes: 1.
Social Cognition Models: A Review of their Relevance for Understanding Participation in Cancer Screening 2.
Identifying Hard to Reach Groups: Review of the Factors (Including Barriers) Associated with Cancer Screening: 3.
Knowledge Translation: A Review of Strategies to Increase Participation in Cancer Screening Special areas of focus included: needs of specific groups that may be under‐
screened, innovative approaches to enhancing participation, barriers to participation and the translation of research findings into action. As there are few screening programs targeted at men, gender issues relevant to participation in screening were also explored. IMPROVING PARTICIPATION IN CANCER SCREENING PROGRAMS: A REVIEW OF SOCIAL COGNITIVE MODELS, FACTORS AFFECTING PARTICIPATION, AND STRATEGIES TO IMPROVE PARTICIPATION ACKNOWLEDGEMENTS The Victorian Cytology Service would like to thank the members of the Literature Review Steering Committee for their oversight of the project and the Victorian Department of Health Under‐screened Working Group for valuable feedback. Victorian Cytology Service Literature Review Steering Committee Dorota Gertig Marion Saville Kate Broun Genevieve Chappell Bianca Barbaro Julia Brotherton Under‐screened Working Group Victorian Department of Health Louise Galloway Janine Coffin Rachael Andersen Celia Gahan Anil Raichur Leanne Lade BreastScreen Victoria Vicki Pridmore Jules Wilkinson PapScreen Victoria Kate Broun The Cancer Council of Victoria Alison Peipers Victorian Cytology Service Marion Saville Dorota Gertig Julia Brotherton Genevieve Chappell Bianca Barbaro The Victorian Cytology Service would also like to thank Janine Coffin, Louise Galloway and Rachael Andersen from the Victorian Department of Health and Penny Allen from the Victorian Cytology Service for their assistance throughout the project. IMPROVING PARTICIPATION IN CANCER SCREENING PROGRAMS: A REVIEW OF SOCIAL COGNITIVE MODELS, FACTORS AFFECTING PARTICIPATION, AND STRATEGIES TO IMPROVE PARTICIPATION REPORT OVERVIEW VOLUME 1 SOCIAL COGNITION MODELS: A REVIEW OF THEIR RELEVANCE FOR UNDERSTANDING PARTICIPATION IN CANCER SCREENING VOLUME 2 IDENTIFYING HARD‐TO‐REACH GROUPS: A REVIEW OF THE FACTORS (INCLUDING BARRIERS) ASSOCIATED WITH CANCER SCREENING
VOLUME 3 KNOWLEDGE TRANSLATION: A REVIEW OF STRATEGIES TO INCREASE PARTICIPATION IN CANCER SCREENING IMPROVING PARTICIPATION IN CANCER SCREENING PROGRAMS: A REVIEW OF SOCIAL COGNITIVE MODELS, FACTORS AFFECTING PARTICIPATION, AND STRATEGIES TO IMPROVE PARTICIPATION IMPROVING PARTICIPATION IN CANCER SCREENING PROGRAMS: A REVIEW OF SOCIAL COGNITIVE MODELS, FACTORS AFFECTING PARTICIPATION, AND STRATEGIES TO IMPROVE PARTICIPATION VOLUME 1 SOCIAL COGNITION MODELS: A REVIEW OF THEIR RELEVANCE FOR UNDERSTANDING PARTICIPATION IN CANCER SCREENING TABLE OF CONTENTS TABLE OF CONTENTS ....................................................................................................................2 LIST OF TABLES..............................................................................................................................3 LIST OF FIGURES............................................................................................................................3 1. INTRODUCTION ..................................................................................................................4 1.1 1.2 1.2 2 HEALTH BELIEF MODEL ......................................................................................................6 2.1 2.2 2.2 3 BACKGROUND ............................................................................................................4 SOCIAL COGNITION .....................................................................................................4 METHODOLOGY .........................................................................................................5 UNDERSTANDING BELIEFS ............................................................................................6 DESCRIPTION OF THE HEALTH BELIEF MODEL ..................................................................6 UTILITY OF THE HEALTH BELIEF MODEL ..........................................................................8 THEORY OF PLANNED BEHAVIOUR .................................................................................14 3.1 3.2 DESCRIPTION .......................................................................................................... 14 UTILITY OF THE THEORY OF PLANNED BEHAVIOUR MODEL ............................................. 16 4 INTENTION–BEHAVIOUR RELATIONSHIP .......................................................................19 5 IMPLICATIONS AND CONCLUSIONS ................................................................................25 5.1 5.2 ISSUES IMPACTING ON THE UTILITY OF THE MODELS ....................................................... 25 CONCLUSIONS ......................................................................................................... 26 REFERENCES ................................................................................................................................28 VOL 1– 2 Social cognition models: a review of their relevance for understanding participation in cancer screening LIST OF TABLES TABLE 1 HEALTH BELIEF MODEL USED TO EXAMINE FACTORS ASSOCIATED WITH BREAST AND CERVICAL SCREENING IN HISPANIC WOMEN ...................................................................................... 7 TABLE 2 ESTIMATED EFFECT SIZES BY TYPE OF SCREENING TEST ......................................................... 17 TABLE 3 ESTIMATED EFFECT SIZES BY PROGRAM FACTORS ................................................................ 17 TABLE 4 CONCEPTUAL FACTORS MODERATING THE INTENTION CHANGE (I∆)–BEHAVIOUR CHANGE (B∆) RELATIONSHIP .............................................................................................................. 20 TABLE 5 ESTIMATED EFFECT SIZES FOR INTERVENTION BASED ON THEORETICAL MODELS ....................... 21 TABLE 6 ESTIMATED EFFECT SIZES FOR TYPES OF INTERVENTIONS AIMED AT CHANGING BEHAVIOUR ........ 21 TABLE 7 ESTIMATED EFFECT SIZES FOR TYPE OF SCREENING TEST FOR INTENTION BEHAVIOUR RELATIONSHIPS ............................................................................................................. 22 TABLE 8 COMPARISON OF RELATIONSHIP BETWEEN BEHAVIOUR AND INTENTIONS BY COST OF SCREENING .................................................................................................................. 22 TABLE 9 CONCEPTUAL SCHEMA OF ACTORS AND ABSTAINERS IN THE INTENTION–BEHAVIOUR RELATIONSHIP .............................................................................................................. 23 TABLE 10 EMPIRICAL RESULTS FOR ACTORS AND ABSTAINERS IN THE INTENTION–BEHAVIOUR RELATIONSHIP .............................................................................................................. 23 LIST OF FIGURES FIGURE 1 HEALTH BELIEF MODEL ................................................................................................... 8 FIGURE 2 SIGNIFICANCE RATIOS FOR HEALTH BELIEF MODEL CONSTRUCTS FOR PREVENTIVE BEHAVIOURS (29 STUDIES) ............................................................................................... 9 FIGURE 3 SCREENING BEHAVIOUR EFFECT SIZES FOR THE FACTORS IN THE HEALTH BELIEF MODEL (BASED ON 5 STUDIES; 2136 PARTICIPANTS) ....................................................................... 9 FIGURE 4 PERCEPTIONS RISK OF DEVELOPING BREAST CANCER AMONG NEVER‐SCREENED, UNDER‐SCREENED AND WELL‐SCREENED WOMEN AGED 50–69 YEARS (POST‐CAMPAIGN) ........ 11 FIGURE 5 THREAT EVALUATION: RISK PERCEPTIONS OF PEOPLE WHO DID AND DID NOT RETURN FAECAL OCCULT BLOOD TESTING KITS IN THE EVALUATION OF THE BOWEL CANCER SCREENING PILOT PROGRAM .......................................................................................................... 12 FIGURE 6 BEHAVIOURAL EVALUATION: AGREEMENT WITH THE STATEMENT THAT THE FAECAL OCCULT BLOOD TESTING WAS ‘MORE TROUBLE THAN IT WAS WORTH’ IN THE EVALUATION OF THE BOWEL CANCER SCREENING PILOT PROGRAM ................................................................... 13 FIGURE 7 MODEL OF THE THEORY OF PLANNED BEHAVIOUR............................................................... 14 FIGURE 8 REVISED THEORY OF PLANNED BEHAVIOUR MODEL ............................................................ 15 FIGURE 9 ATTITUDES OF AUSTRALIAN WOMEN TOWARDS SCREENING BY BREASTSCREEN AUSTRALIA ....... 18 FIGURE 10 BREAST CANCER SCREENING: DIFFERENCES IN THE INTENTIONS BETWEEN NEVER‐SCREENED AND UNDER‐SCREENED AUSTRALIAN WOMEN .................................................................... 24 Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 3 1.
INTRODUCTION
1.1 Background The Victorian Cytology Service (VCS) was commissioned by the Victorian Department of Health to undertake a review of international literature (peer‐reviewed and grey literature) focusing on key areas that inform thinking, knowledge and approaches to engaging individuals, groups and communities to participate in organised programs, including population‐based screening programs. The aim of the review was to consider:  factors that enhance or hinder engagement in health screening and other preventative programs, including perceptions held by participants  needs of specific groups (e.g. information needs of specific groups, such as those with low literacy skills)  health beliefs and their impact on cognitive constructs and behaviour in relation to screening and engagement in positive health behaviours  structural and environmental barriers to participation  key work in other disciplines or settings that provide insight into innovative thinking and practice in engaging individuals and communities; especially within organised, voluntary programs or initiatives  work completed in health and other disciplines on translating research findings into action. The VCS contracted the Centre for Health Policy, Programs and Economics (CHPPE) to undertake the identification, synthesis and summary of the evidence. Taken together, the three volumes cover the Department’s requirements relating to the role of health beliefs, the factors associated with screening and knowledge translation. The three volumes are: 1. Social cognition models: a review of their relevance for understanding participation in cancer screening. 2. Identifying hard‐to‐reach groups: review of the factors (including barriers) associated with cancer screening 3. Knowledge translation: a review of strategies to increase participation in cancer screening In Volume 2, a variety of factors were shown to be associated (either positively or negatively) with the uptake of screening. The factors were categorised under six domains. This first volume of the review concentrates on the cognitive domain. 1.2 Social cognition Social cognition is concerned with how individuals make sense of social situations. The approach focuses on individual … thoughts as processes which intervene between observable stimuli and responses in specific real world VOL 1– 4 Social cognition models: a review of their relevance for understanding participation in cancer screening situations. A significant proportion of social psychology over the past quarter century has started from this assumption that social behaviour is best understood as a function of people's perceptions of reality, rather than as a function of an objective description of the stimulus environment.1 Social cognition models (SCMs) describing the key cognitions (i.e. thoughts or perceptions) and their inter‐relationships in the regulation of behaviour have been extensively applied to the understanding of health behaviours. The most widely applied SCMs include:  the Health Belief Model (HBM)  Protection Motivation Theory (PMT)  Theory of Planned Behaviour (TPB) and its forerunner the Theory of Reasoned Action (TRA)  Social Cognitive Theory (SCT)  health locus of control  a set of models that focus on the idea that behaviour change occurs through a series of qualitatively different stages (e.g. the Transtheoretical Model (TTM), the precaution‐
adoption process model (PAPM), the health action process approach).1 1.2 Methodology This review has focused on the HBM and the TPB. 1 It is organised as follows:  Chapter 2: The Health Belief Model  Chapter 3: The Theory of Planned Behaviour  Chapter 4: An examination of the intention–behaviour (I–B) relationship  Chapter 5: The implications of the review for understanding screening rates and developing interventions. Chapters 2 and 3 begin with a brief description of the models and how they have changed over time. The descriptions rely heavily on the work of Connor and Norman (2005) to provide an up‐to‐date description of the models.1 The second section in each of these chapters looks at the utility of the models. This analysis draws on quantitative studies (mainly, but not entirely meta‐analyses) in the peer‐reviewed literature that show the strength of the relationships between the concepts included in each of the models. Throughout the review, a small amount of Australian data is included to demonstrate some of the concepts. These examples are indicative only and should not be over‐interpreted as the aggregated nature of the data precluded multivariate analyses. 1
2
For an overview of other SCMs, see Conner and Norman (2005) Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 5 2 HEALTH BELIEF MODEL 2.1 Understanding beliefs Beliefs are individual cognitive characteristics (perceptions) that shape behaviour. They can be acquired by:  internalising the beliefs of the people around us during childhood (primary socialisation)  adopting the beliefs of significant others (e.g. peers and leaders)  being exposed to repetitive messages, and association of beliefs with images of sex, love and other strong positive emotions (the primary thrust of the advertising industry)  physical trauma. Beliefs can differentiate between individuals from the same background and, although there is an assumption that beliefs can be modified, people often cling to beliefs and act on them even against their own self‐interest. An example is provided by Yarbrough and Braden (2001) in their article on the utility of the health belief model as a guide for explaining or predicting breast cancer screening behaviour noted: Residents of a community with a documented increased incidence of cancer were very resistant to health education interventions in spite of knowing about their increased risk. They argued that because the cause of cancer is unclear, health promotion and protection activities might not be effective for reducing cancer risk.3 2.2 Description of the Health Belief Model The HBM emerged in the late 1950s. Rosenstock (1974) attributed the first HBM research to Hochbaum's 1958 studies of the uptake of tuberculosis X‐ray screening. The HBM was originally used to understand why people did not participate in preventive services, and more recently to understand decisions around the use of health services.4,5 In the early models, the key beliefs considered to shape health behaviours were defined as the:  perceived likelihood of experiencing a health problem  perceived severity or seriousness of the consequences of experiencing the health problem  perceived benefits of the health action  perceived barriers or costs association with performing the health action.4,6 Over time, the model has been expanded to include other factors:, such as the inclusion of self‐efficacy. This concept asserts that whether or not people undertake a task or health behaviour will depend, in part, on their judgements of their own ability to organise and execute the actions or steps required to complete the task.7 Rosenstock, Strecher and Becker (1988) believed that including self‐efficacy would provide a more powerful approach in understanding health‐related behaviour. 8 According to Abraham and Sheeran (2005), a number of studies have tested the predictive utility of including self‐efficacy in the HBM, and generally confirm that ‘self‐efficacy is a useful additional predictor’. However, the authors do VOL 1– 6 Social cognition models: a review of their relevance for understanding participation in cancer screening point out that ‘self‐efficacy may not always enhance the predictive utility of the model’ when floor or ceiling effects are observed. This would occur if people were either uniformly very confident (ceiling), or uniformly very unconfident (floor), that they could perform the required actions.6 Other factors have been added to the model over time. Table 1 Health Belief Model used to examine factors associated with breast and cervical screening in Hispanic women1 shows the HBM used by Austin, McNally and Stewart (2002) to guide their literature review into the factors associated with breast and cervical screening in Hispanic women.5 This model included not only self‐efficacy but also ‘cues to action’. Table 1 Health Belief Model used to examine factors associated with breast and cervical screening in Hispanic women Concept Definition Application Perceived susceptibility One’s opinion of chances of getting a condition Define populations at risk, risk level; personalise risk based on a person’s features or behaviour; heighten perceived susceptibility if too low Perceived severity One’s opinion of how serious a condition and its sequelae are Specify consequences of the risk and the condition. Perceived benefits One’s opinion of the efficacy of the advised action to reduce risk or seriousness or impact Define action to take; how, where, when; clarify the positive effects to be expected. Perceived barriers One’s opinion of the tangible and psychological costs of the advised action Identify and reduce barriers through reassurance, incentives, assistance Cues to action Strategies to activate ‘readiness’ Provide how‐to information, promote awareness, reminders Self‐efficacy Confidence in one’s ability to take action Provide training, guidance in performance action Source: Austin, McNally and Stewart (2002)5 Figure 1 shows the HBM, as outlined by Abraham and Sheeran in 2005, which appears to be the most comprehensive exposition of the model. This exposition includes four major domains, as outlined below. 1. Threat evaluation Threat evaluation consists of two key beliefs: (i) perceived susceptibility to illness or health problem; (ii) the perceived severity of the consequences of illnesses. 2. Behavioural evaluation Behavioural evaluation also consists of two key beliefs: (i) perceptions concerning the benefits, or efficacy, of a recommended health behaviour; (ii) perceptions about the costs of, or barriers to, enacting the behaviour. 3. Health motivation Health motivation is the readiness to be concerned about health matters. 4. Cues to action Cues to action are a diverse range of triggers that can activate health behaviour when Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 7 appropriate beliefs are held. These include such things as an individual’s perceptions of symptoms, social influences and health education campaigns.6 This exposition of the model does not specifically include self‐efficacy, although it could perhaps be included under one of the other headings (e.g. perceived barriers). Figure 1 Health Belief Model Perceived
Susceptibility
Demographic
Variables
Perceived
Severity
Health
Motivation
Psychological
Characteristics
Action
Perceived
Benefits
Perceived
Barriers
Cues to
Action
Source: Abraham and Sheeran (2005)6 2.2 Utility of the Health Belief Model A number of uses have been outlined for the HBM. According to Abraham and Sheeran (2005), there were a number of studies prior to the early 1970s which indicated that the key health beliefs underlying the threat and behavioural evaluations provide a useful framework for understanding individual differences in health behaviour, and for designing interventions to change behaviour.6 According to Glanz and Rimer (1995), the HBM can be used effectively to guide the development of messages aimed at persuading individuals to undertake health actions such as screening mammography. 9 However, there are a number of difficulties that make any evaluation of the utility of the model difficult. The first relates to the fact that, in the evaluations of the HBM, most attention has focused on the four factors contained in the threat and behavioural evaluation. Cues to action and health motivation have received less attention.6 The implications of the HBM have generally been stated as the fact that people are more likely to undertake a health action if the threat evaluation is high (e.g. high risk with serious consequences) and the behavioural evaluation is positive (e.g. large benefits and few barriers).10 Two meta‐analyses have calculated significance ratios and effect sizes for the impact on behaviour for the four factors (Figure 2 and Figure 3), but there does not appear to have been any such analysis for the other constructs (health motivation and cues to action). VOL 1– 8 Social cognition models: a review of their relevance for understanding participation in cancer screening Figure 2 Significance ratios for Health Belief Model constructs for preventive behaviours (29 studies) 100%
Significance
Ratio
75%
50%
93%
86%
74%
25%
50%
0%
Perceived
Susceptibility
Perceived
Severity
Perceived
Benefits
Perceived
Barriers
Source: Janz and Becker (2002)10 Notes: Significance ratio: (number of studies with statistically significant results) ÷ (number of studies reporting significance tests) Figure 3 Screening behaviour effect sizes for the factors in the Health Belief Model (based on 5 studies; 2136 participants) 0.15
0.10
0.13
0.05
0.09
0.03
Effect Size
0.00
-0.05
Perceived
Susceptibility*
-0.10
Perceived
Severity
Perceived
Benefits*
Perceived
Barriers*
-0.22
-0.15
-0.20
-0.25
Source: Harrison et al. (1992)11 Notes: *p < 0.001
Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 9 The second difficulty in determining the utility of the HBM is the idiosyncratic interpretation and implementation of the model in practice. Examples are provided below.  In a review of the utility of the HBM as a guide for explaining or predicting breast cancer screening behaviours, Yarbrough and Braden (2000) found that the maximum amount of variance explained by HBM variables in the studies was 47%. The study that reported this result included the non‐HBM variable of socioeconomic status as a factor. In the other studies, the variability accounted for by any one, or a combination, of HBM variables was only 15–27%.3  Abraham and Sheeran (2005) reported that four theory‐based interventions (included in two studies) increased screening mammography utilisation on average by 23% compared to usual care.6 On checking the primary studies on which this conclusion was based, it was found that: o
one study used a model containing six belief concepts: susceptibility, severity, benefits, barriers, motivation and control; ‘control’ is not an HBM variable but was derived from the TPB outlined in the next chapter o
the other study included ‘intentions to undertake screening’ as a predictor in the analysis; once again, this is not an HBM factor but is also included in the TPB model. In 2000, Yarbrough and Braden concluded that the HBM did not have the power to predict behaviours consistently. According to these authors, the reasons for the HBM’s apparently limited predictive power stemmed from the methodological shortcomings of the model, namely:  theoretical descriptions not being strong enough to predict points for targeting interventions  a lack of evidence of the essential conditions for each ‘concept’: its boundaries, and the background factors that would influence it  the relationships between concepts have not been validated  most of the studies addressed linear relationships between the HBM factors rather than multiplicative interactive influences of variables on one another. In 2005, Abrahams and Sheeran (2005) reviewed findings from the quantitative reviews of the four factors. They concluded that, although the four factors are often found to be statistically significant predictors of health‐related behaviours, the effect size is small. 2 The authors point out a number of caveats that need to be borne in mind when considering the results of these analyses. One is that the effects of individual health beliefs should be combined in some way and that the combined effect may be greater than the sum of individual effects. The second is that the effect sizes show ‘considerable heterogeneity’, implying that differences in the design and measurement (i.e. operationalization) of the factors influence the results.6 Finally, this section on the utility of the HBM ends with the presentation of some descriptive data relating to the uptake of screening mammography and faecal occult blood testing (FOBT) in the Australian context. Although the data appear to indicate that there are differences 2
Cohen's (1992) recommendations for judging effect sizes (ES) 12 Test Statistic Small ES d 0.20 ma vs mb for independent means Significance of produce moment r 0.10 q 0.10 ra vs rb for independent rs Medium ES 0.50 0.30 0.30 Large ES 0.80 0.50 0.50 VOL 1– 10 Social cognition models: a review of their relevance for understanding participation in cancer screening between well‐screened, under‐screened and never‐screened groups across some of the factors outlined in the HBM, the data need to be interpreted with caution. Firstly, they represent bivariate analyses and the apparent relationships may be mediated by other variables. Secondly, the nature of the data in one of the primary sources does not allow for significance testing. This means that apparent differences may not be significant.  Screening mammography The data presented in Figure 4 indicate that there may be differences in perceived risk among women who have never had a mammogram (never‐screened), those who do not screen regularly (under‐screened) and those who are up‐to‐date with their screening (well‐screened). Never‐screened women were less likely to perceive themselves at risk of developing breast cancer than under‐screened and well‐screened women (68% vs 74% and 88%). Never‐screened and under‐screened women were more likely to indicate that they had no chance of developing breast cancer (14% and 13% vs 3%).13 However, these data needed to be treated with some caution as they are cross‐sectional rather than prospective. This means there is no way of knowing the impact of screening on the risk perceptions of the participants. Figure 4 Perceptions risk of developing breast cancer among never‐screened, under‐
screened and well‐screened women aged 50–69 years (post‐campaign) 100%
Percentage
of Respondents
80%
60%
40%
20%
0%
Never-Screened
Under-Screened
Well-Screened
Don't Know
18%
13%
9%
No Chance
14%
13%
3%
Slight Chance
60%
61%
69%
High Chance
8%
13%
19%
13
Source: King et al. (2003) Notes: Percentages exclude those who indicated they had already had a diagnosis of breast cancer, and may not sum to 100% within the groups due to rounding Slight chance: slight chance + very slight chance High chance: High chance + very high chance Sample sizes not included in the original Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 11  Faecal occult blood testing The data presented in Figure 5 indicate that there was no statistically significant difference in the risk perceptions of people who returned FOBT kits in the Bowel Cancer Screening Pilot Program (BCSPP) and those who did not return the kits. However, as shown in Figure 6, those who did not return kits had a less positive behavioural evaluation than those who did not. They were nearly 12 times more likely to agree with the statement that ‘Having a test like FOBT seems like more trouble than it's worth’ (23% vs 2%). Those who returned the kits were more than twice as likely to strongly disagree with the statement (72% vs 30%).14 Figure 5 Threat evaluation: risk perceptions of people who did and did not return faecal occult blood testing kits in the evaluation of the Bowel Cancer Screening Pilot Program 80%
Percentage
of Respondents
60%
40%
20%
0%
Returned FOBT
(N=1240)
Did Not Return FOBT
(N=296)
Likely
22%
19%
Unlikely
64%
64%
Others
14%
17%
Source: Department of Health and Ageing (2004)14 Notes: Likely: Likely + very likely Unlikely: Unlikely + very unlikely Others: Neither likely nor unlikely + not sure/don't know/it depends FOBT: Faecal occult blood testing Post‐hoc significance test using three response categories conducted by CHPPE: Chisq = 2.76, df = 2, p = 0.251 Although the significance test using all six response categories shows a significant difference (p = 0.03), it appears that the ‘Others’ categories are responsible for the statistical significance. Post‐
hoc significance testing using five response categories showed no statistically significant difference between the two groups (Chisq = 6.01, df = 4, p = 0.198). VOL 1– 12 Social cognition models: a review of their relevance for understanding participation in cancer screening Figure 6 Behavioural evaluation: agreement with the statement that the faecal occult blood testing was ‘more trouble than it was worth’ in the evaluation of the Bowel Cancer Screening Pilot Program Percentage
of Respondents
80%
60%
40%
20%
0%
Did not Return FOBT
(N=303)
Returned FOBT
(N=1261)
Strongly Disagree
30%
72%
Disagree
38%
24%
Agree
18%
2%
Strongly Agree
5%
0%
Others
10%
2%
Source: Department of Health and Ageing (2004)14 Notes: Others: Neither agree nor disagree + unsure/don't know FOBT: Faecal occult blood testing Post‐hoc significance test using five response categories conducted by CHPPE: Chisq = 305.54, df = 4, p < 0.001. Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 13 3 THEORY OF PLANNED BEHAVIOUR 3.1 Description Because the Theory of Reasoned Action (TRA) suffered from limitations in dealing with behaviour over which people do not have complete control, the model was revised and renamed the Theory of Planned Behaviour (TPB).15‐17 The TRA proposed that a person’s intention (I) to adopt a particular behaviour was influenced by the person’s attitudes (A) toward that behaviour, and subjective norms (SN). The TPB included a third component: perceived behavioural control (PBC).18, 19 According to Cooke and French (2008), the TPB and the TRA (Ajzen 1997), have been utilised ‘extensively’ to predict health behaviour.17 The three conceptually independent determinants of intention originally proposed by the TBP are summarised below. 1. Attitude toward the behaviour This refers to the extent to which a person has a favourable, or unfavourable, evaluation of the behaviour in question. 2. Subjective norm This refers to the perceived social pressure from others to perform, or not to perform, the behaviour. 3. Perceived behavioural control This refers to the perceived ease or difficulty of performing the behaviour and is assumed to reflect past experience as well as anticipated impediments and obstacles. PBC was used as proxy measures of actual control. There was overlap in definition of PBC with Bandura’s (1977) definition of self efficacy (i.e. ‘the conviction that one can successfully execute the behaviour required to produce the outcomes’).15, 16 PBC also had a direct influence on behaviour (Figure 7). Figure 7 Model of the theory of planned behaviour Source: Ajzen (1991)15 VOL 1– 14 Social cognition models: a review of their relevance for understanding participation in cancer screening The relative importance of attitude SN and PBC in the prediction of intention was expected to vary across behaviours and situations. However, as a general rule, the more favourable the attitude and SN and the greater the PBC, the stronger would be an individual’s intention to perform the behaviour. The TPB depicts behaviour as a linear regression function of behavioural intention and perceived behavioural control. B = w1I + w2PBC where B is behaviour, I is behavioural intention, PBC is perceived behavioural control and w1 and w2 are regression weights.16 According to Ajzen (1991), accurate prediction of behaviour requires that: 1. measures of intention and PBC correspond to, or are compatible with, the behaviour that is to be predicted 2. intention and PBC remain stable in the interval between their assessment and observation of the behaviour; intervening events may produce changes in intentions or PBC with the effect that the original measures of these variables no longer permit accurate prediction of behaviour 3. PBC must be an accurate proxy for actual behavioural control.15 Connor and Sparks (2005) presented a revised TPB model that still contained the original constructs, but included a construct called ‘actual control’ and outlined how other beliefs and external variables (demographic, personality and environment) influence attitude, SN and PBC (Figure 8).16 Figure 8 Revised Theory of Planned Behaviour model EXTERNAL
VARIABLES
BELIEFS
Demographic
Behavioural
Attitude Towards
the Behaviour
Personality
Characteristics
Normative
Subjective
Norm
Environmental
Influences
Control
Intention
Behaviour
PBC
Actual
Control
Source: Based on Connor and Sparks (2005)16 Notes: PBC: Perceived behavioural control According to this model, intention does not necessarily result in the desired behaviour. In addition, the concordance between intention and behaviour depends on the: Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 15  PBC of the person, based on an appraisal of internal control factors, such as beliefs about having the necessary skills and resources to perform the behaviour  actual external factors that affect the performance of the behaviour.19 As noted by Peyman and Oakley (2009) both internal and external factors can facilitate or hinder behaviour.20 3.2 Utility of the Theory of Planned Behaviour Model If the TPB is to be used to develop an intervention, it is important to determine which variables should be targeted (i.e. attitude, SN, PBC, or actual control factors). It would be counter‐productive to target variables that did not account for variance in intention or behaviour.16 There have been a number of studies investigating the relationships between the variables in the TPB and intentions. 3 For example, Connor and Sparks (2005) quote an unpublished meta‐analysis by McEachan, Connor and Lawton that included 12 prospective applications of the TPB to breast or testicular self‐examination (5 studies), cervical screening (2 studies) and health screening (5 studies). Across all the studies:  44% of the variance in intention was explained by attitude, SN and PBC  attitude was the strongest predictor (r+ = 0.56), PBC the second strongest (r+ = 0.43), and SN the weakest (r+ = 0.34). 4 Vart (2010) conducted a survey of 334 men and women aged ≥ 60 years in the UK as part of her, as yet, unpublished doctoral thesis. Preliminary findings indicated that the reasoning behind the intention to participate in FOBT was significantly affected by individuals’ attitudes towards bowel cancer and bowel cancer screening.21 A meta‐analysis by Cooke and French (2008) indicated that:  the strength of SN–intention and PBC–intention relationships varied across type of screening test (Table 2 Estimated effect sizes by type of screening test)  receiving an invitation increased the strength of the attitude–intention and SN–
intention relationships but decreased the PBC–intention relationship (Table 3)  having to pay for screening decreased the strength of all the relationships but the decrease was outside statistical significance for the SN–intention relationship at p ≤ 0.05 (Table 3).17 3
4
The relationship between intentions (I) and behaviour (B) are considered in Chapter 3 r+ = sample weighted average correlation VOL 1– 16 Social cognition models: a review of their relevance for understanding participation in cancer screening Table 2 Estimated effect sizes by type of screening test Variable Screening mammography Pap tests A‐I r+ 0.46 0.43 0.43 SN‐I r+ 0.30* 0.43* 0.52* PBC‐I r+ 0.45* 0.58 0.62 Relationship FOBT 17
Source: Cooke and French (2008) Notes: * Denotes correlations within the row that are significantly different from one another (p < 0.05) FOBT: Faecal occult blood testing r+: Sample‐weighted average correlation A–I: Attitude–intention relationship SN–I = Subjective norm–intention relationship PBC–I = Perceived behavioural control–intention relationship Table 3 Estimated effect sizes by program factors Invitation No invitation p value r+ 95% CI r+ 95% CI A–I 0.55 0.53–0.56 0.39 0.36–0.42 < 0.001 SN–I 0.39 0.38–0.41 0.44 0.41–0.47 < 0.001 PBC–I 0.44 0.42–0.45 0.57 0.53–0.61 < 0.01 Free Paid p value A–I 0.53 0.51–0.54 0.42 0.38–0.45 < 0.001 SN–I 0.40 0.39–0.42 0.43 0.39–0.46 > 0.05 PBC–I 0.47 0.45–0.48 0.38 0.33–0.43 < 0.001 17
Source: Cooke and French (2008) Notes: r+: Sample‐weighted average correlation A–I: Attitude–intention relationship SN–I: Subjective norm–intention relationship PBC–I: Perceived behavioural control–intention relationship Finally, the evaluation of the 2000/2001 phase of the BreastScreen Australia media campaign indicated differences in attitudes towards the provision of screening mammography by BreastScreen Australia between well‐screened, under‐screened and never‐screened women. Although over 90% of women in each group agreed that BreastScreen Australia screening was a good idea (99% vs 93% vs 91%), the strength of the attitude was lower in the under‐screened and never‐screened groups; 84% and 70%, respectively thought it was a ‘very good idea’ compared to 96% of the well‐screened groups (Figure 9). Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 17 Figure 9 Attitudes of Australian women towards screening by BreastScreen Australia 100%
Percentage
of Respondents
80%
60%
40%
20%
0%
Well-Screened
Under-Screened
Never-Screened
Very Good Idea
96%
84%
70%
Fairly Good Idea
3%
9%
21%
Bad Idea
0%
4%
2%
Don't Know
0%
4%
7%
13
Source: King et al. (2003) Notes: Percentages may not sum to 100% due to rounding Sample sizes not included in the original VOL 1– 18 Social cognition models: a review of their relevance for understanding participation in cancer screening 4 INTENTION–BEHAVIOUR RELATIONSHIP Intentions can be defined as the instruction that people give themselves to perform particular behaviours in order to achieve certain goals. They are characteristically measured by items of the form ‘I intend to do/achieve X’. Intentions are regarded as the culmination of a decision‐making process.22‐25 Godin and Kok (1996) reviewed the literature applying the TPB to attendance at screening programs. Across six studies, they reported sample‐weighted average correlations (r+) between intentions and behaviours of 0.35 – a moderate effect according to the Cohen (1992) scale in footnote 3.12, 26 In 2002, Sheeran published the results of a meta‐analysis of ten individual meta‐analyses of the intention–behaviour (I–B) relationship and found that intention accounted for 28% of the variance in behaviour (r = 0.53). This was a ‘large’ effect size and suggested that ‘intentions are good predictors of behaviour’.27 However, it is difficult to gauge from correlational studies such as this the impact that a change in intention (I∆) would have on behaviour (B∆). In 2006, Webb and Sheeran published a meta‐analysis of intervention studies that assessed both I∆ and subsequent B∆. The forty‐five included studies contained forty‐seven tests of the I∆–B∆ relationship. A major finding of the meta‐analysis was that a medium‐to‐large I∆ (d+=0.66) engendered a small‐to‐medium B∆ (d+=0.36). 5 The authors concluded that intention has a significant impact on behaviour but the effect is considerably smaller than correlational tests have suggested.23, 28 Webb and Sheeran (2006) investigated the impact of various factors that may mediate the I∆– B∆ relationship and, as shown in Table 44, found that the impact of I∆ on behaviour was:  larger when (i) participants had more actual or perceived control over the health behaviour; (ii) the interventions resulted in significant I∆ rather than significant, simultaneous I∆ and PBC∆; (ii) the behaviour was measured objectively rather than by self‐report  smaller when (i) circumstances supported the development of habitual behaviour; (ii) the study involved risky behaviours performed in social contexts; (iii) the interval between intention and behaviour measures was greater than 11.5 weeks; (iv) the comparison group received no intervention. The I∆–B∆ relationship was not mediated by (i) the way in which intentions were measured; (ii) the publication status of the study (peer‐reviewed vs grey literature); (iii) the nature of the participants (students vs non‐students).23 5
12
The magnitude of the effect sizes was based on Cohen's (1992) recommendations for d shown in footnote 3.
Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 19 VOL 1–20 Social cognition models: a review of their relevance for understanding participation in cancer screening 23
Non‐student 0.33 0.38 0.25 0.30 0.46 0.45 0.74 0.38 0.55 0.32 0.25 d+ Student Unpublished Alternative Objective > 11.5 Likely Likely BE I + PBC High High Rating Level 2 0.38 0.25 0.41 0.67 0.23 0.19 0.22 0.35 0.33 0.51 0.45 d + No significant differences in the relationship between student and non‐student samples No significant differences in the relationship between published and unpublished studies I∆ had a smaller impact on behaviour when there was no comparison intervention (i.e. comparison group received no treatment) I∆ had a greater impact on behaviour when behaviour was measured objectively I∆ had smaller effect on behaviour when the time interval between intention and behaviours was greater than 11.5 weeks I∆ has smaller effect on behaviour in the case of risky behaviours performed in social contexts compared with health‐protective behaviours I∆ had smaller effect on behaviour when circumstances supported the development of habits The type of intention measure (intention (BI) vs expectation (BE)) did not influence the I∆–B∆ relationship I∆ had a larger effect on behaviour when interventions produced significant I∆ only compared to those that generated significant I∆ and PBC∆ I∆ had a larger effect on behaviour when participants perceived they had more control over the behaviour I∆ had a larger effect on behaviour when participants were rated as possessing more control over the behaviour Interpretation of the results VOL 1– 20 Social cognition models: a review of their relevance for understanding participation in cancer screening Notes: ∆: Change; d+: Sample weighted effect size; PBC: Perceived behavioural control; BE: Behavioural Expectation ; I: Intention; B: Behaviour Source: Webb and Sheeran (2006) Type of sample Publication status Published None Comp. intervention STUDY CHARACTERISTICS Self‐report Behaviour measure Time interval ≤ 11.5 Unlikely MEASUREMENT FACTORS Unlikely Social reaction I Habitual control Intention measure I Only PBC effect Low Low PBC Assessed control Volitional control Rating Level 1 Conceptual factors moderating the intention change (I∆)–behaviour change (B∆) relationship CONCEPTUAL FACTORS Table 4 Webb and Sheeran (2006) also examined the impact of the use of the different SCMs to develop the interventions, and the impact of different types of interventions on both intentions and behaviours.  As shown in Table 55, there were no statistically significant differences between the impact of the TPB (and its forerunner the TRA) and the HBM. Both had medium effects on intention and small effects on behaviour.  As shown in Table 6, incentives for behaving or remaining in a program had the largest impact on intention but there was no difference between incentives and social encouragement/pressure/support on behaviour. However, both these types of interventions had a larger impact on behaviour than the other types of intervention.23  Although it has been argued that planning and forming implementation intentions narrows the gap between intention and behaviour,24 the intervention involving planning and implementation intentions had the third largest effect on intention but the smallest effect on behaviour (Table 66).23 Table 5 Estimated effect sizes for intervention based on theoretical models Sample weighted effect sizes (d+) Intention Behaviour Protection Motivation Theory 0.69 0.46a Theory of Reasoned Action/Theory of Planned Behaviour 0.58 0.40a Health Belief Model 0.52 0.29a 0.53 0.22a Stage Models (MAP, TTM, HAPA, ARRM) 23
Source: Webb and Sheeran (2006) Notes: Stage models: Model of Action Phases (MAP), Transtheoretical Model (TTM), Health Action Process Approach (HAPA), Aids Risk Reduction Models (ARRM) a: no statistically significant differences between the effect sizes
Table 6 Estimated effect sizes for types of interventions aimed at changing behaviour Effect sizes 1
Incentives Intention Behaviour 0.96 0.58a Social encouragement/pressure/ support 0.64 0.54a Information on behaviour/outcome 0.60 0.32b Persuasive communication 0.38 0.29b Modelling/demonstration 0.41 0.28b Environmental changes 0.77 0.27b Personalised message 0.10 0.26b Risk awareness material 0.56 0.25b 0.68 0.20c Planning, implementation 23
Source: Webb and Sheeran (2006) Notes: 1 Incentives for behaving or remaining in the program a,b,c: similar subscripts: no statistically significant difference; different subscripts: statistically significant differences Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 21 There was a wide range of behaviour included in the Webb and Sheeran (2008) analysis (e.g. condom use, smoking, course enrolment, sun protection, exercise, testicular self‐
examination, cycle helmet use, breast self‐examination, donating behaviour) and care needs to be taken in generalising the results to all types of behaviour. Cooke and French (2008) undertook a quantitative review of applications of the TRA/TPB to screening attendance. Their review included nineteen tests of the relationship between intention and behaviour in the peer‐reviewed literature. Their results indicated that intention differed in effectiveness as a predictor of behaviour depending on the screening test; intention was a better predictor of behaviour for colorectal cancer screening (not exclusively FOBT) and screening mammography than for Pap tests (Table 77). They also found that, across all screening tests, intention was a better predictor of attendance in paid contexts than in free contexts (Table 88).17 Table 7 Estimated effect sizes for type of screening test for intention behaviour relationships Sample weighted average correlation r+ Participants Number of tests 8148 19 0.42 240 2 0.44 Screening mammography 4587 6 0.37 Cervical 1013 4 0.212 Overall (all types) 1
Colorectal Source: Cooke and French (2008) 17 Notes: 1 Not exclusively faecal occult blood testing 2 Difference between cervical and other screening tests statistically significant; no statistically significant difference between colorectal and mammography screening r+: sample weighted average correlation Table 8 Comparison of relationship between behaviour and intentions by cost of screening Number of tests(1) r+ 95% CI Paid 4 0.58 0.53–0.61 Free 15 0.40 0.38–0.41 <0.001 p value Sample weighted average correlation 17
Source: Cooke and French (2008) Notes: 1 Types of screening included in the analysis were: cervical, colorectal, genetic test, health check, mammography, and prenatal. There was significant heterogeneity in the r+ values between tests but the r+ values for each type of test were not reported In a conceptual and empirical review of the intention–behaviour relationship, Sheeran (2002) divided people conceptually into four groups according to their intention and subsequent behaviour. Conceptually it was the ‘disinclined actors’ and the ‘inclined abstainers’ who were responsible for the discrepancy between intention and behaviour (Table 9). VOL 1– 22 Social cognition models: a review of their relevance for understanding participation in cancer screening Table 9 Conceptual schema of actors and abstainers in the intention–behaviour relationship Subsequently screened Subsequently not screened Inclined actors Inclined abstainers Disinclined actors Disinclined abstainers Intention to screen Do not Intend to screen 27
Source: Sheeran (2002) Empirically, Sheeran (2002) found that, except for screening mammography, the proportion of inclined abstainers (those people who intended to screen but subsequently did not) was larger than the 'disinclined screeners' (those who did not intend to screen but subsequently did so).27 In one study of FOBT in France a similar effect was also seen (Table 100).29 Table 10 Empirical results for actors and abstainers in the intention–behaviour relationship Inclined to screen (100%) Screened Not screened Cervical 43% Cervical 70% Breast 74% FOBT 48% 27
Do not intend to screen (100%) Not screened Screened 57% 88% 12% 30% 100% 0% 26% 65% 35% 52% 82% 27% 29
Sources: Cervical and Breast Sheeran (2002) ; FOBT Herbert et al. (1997) Notes: FOBT: Faecal occult blood testing Although these results are interesting, they should be treated with some caution. They are based on a small number of studies, and, as shown in Figure 10, people do not always fit neatly into a dichotomised version of intentions. In Australia, there is a not insignificant ‘unsure’ group.13 Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 23 Figure 10 Breast cancer screening: differences in the intentions between never‐screened and under‐screened Australian women 100%
Percentage
of Respondents
80%
60%
40%
20%
0%
Never-Screened
(N=61)
M ore than 2 years since last screen
(N=57)
Likely
33%
62%
Unlikely
58%
27%
Others
8%
12%
Source: King et al. (2003)13 Notes: Post‐hoc statistical significance test based on three response categories undertaken by CHPPE: Chisq = 12.31, df = 2, p = 0.002; Others: Neither likely nor unlikely + don't know
VOL 1– 24 Social cognition models: a review of their relevance for understanding participation in cancer screening 5 IMPLICATIONS AND CONCLUSIONS As was shown in Volume 2 of the literature review, a broad range of factors have been shown to be associated with uptake of cancer screening. These include demographic factors, cognitive factors, psychosocial factors, health and lifestyle factors, cultural factors and health system factors. SCMs have been used to understand and identify reasons for low compliance rates with cancer screening. The justifications that have been advanced for focusing on cognitions (cognitive) are that cognitions are:  important determinants that mediate other determinants, such as demographic factors  more amenable to change than other factors including psychological factors. These justifications imply that effective interventions should be based on manipulations of the cognitive factors that have been shown to determine health behaviours.1, 5A number of issues need to be considered in relation to how the models can be used to help design interventions. 5.1 Issues impacting on the utility of the models  Integrity of the underlying assumptions Both the models presented in this review emphasise the rationality of human behaviour. Intentions in the TPB and behaviour in the HBM are considered to be the end result of a rational decision‐making process based upon a ‘deliberative, systematic processing of the available information’. The most transparent examples of this underlying assumption are the threat and behavioural assessments in the HBM.1 It cannot be judged from the evidence presented whether or not this is an adequate model for screening behaviour. However, the following quote by Lende and Lachiondo (2009) indicates that models based on cost–benefit analysis may not be appropriate. ‘We saw a sharp divergence between what a person understands about breast cancer and screening and how that person actually deals with these things on a personal level. It was usually easy to talk about others in purely biomedical terms. That afforded a distance and a perspective from which it was easy to explain what needs to be done to maintain health. However, this biomedical model became clouded when the conversation turned personal, forcing women to leave behind bodies in general and focus on their own body specifically. Each woman brought an embodied understanding of breast cancer to the table, as the experiences they had in their bodies and the subsequent meanings they held for their bodies guided many of their thoughts about their bodies… Consequently, their explanations moved well beyond a cost–benefit analysis that weighed time and money against the upside of early detection.’30 Both the models are health‐oriented. They assume that health decisions and behaviour are fuelled by motives to protect one’s health and to regulate threats associated with health. But there is another line of research in which perceptions of the ‘self’ (not health) assume a major role in influencing health judgements and Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 25 behaviours. A small number of findings suggest that people’s conceptions of their physical bodies can have implications for health behaviours. For example, research by Burris and Rempel (2004) indicates that people’s sensitivity to activities involving penetration of the boundaries of the physical self predicts decreased blood donation.31 Although no research specifically relating to cancer screening was included in the article to demonstrate this point, the authors indicate that the findings suggest that there may be psychological implications associated with the physicality of many health behaviours that will have an impact on screening behaviour.32  Specificity of the constructs As Yarbrough and Braden (2000) pointed out in relation to the HBM, the theoretical descriptions of the key constructs in the models (e.g. barriers, benefits, risk) are not specific enough to use for targeting interventions. If, for example, an intervention were to be implemented targeting ‘attitudes’ towards screening behaviour, there is nothing in the model to suggest what the key attitudes are and how they can be changed. The analyses for the impact of different types of interventions on intentions and behaviour in the TPB indicated that they are not all equal. For example, social encouragement/pressure/support would appear to have a greater impact on intentions than either persuasive communications or personalised messages. However, this was discovered empirically and cannot be adduced from the model.  Relationships between the constructs There is very little in the descriptions of the models and the accompanying analyses to determine how the constructs (and underlying factors) combine to produce behaviour change. Yet there is no reason to assume that they will combine either additively or multiplicatively to produce a desired change in behaviour. For example, there was an example in the analysis for the TPB which indicated that interventions that resulted in significant changes in intentions had a larger impact on behaviour than those that produced significant changes in both intentions and perceived behavioural control.23 There are a number of instances where the data analyses relating to the relationship between intentions and behaviour indicated that the relationships between the constructs varied according to: (i) the type of screening test; (ii) factors associated with the health system (e.g. issuing screening invitations and the cost of the test). It cannot be automatically assumed, therefore, that the evidence obtained in one health system will generalise to another with different structural features. 5.2 Conclusions Overall, the models are interesting and provide a way of visualising or thinking about how to intervene to increase screening rates, but cannot provide the finer details. On balance, the analysis in relation to the TPB appears to have somewhat greater utility for designing interventions. Some tentative conclusions arising from this model are as follows:  interventions that succeed in changing intentions to screen will also succeed in changing behaviour but the impact on behaviour will be much smaller than on intentions  people who are inclined to screen but who, for whatever reason, do not ultimately attend screening, are the major reason for the lack of concordance between intentions and behaviour  whether the intervention should target attitudes, social norms or perceived behavioural control (self‐efficacy) will depend to some extent on the screening test VOL 1– 26 Social cognition models: a review of their relevance for understanding participation in cancer screening  there is nothing to indicate that targeting all three constructs will produce a greater effect and a very small amount of data to indicate that it may be counter‐productive  issuing invitations to screen appears to strengthen the relationship between attitudes and intentions. Financial incentives would appear to have the largest impact on intentions but both incentives and social encouragement/pressure/support appear to have the greatest impact on behaviour. Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 27 REFERENCES 1. Connor M, Norman P. Predicting Health Behaviour: A Social Cognition Approach. In: Conner M, Norman P, Eds. Predicting Health Behaviour. Maidenhead: Open University Press; 2005. 2. Conner M, Norman P, eds. Predicting Health Behaviour. Second ed. New York: Open University Press; 2005. 3. Yarbrough SS, Braden CJ. Utility of health belief model as a guide for explaining or predicting breast cancer screening behaviours. Journal of Advanced Nursing 2001;33(5):677‐688. 4. Rosenstock I. The health belief model and preventive health behavior. Health Education Monographs 1974(2):354‐386. 5. Austin LT, McNally M, Stewart DE. Breast and cervical cancer screening in Hispanic women: a literature review using the health belief model. Women's Health Issues 2002;12(3):122‐128. 6. Abraham C, Sheeran P. The Health Belief Model. In: Conner M, Norman P, Eds. Predicting Health Behaviour. Maidenhead: Open University Press; 2005:28‐80. 7. Bandura A. Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, New Jersey: Prentice‐Hall, Inc; 1986. 8. Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the health belief model. Health Education Quarterly 1988;15(2):175‐183. 9. Glanz K, Rimer BK. Theory at a Glance: A Guide for Health Promotion Practice. Washington, DC: US Department of Health and Human Services, Public Health Service, National Institutes of Health; 1995. 10. Janz NK, Becker MH. The health belief model: A decade later. Health Education Quarterly 1984;11:1‐47. 11. Harrison JA, Mullen PD, Green LW. A meta‐analysis of studies of the Health Belief Model with adults. Health Education Research 1992;7(1):107‐116. 12. Cohen J. A Power Primer. Psychological Bulletin 1992;112(1):155‐159. 13. King E, Ball J, Carroll T. Evaluation Report for the 2000/2001 Phase of the BreastScreen Australia Campaign. Canberra: Australian Government, Department of Health and Ageing; 2003. Screening Monograph No. 1/2004. 14. Department of Health and Ageing. Bowel Cancer Knowledge, Perceptions and Screening Behaviours: Knowledge, Attitudes and Practices, Pre‐and Post‐Intervention Surveys (2002 and 2004). Canberra: Australian Government, Department of Health and Ageing; 2004. Screening Monograph No. 4/2005. 15. Ajzen I. The theory of planned behavior. Organizational Behavior and Human Decision Processes 1991;50(2):179‐211. 16. Connor M, Sparks P. Theory of Planned Behaviour and Health Behaviour. In: Conner M, Norman P, Eds. Predicting Health Behaviour: Research and Practice with Social Cognition Models. Second Ed. Maidenhead: Open University Press; 2005. 17. Cooke R, French DP. How well do the theory of reasoned action and theory of planned behaviour predict intentions and attendance at screening programmes? A meta‐analysis. Psychology & Health Oct 2008;23(7):745‐765. 18. Sirur R, Richardson J, Wishart L, Hanna S. The role of theory in increasing adherence to prescribed practice. Physiotherapy Canada 2009;61(2):68‐77. 19. Ajzen I, Madden T. Prediction of goal‐directed behaviour: attitudes, intentions, and perceived behavioural control. Journal of Experimental Social Psychology 1986;22:453‐474. VOL 1– 28 Social cognition models: a review of their relevance for understanding participation in cancer screening 20. Peyman N, Oakley D. Effective contraceptive use: an exploration of theory‐based influences. Health Education Research 2009;24(4):575‐585. 21. Vart GF. Factors Influencing the Decision to Participate in Bowel Cancer Screening. Surrey, University of Surrey, 2010. 22. Sheeran P, Milne S, Webb TL, Gollwitzer PM. Implementation Intentions and Health Behaviour. In: Conner M, Norman P, Eds. Predicting Health Behaviour. Maidenhead: Open University Press; 2005. 23. Webb TL, Sheeran P. Does changing behavioral intentions engender behavior change? A meta‐analysis of the experimental evidence. Psychological Bulletin 2006;132(2):249‐268. 24. Gallo IS, Gollwitzer PM. Implementation intentions: A look back at fifteen years of progress. Psicothema 2007;19(1):37‐42. 25. Gollwitzer PM, Sheeran P, Mark PZ. Implementation Intentions and Goal Achievement: A Meta‐Analysis of Effects and Processes. Advances in Experimental Social Psychology. Vol Volume 38: Academic Press; 2006:69‐119. 26. Godkin G, Kok G. The theory of planned behavior: A review of its applications to health‐
related behaviors. American Journal of Health Promotion. 1996;11:87‐98. 27. Sheeran P. Intention‐Behavior Relations: A Conceptual and Empirical Review. European Review of Social Psychology 2002;12:1 ‐ 36. 28. McCaul KD, Branstetter AD, Glasgow RD, Schroeder DM. What is the relationship between breast cancer risk and mammography screening? A meta‐analytic review. Health Psychology. 1996;15:423‐429. 29. Herbert C, Launoy G, Gignoux M. Factors affecting compliance with colorectal cancer screening in France: differences between intention to participate and actual participation. European Journal of Cancer Prevention 1996;6:44‐52. 30. Lende DH, Lachiondo A. Embodiment and breast cancer among African American Women. Qualitative Health Research 2009;19(2):216‐228. 31. Burris CT, Rempel JK. "It's the end of the world as we know it": Threat and the spatial‐
symbolic self. Journal of Personality and Social Psychology 2004;86:19‐42. 32. Goldenberg JL, Arndt J. The Implications of Death for Health: A Terror Management Health Model for Behavioral Health Promotion. Psychological Review 2008;115(4):1032‐1053. Social cognition models: a review of their relevance for understanding participation in cancer screening VOL 1– 29 VOL 1– 30 Social cognition models: a review of their relevance for understanding participation in cancer screening VOLUME 2 IDENTIFYING HARD‐TO‐REACH GROUPS: A REVIEW OF THE FACTORS (INCLUDING BARRIERS) ASSOCIATED WITH CANCER SCREENING TABLE OF CONTENTS TABLE OF CONTENTS ............................................................................................................................. 2 LIST OF TABLES ...................................................................................................................................... 4 LIST OF FIGURES .................................................................................................................................... 5 1. INTRODUCTION .......................................................................................................................... 6 2 METHODOLOGY.......................................................................................................................... 8 2.1 2.2 3 DEMOGRAPHIC FACTORS .......................................................................................................... 10 3.1 3.2 3.3 3.4 3.5 4 ACCULTURATION ..........................................................................................................................59 FATALISM ...................................................................................................................................61 MODESTY/EMBARRASSMENT/SHAME ..............................................................................................62 MEDICAL MISTRUST ......................................................................................................................63 COLLECTIVISM/COMMUNALISM ......................................................................................................64 SPIRITUALITY/RELIGIOSITY ..............................................................................................................64 HEALTH SYSTEM FACTORS ........................................................................................................ 66 8.1 8.2 8.3 8.3 9. PHYSICAL/MENTAL HEALTH ............................................................................................................51 DRUG USE: ALCOHOL AND TOBACCO CONSUMPTION ...........................................................................54 PREVIOUS PARTICIPATION IN SCREENING ...........................................................................................54 Cultural FACTORS...................................................................................................................... 59 7.1 7.2 7.3 7.4 7.5 7.6 8 FEAR/ANXIETY/WORRY .................................................................................................................45 COPING STYLE (EMOTION REGULATION)............................................................................................46 SOCIAL NETWORKS .......................................................................................................................46 HEALTH AND LIFESTYLE............................................................................................................. 51 6.1 6.1 6.3 7 LITERACY ....................................................................................................................................31 HEALTH LITERACY .........................................................................................................................33 CANCER‐RELATED KNOWLEDGE .......................................................................................................36 PERCEIVED RISK ...........................................................................................................................38 PERCEIVED AMBIGUITY ..................................................................................................................40 PSYCHOSOCIAL FACTORS .......................................................................................................... 44 5.1 5.2 5.3 6 AGE ..........................................................................................................................................10 GENDER .....................................................................................................................................16 ETHNICITY ..................................................................................................................................18 SOCIOECONOMIC STATUS ..............................................................................................................22 EDUCATION ................................................................................................................................27 COGNITIVE FACTORS................................................................................................................. 30 4.1 4.2 4.3 4.4 4.5 5 LOCATING THE LITERATURE ..............................................................................................................8 REVIEW STRATEGY ..........................................................................................................................8 AVAILABILITY ...............................................................................................................................66 ACCESSIBILITY..............................................................................................................................70 AFFORDABILITY ............................................................................................................................72 ACCEPTABILITY ............................................................................................................................74 DISCUSSION AND IMPLICATIONS............................................................................................... 77 VOL 2– 2 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 9.1 9.2 DISCUSSION ................................................................................................................................77 IMPLICATIONS .............................................................................................................................81 REFERENCES ........................................................................................................................................ 84 APPENDIX: CANCER SCREENING LITERATURE REVIEWS ........................................................................ 90 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 3 LIST OF TABLES TABLE 1 TABLE 2 TABLE 3 TABLE 4 TABLE 5 TABLE 6 TABLE 7 TABLE 8 TABLE 9 TABLE 10 TABLE 11 TABLE 12 TABLE 13 TABLE 14 TABLE 15 TABLE 16 TABLE 17 TABLE 18 TABLE 19 TABLE 20 TABLE 21 TABLE 22 TABLE 23 TABLE 24 TABLE 25 TABLE 26 TABLE 27 TABLE 28 TABLE 29 TABLE 30 TABLE 31 TABLE 32 TABLE 33 TABLE 34 TABLE 35 INITIAL NUMBER OF CITATIONS JUDGED PERTINENT TO THE THREE REVIEWS .....................................................8 CLASSIFICATION OF FACTORS ASSOCIATED WITH CANCER SCREENING UPTAKE ...................................................9 UPTAKE OF CANCER SCREENING TESTS AND AGE: JEPSON ET AL. (2000) REVIEW ............................................10 UK: UPTAKE OF FOBT SCREENING (FEB 2003 – NOV 2004) BY AGE IN THE SECOND ROUND OF THE ENGLISH PILOT OF BOWEL CANCER SCREENING ...................................................................................................15 AUSTRALIA: ESTIMATED PARTICIPATION RATES FOR FOBT SCREENING BY AGE (2008) ....................................16 DIFFERENCES IN MALE AND FEMALE FOBT PARTICIPATION RATES: USA, UK AND AUS ...................................17 FEMALE–MALE DIFFERENCES IN UPTAKE OF FOBT SCREENING IN THE SECOND ROUND OF THE ENGLISH PILOT OF BOWEL CANCER SCREENING (FEB 2003–NOV 2004)..........................................................................17 CANCER SCREENING RATES BY ETHNICITY: USA (2005).............................................................................18 UPTAKE OF CANCER SCREENING TESTS AND RACE/ETHNICITY: JEPSON ET AL. (2000) REVIEW ............................19 AGE STANDARDISED PARTICIPATION RATES IN BREASTSCREEN AUSTRALIA: VICTORIAN WOMEN AGED 50–69 YEARS ...................................................................................................................................20 MAMMOGRAM UPTAKE BY ETHNIC AND NON‐HISPANIC WHITE PARTICIPANTS: PURC‐STEPHENSON ET AL. (2008) REVIEW ................................................................................................................................20 PARTICIPATION RATES IN BREASTSCREEN SOUTH AUSTRALIA BY ETHNIC GROUPS ............................................21 UPTAKE OF CANCER SCREENING TESTS AND SOCIOECONOMIC STATUS: JEPSON ET AL. (2000) REVIEW ................23 RELATIONSHIP BETWEEN FINANCIAL FACTORS AND UPTAKE OF SCREENING MAMMOGRAPHY: SCHUELER ET AL. (2008) REVIEW ................................................................................................................................23 AGE STANDARDIZED PARTICIPATION RATES IN BREASTSCREEN AUSTRALIA BY SOCIOECONOMIC STATUS OF AREA OF RESIDENCE (2005‐2006) ...................................................................................................................24 AGE STANDARDIZED PARTICIPATION RATES OF WOMEN AGED 20–69 YEARS IN THE AUSTRALIAN NATIONAL CERVICAL SCREENING PROGRAM, BY SOCIOECONOMIC STATUS OF AREA OF RESIDENCE (2005–2006)...............25 CRUDE PARTICIPATION IN AUSTRALIAN NBCSP BY SOCIOECONOMIC STATUS OF AREA OF RESIDENCE (2008).......25 SECOND ROUND OF THE ENGLISH PILOT OF BOWEL CANCER SCREENING: UPTAKE OF FOBT SCREENING BY DEPRIVATION CATEGORY OF AREA OF RESIDENCE (FEB 2003–NOV 2004) ....................................................26 SCREENING RATES AND EDUCATION: USA (2005) ...................................................................................27 RELATIONSHIP BETWEEN EDUCATION AND SCREENING UPTAKE: JEPSON ET AL. (2000) REVIEW .........................27 PERCENTAGE OF VARIANCE (R SQ) IN LITERACY PROFICIENCY ACCOUNTED FOR BY 12 PREDICATOR VARIABLES: AUSTRALIA AND FINLAND ....................................................................................................................31 MULTIPLE REGRESSION MODELLING OF COMPREHENSION OF INTERNET‐BASED COLORECTAL CANCER INFORMATION
34 ASSOCIATION BETWEEN FUNCTIONAL HEALTH LITERACY AND CANCER‐RELATED KNOWLEDGE .............................35 RELATIONSHIP BETWEEN HEALTH LITERACY AND UNDERSTANDING OF MEDICAL LEAFLETS (N = 321)...................35 RELATIONSHIP BETWEEN CANCER RELATED KNOWLEDGE AND SCREENING UPTAKE: JEPSON ET AL. (2000) REVIEW ............................................................................................................................................36 RELATIONSHIP BETWEEN BREAST CANCER KNOWLEDGE ON SCREENING MAMMOGRAPHY UPTAKE: SCHUELER ET AL. (2008) REVIEW ................................................................................................................................37 RELATIONSHIP BETWEEN PERCEPTIONS OF RISK AND SCREENING UPTAKE: JEPSON ET AL. (2000) REVIEW ............38 RELATIONSHIP BETWEEN MARITAL STATUS AND SCREENING UPTAKE: JEPSON ET AL. (2000) REVIEW ..................48 ASSOCIATION BETWEEN PERCEIVED HEALTH STATUS AND THE UPTAKE OF SCREENING: JEPSON ET AL. (2000) REVIEW .............................................................................................................51 IMPACT OF BREAST DISEASE ON SCREENING MAMMOGRAPHY UPTAKE: SCHUELER ET AL. (2008) REVIEW ............52 HEALTH FACTORS ASSOCIATED WITH LOWER UPTAKE OF COLORECTAL CANCER SCREENING TESTS: PETERSEN (2002) REVIEW ................................................................................................................................53 AUSTRALIAN NATIONAL BOWEL CANCER SCREENING PROGRAM (NBCSP): CRUDE PARTICIPATION RATE BY DISABILITY STATUS (2008) ..................................................................................................................54 ASSOCIATION BETWEEN PREVIOUS SCREENING BEHAVIOUR AND THE UPTAKE OF SCREENING: JEPSON ET AL. (2000) REVIEW ................................................................................................................................55 CANCER SCREENING RATES BY IMMIGRATION STATUS: USA (2005).............................................................60 IMPACT OF ACCULTURATION ON THE MAMMOGRAM UPTAKE: SCHUELER ET AL. (2008) REVIEW .......................61 VOL 2– 4 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening TABLE 36 GEOGRAPHIC VARIATIONS IN SCREENING RATES IN AUSTRALIA ....................................................................67 TABLE 37 GEOGRAPHIC VARIATION IN SCREENING RATES IN THE USA ........................................................................67 TABLE 38 FIVE‐YEAR SCREENING RATES FOR NHS BREAST SCREENING PROGRAMME: UPTAKE OF INVITATIONS TO SCREEN FOR WOMEN AGED 50–70 YEARS IN SELECTED GEOGRAPHIC AREAS .............................................................68 TABLE 39 GEOGRAPHIC VARIATION IN BREASTSCREEN VICTORIA 2‐YEAR PARTICIPATION RATES (50–69 AGE STANDARDISED RATES)......................................................................................................68 TABLE 40 RELATIONSHIP BETWEEN AVAILABILITY AND THE UPTAKE OF MAMMOGRAMS: SCHUELER ET AL. (2008) REVIEW ...69 TABLE 41 COMPARISON OF DIFFICULTY IN THE STEPS NECESSARY TO OBTAIN A PAP TEST WITH WELL‐SCREENED WOMEN .....70 TABLE 42 RELATIONSHIP BETWEEN ACCESSIBILITY FACTORS AND UPTAKE OF SCREENING MAMMOGRAPHY: SCHUELER ET AL. (2008) REVIEW .........................................................................................................71 TABLE 43 UPTAKE OF CANCER SCREENING TESTS AND INSURANCE STATUS: JEPSON ET AL. (2000) REVIEW ........................72 TABLE 44 ASSOCIATION BETWEEN HEALTH INSURANCE AND THE UPTAKE OF CANCER SCREENING IN THE USA (2005) ...........................................................................................................................................73 TABLE 45 RELATIONSHIP BETWEEN FINANCIAL FACTORS AND UPTAKE OF SCREENING MAMMOGRAPHY: SCHUELER ET AL. (2008) REVIEW ................................................................................................................................73 TABLE 46 EFFECT OF COST ON WILLINGNESS TO JOIN A COLORECTAL CANCER SCREENING PROGRAM IN HONG KONG (YOUNG AND OLD AGE GROUPS ONLY)....................................................................................................74 TABLE 47 SUMMARY OF THE EVIDENCE ................................................................................................................78 LIST OF FIGURES FIGURE 1 FIGURE 2 FIGURE 3 FIGURE 4 FIGURE 5 PARTICIPATION IN THE BREASTSCREEN AUSTRALIA PROGRAM (FEMALES 2005‐2006) ...................................11 SCREENING MAMMOGRAPHY RATES BY AGE: VIC (2006–07), USA (2005), AND UK (2009) ........................12 LONG‐TERM SCREENING MAMMOGRAPHY AND PAP TEST MONITORING DATA FOR VICTORIA .............................12 PAP TEST SCREENING RATES BY AGE: VICTORIA (2006–07), USA (2005), AND UK (2009)............................14 MODEL OF THE RELATIONSHIP BETWEEN FEAR AND BREAST CANCER SCREENING BEHAVIOUR: CONSEDINE ET AL. (2004) REVIEW ................................................................................................................................45 FIGURE 6 ASSOCIATION BETWEEN SOCIAL NETWORK SIZE AND SCREENING MAMMOGRAM STATUS ..................................47 FIGURE 7 ASSOCIATION BETWEEN PREVIOUS CANCER SCREENING ACTIVITIES AND THE UPTAKE OF FAECAL OCCULT BLOOD TESTING ...........................................................................................................................................57 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 5 1. INTRODUCTION The Victorian Cytology Service (VCS) was commissioned by the Victorian Department of Health to undertake a review of international literature (peer‐reviewed and grey literature) focusing on key areas that inform thinking, knowledge and approaches to engaging individuals, groups and communities to participate in organised programs, including population‐based screening programs. The aim of the review was to consider:  factors that enhance or hinder engagement in health screening and other preventative programs, including perceptions held by participants  needs of specific groups (e.g. information needs of specific groups, such as those with low literacy skills)  health beliefs and their impact on cognitive constructs and behaviour in relation to screening and engagement in positive health behaviours  structural and environmental barriers to participation  key work in other disciplines or settings that provide insight into innovative thinking and practice in engaging individuals and communities; especially within organised, voluntary programs or initiatives  work completed in health and other disciplines on translating research findings into action. The VCS contracted the Centre for Health Policy, Programs and Economics (CHPPE) to undertake the identification, synthesis and summary of the evidence. The review is presented as three stand‐alone documents. Each volume has a specific focus:  volume 1 provides an overview of the theoretical models underpinning thinking in relation to participation in screening programs.  volume 2 focuses on the factors and barriers to participation in screening programs  volume 3 focuses on the impact of interventions designed to increase participation in organised screening/health programs. According to Consedine et al. (2004), screening rates have been linked to a wide range of factors, including background variables such as age, socioeconomic status, and education. Despite the large amount of research that has been generated in the consideration of these background variables, Consedine et al. (2004) believe that it is important to undertake research into other factors that may have an impact on screening. In part, this is because the research into these background factors helps to identify those at risk for a poor screening profile. However, they offer little direction in terms of viable interventions. Consedine et al. (2004) believe that improvements in screening will require interventions that target factors that are amenable to change.1 For this reason, this volume looks at a range of factors that have been found to be associated (either positively or negatively) with the uptake of screening for breast, cervical and colorectal cancer. VOL 2– 6 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Section 2 of this report provides an overview of the methodology, including a typology of the factors developed to structure the report. Sections 3 to 8 contain the analysis in relation to socio‐demographic factors, cognitive factors, psychosocial factors, health and lifestyle factors, cultural factors and, finally, health system factors. The report concludes with a summary of the evidence and a discussion of the results (Section 9). Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 7 2 METHODOLOGY 2.1 Locating the literature A range of databases and indexes available through the University of Melbourne library were investigated to determine search terms for the literature review that would give a broad, multi‐
disciplinary coverage, coupled with the most efficient search and retrieval capabilities. The five databases included in the search were: 1. Medline: medical research literature 2. ISI Web of Science/ISI Web of Knowledge: Science Citation Expanded®, Social Sciences Citation Index®, and Arts & Humanities Citation Index™ 3. CINAHL: nursing, biomedicine, health sciences, consumer health and seventeen allied health disciplines in journals, books and book chapters, dissertations, selected conference proceedings, standards of practice, pamphlets, educational software packages and audiovisual material 4. SCOPUS: medical and social science literature 5. Cochrane Library: systematic reviews of health care treatments and interventions. Broad search terms used to locate the screening literature included combinations of the following: (breast cancer OR cervical cancer OR colorectal cancer OR bowel cancer OR colon cancer); screen*; (participation OR engagement). Articles were also located by a scrutiny of reference lists and grey literature databases identified by the Steering Committee and through an environmental survey undertaken by the VCS. Only articles in English and published in the period 1999–2009 were included. The original database comprised just over 1700 articles (Table 1). During the preparation of the first two reviews, the database grew to over 2000 references, over 1500 of which were pertinent to a review of factors associated with the uptake of screening mammography, Pap tests and faecal occult blood testing (FOBT). Table 1 Initial number of citations judged pertinent to the three reviews Strategy Peer‐reviewed literature: Number Database search Scrutiny of reference lists and Steering Committee Grey literature Database searches and Steering Committee Total 2.2 % 1491 86% 165 10% 69 4% 1725 100% Review strategy Because of time and resource constraints it was not possible to base this volume of the review for the VCS on over 1500 primary studies. Therefore, it was based primarily on monitoring data and literature reviews published between 1999 and 2009 inclusive. The reviews were of factors associated with uptake of screening for breast, cervical or colorectal cancer and/or cancer VOL 2– 8 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening screening in general. Because of the lack of reviews relating to FOBT, two reviews published outside this timeframe were included; one by Vernon published in 1997 and one by Guessous et al. published in 2010.2, 3 A review published in 1996 relating to screening mammography has also been included. It was referred to in a review published within the given timeframe and was pertinent to the analysis.4 A total of 26 cancer screening reviews were included (see Appendix for details). The search strategies used to locate the primary studies contained in the 26 reviews included citations published from 1872 to 2009. The focus of the reviews was as follows.  46% (12) related to breast cancer only  4% (1) related to cervical cancer only  12% (3) related to breast and cervical cancer  12% (3) related to colorectal cancer only,  19% (5) related to breast, cervical and colorectal cancer  8% (2) to cancer in general. One of the reviews relating to colorectal cancer published in 1997 focused exclusively on FOBT.2 The other two (one published in 2002 and the other in 2010) focused on factors associated with colorectal cancer screening and not exclusively on FOBT.3, 5 However, in the review by Guessous et al. (2010), 71% of the primary studies reported on FOBT alone or in combination with other screening.3 Most, but not all, the reviews were based on empirical studies – those that contained univariate and multivariate analyses of the relationships between particular factors and screening rates. For this reason, the VCS review has not explored qualitative data in great depth. The reliance on literature reviews and the plethora of factors included in the reviews also means that the review provides a broad overview of all factors rather than an in‐depth analysis of each. Categorising the factors associated with cancer screening uptake
For the VCS review, the actors associated either positively or negatively with cancer screening have been classified under six domains: demographic, cognitive, psychosocial, health and lifestyle, cultural and health system factors (Table 2). Table 2 Classification of factors associated with cancer screening uptake Domain Factors Demographic factors Age, gender, ethnicity, socioeconomic status, education Cognitive factors Literacy, health literacy, cancer‐related knowledge, perceived risk, perceived ambiguity Psychosocial factors Fear/anxiety/worry, coping style (emotion regulation), social networks Health and lifestyle factors Physical/mental health, drug use, previous participation in screening Cultural factors Acculturation, fatalism, modesty/embarrassment/shame, medical mistrust, collectivism/communalism, spirituality/religiosity Health system factors Availability, accessibility, affordability, acceptability Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 9 3 DEMOGRAPHIC FACTORS Demographic data relate to the characteristics of a population. Commonly used demographics include sex, race, age, income, disability, mobility (in terms of travel time to work or number of vehicles available), educational attainment, home ownership, employment status and location. 1 The demographic factors included in this review were: age, gender, ethnicity (including Indigenous status), socioeconomic status and education. Data relating to these factors come from two sources:  monitoring data for population‐based screening programs published on the internet in the form of Excel spreadsheets and in grey literature reports  systematic reviews included in the peer‐reviewed literature, supplemented, where appropriate, with a small number of recently published primary studies. 3.1 Age As shown in Table 3, just under two‐thirds (61%) of the 31 studies relating to screening mammography included in the Jepson et al. (2000) review did not find an association between mammogram uptake and age. However, over three‐quarters of the studies relating to Pap tests and the studies relating to FOBT reported a significant association. 2 The authors stated that the results relating to Pap tests were difficult to interpret. However, for FOBT, ‘many’ studies showed that uptake increased with age.6 The evidence in relation to each of the screening tests is considered in more detail below. Table 3 Uptake of cancer screening tests and age: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with 12/31 studies (39%) statistically significant effects (p ≤ 0.05) 7/9 studies (78%) 7/9 studies (78%) Authors’ conclusion The majority of studies found an association between uptake and age. However, they reported conflicting effects. It was not clear whether older or younger women were more likely to attend The majority of studies found a significant association between uptake of colorectal cancer screening and age. In many the return of FOBT was highest among participants aged ≥ 65 years The majority of studies did not find an association between uptake and age, and those that did reported conflicting results. It was not clear whether older or younger women were more likely to attend Source: Jepson et al. (2000)6 Notes: FOBT: Faecal occult blood testing 1
2
http://en.wikipedia.org/wiki/Demographics See the Appendix for an overview of the 26 original literature reviews. VOL 2– 10 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening  Screening mammography BreastScreen Australia actively recruits women between the ages of 50 and 69 (inclusive). As shown in Figure 1, there were relatively high participation rates for the target group in the period 2005‐2006. – 54% to 59% compared to 19% for 45‐49 year old women and 32% for 70‐74 year olds. 500
65
400
55
300
45
200
35
100
25
0
Number of Females (000s)
% of Females
45-49
50-54
55-69
60-64
65-69
70-74
140.69
364.52
360.14
287.07
230.07
105.72
19
54
58
60
59
32
% Screened
Participation in the BreastScreen Australia Program (females 2005‐2006) Number of Females (000s)
Figure 1 15
Source: Australian Institute of Health and Welfare (2009) 7 USA monitoring data are highly aggregated and indicate that screening is highest among the 50–64 year age group. UK and Victorian monitoring data are more disaggregated. These data indicate that screening rates are lowest for those who become eligible to enter screening programs (i.e. women in their early 50s) with a slight diminution for 65–69 year olds as women move out of the target group (Figure 2). 3 Long‐term Victorian data indicate that, although there has been some variation in uptake by different age groups over time, uptake across the age groups has remained relatively stable (Figure 3). 3
There is no centrally organized screening program in the USA, but the American Cancer Society indicates that it is important women age 40 years of age and older have annual mammograms at an accredited mammography screening facility.8 In the UK, the NHS Breast Screening Programme provides free breast screening for all women in the UK aged 50 and over. Women aged between 50 and 70 are routinely invited. (http://www.cancerscreening.nhs.uk/breastscreen/#whatdoes)
Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 11 Figure 2 Screening mammography rates by age: VIC (2006–07), USA (2005), and UK (2009) 59
53
56
50–54
55–59
64
58
76
72
76
77
75
55-59
60-64
65-69
64
46
60–64
65–69
40-49
VIC (1)
50-64
65+
50-52
53-54
USA (2)
UK (3)
Sources: VIC (Victoria): Productivity Commission Excel spreadsheet downloaded 15 October 2009 from http://www.pc.gov.au/__data/assets/excel_doc/0004/85414/53‐chapter12‐attachment‐only.xls; USA: Cokkinides et al. (2007)8; UK: Table 2 of National Health Service Excel spreadsheet downloaded on 24 Dec 2009 from http://www.ic.nhs.uk/webfiles/publications/breastscreening0708/2007‐08 Breast Screening Tables.xls Notes: (1) Participation in BreastScreen Australia program over a 24‐month period; (2) Mammogram in past 2 years; (3) Less than 3 years since last test Figure 3 Long‐term screening mammography and Pap test monitoring data for Victoria 80
75
70
65
60
55
50
45
20–24
25–29
30–34
35–39
40–44
45–49
50–54
55–59
60–64
65–69
Mammography 2002-03
55.4
59.4
60.2
58.1
Mammography 2006-07
53.1
55.8
59
57.8
Pap Tests 1996-97
48.5
60.2
66.4
68.4
69.4
71.8
70.4
73.8
64.9
56.2
Pap Tests 2006-07
47.3
57.9
63.8
66.6
67.6
71.5
70.6
75
65.2
57.9
Sources: Mammography ‐ Productivity Commission Excel spreadsheet downloaded 15th October 2009, available at http://www.pc.gov.au/__data/assets/excel_doc/0004/85414/53‐chapter12‐attachment‐only.xls, Pap tests – AIHW (2009)9 Notes: Residents of Victorian postcodes allocated to the Albury/Wodonga catchment (NSW jurisdiction) are included in Victoria’s population estimate, accounting for the slight decrease in participation rates compared to those published by BreastScreen Victoria VOL 2– 12 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening In the Wells et al. review of factors associated with mammogram uptake by Latin American women in the USA published in 2007: o 57% of studies (8/14) reported that age was a significant predictor of screening mammography in multiple logistic regression analysis o 43% of studies (6/14) did not find an association between the uptake of mammograms and age. The authors indicated that it was difficult to interpret the results because the studies used different age ranges in the regression analyses.10 
Pap tests In Australia, the National Cervical Screening Program promotes screening for women between 18 (or two years after first sexual intercourse, whichever is later) and 69 years. 4 Victorian monitoring data indicate that Pap test participation rates peak in the 45–59 age group and then decrease (Figure 4) Although some variations have been seen over time, differences across age groups are relatively consistent (Figure 3). In the USA, there is no centrally organised program and the most recent American Cancer Society cervical cancer screening guidelines are shown in Box 1. USA survey data published in 2007 (before the new guidelines were issued) indicate that the uptake of Pap tests appeared to peak in the 30–39 year age group, followed by a decline in participation (Figure 4). Box 1 Current American Cancer Society cervical cancer screening guidelines (2010) All women should begin cervical cancer screening about 3 years after they begin having vaginal intercourse, but no later than 21 years old. Screening should be done every year with the regular Pap test or every 2 years using the newer liquid‐based Pap test. Beginning at age 30, women who have had 3 normal Pap test results in a row may get screened every 2 to 3 years. Women older than 30 may also get screened every 3 years with either the conventional or liquid‐based Pap test, plus the human papilloma virus (HPV) test. Women 70 years of age or older who have had 3 or more normal Pap tests in a row and no abnormal Pap test results in the last 10 years may choose to stop having Pap tests. Source: http://www.cancer.org/docroot/ped/content/ped_2_3x_acs_cancer_detection_guidelines_36.asp In the UK all women between the ages of 25 and 64 are eligible for a free Pap test every three to five years, depending in their age. 5 The data indicate that Pap test participation rates increase up to the 40–59 year age group. However, because the recommended screening interval differs for the 25–49 group and the 50–64 year age groups, rates across the whole range cannot be compared (Figure 4). 4
5
http://www.cancerscreening.gov.au/internet/screening/publishing.nsf/Content/cervical‐about http://www.cancerscreening.nhs.uk/cervical/index.html
Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 13 65
58
59
62
75
77 78
80
83 80
77
60-64
71
50-54
82 79
50-64
71
87
75
25-29
58
64 67
68 72
89
40-44
85
35-39
Pap test screening rates by age: Victoria (2006–07), USA (2005), and UK (2009) 65+
Figure 4 VIC (1)
USA (2)
55-59
45-49
30-34
60-64
50-59
40-49
30-39
21-29
65–69
60–64
55–59
50–54
45–49
40–44
35–39
30–34
25–29
20–24
47
UK (3)
Sources: VIC: AIHW (2009)9 8
USA: Cokkinides et al. (2007) UK: Table 1of National Health Service Excel spreadsheet downloaded on 25 Nov 2009 from http://www.ic.nhs.uk/webfiles/publications/cervscreen0809/Cervical_ Screening_Programme_2008_09_Data_Tables.xls Notes: (1) Victorian 2‐year participation rates by age in the National Cervical Screening Program (2) In 2005, proportion of women who had a Pap test within the previous 3 years (3) Coverage of the NHS Cervical Screening Programme in England as at 31st March 2009. For ages 25–48, coverage represents the proportion of women with an adequate test result who have had a test within the last 3 years and for ages 50–64, coverage represents the proportion within the last 5 years Faecal occult blood testing A review of the factors associated with the uptake of FOBT published in 1997 contained thirteen studies that reported the relationship between age and uptake. Seven (54%) measured prospective behaviour and reported that uptake was lowest among persons aged 70 years and over. Three studies (23%) found that completion was higher in younger groups and three (23%) found no consistent pattern.2 In 2005 in the USA, the rate of having an FOBT within the past year was 10.6% for those aged 50–64 and 13.8% for those aged 65 years and over.8 The relatively low FOBT screening rates in the USA are due to the fact that there is no centrally organised screening program, and the current American Cancer Society screening guidelines do not unequivocally state that people aged 50 and over should have yearly FOBT (Box 2). VOL 2– 14 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Box 2 Current American Cancer Society colorectal cancer screening guidelines Beginning at age 50, both men and women should follow one of these testing schedules: (i) Tests that find polyps and cancer Flexible sigmoidoscopy every 5 years*, or Colonoscopy every 10 years, or Double‐contrast barium enema every 5 years*, or CT colonography (virtual colonoscopy) every 5 years* (ii) Tests that primarily find cancer Yearly faecal occult blood test (gFOBT)**, or Yearly faecal immunochemical test (FIT) every year**, or Stool DNA test (sDNA), interval uncertain** * If the test is positive, a colonoscopy should be done. ** The multiple stool take‐home test should be used. One test done by the doctor in the office is not adequate for testing. A colonoscopy should be done if the test is positive. The tests that are designed to find both early cancer and polyps are preferred if these tests are available to you and you are willing to have one of these more invasive tests. Source: http://www.cancer.org/docroot/ped/content/ped_2_3x_acs_cancer_detection_guidelines_36.asp In the UK, FOBT uptake increased with increasing age for both women and men in the second round of the English pilot (Table 4). Table 4 UK: Uptake of FOBT screening (Feb 2003 – Nov 2004) by age in the second round of the English Pilot of Bowel Cancer Screening Age (years) Males % Adj OR
<55 41.3 1 55–59 46.9 1.23 60–64 51.0 65–69 56.2 Females 1
95% CI % Adj OR 50.2 1 1.18–1.28 56.2 1.26 1.20–1.31 1.47 1.41–1.54 60.0 1.49 1.43–1.56 1.82 1.74–1.91 60.6 1.55 1.48–1.63 95% CI 11
Source: Weller et al (2006) Downloaded from the NHS website on 26 November 2009 http://www.cancerscreening.nhs.uk/bowel/pilot‐2nd‐round‐evaluation.pdf Notes: Adj OR: Adjusted odds ratio; CI: Confidence interval The monitoring data for people invited to participate in the Australian National Bowel Cancer Screening Program (NBCSP) also indicated that participation rates increased with age – 50 years 31.8%, 55 years 38.9% and 65 years 47.7% at 26 weeks (Table 5). Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 15 Table 5 Australia: Estimated participation rates for FOBT screening by age (2008) 50 years 55 years 65 years 31.8 38.9 47.7 31.6–32.0 38.7–39.1 47.5–48.0 26 weeks People participating (per 100 invitations) (1) 95% CI 52 weeks (1) (2)
People participating (per 100 invitations) 39.3 48.1 95% CI 39.1–39.5 47.8–48.3 12
Source: Australian Institute of Health and Welfare and Department of Health and Ageing (2009) Notes: (1) Participation rates equal the estimated Kaplan‐Meier participation rate of people who returned a completed FOBT kit as a proportion of the total number of the eligible population who were invited to screen, excluding people who suspended or opted off the Program. (2) People aged 50 years were invited to screen from 1 July 2008, so no 52‐week data were available.. 3.2 Gender Faecal occult blood testing In 1997, Vernon’s review of factors associated with the uptake of FOBT indicated that the impact of gender differed across different methodologies:  prospective studies (15 studies): participation was higher for women in 13 studies (87%), and for men in 2 studies (13%)  experimental interventions (11 studies): women were more likely than men to complete FOBT, although the differences were not pronounced and, when tested, were not always statistically significant  other studies (5 studies): data from these studies, including national surveys of recent past behaviour, showed inconsistent patterns by gender.2 In the 2000 review of factors associated with FOBT screening, Jepson et al. (2000) found that the majority of studies (6 of 7 studies; 86%) did not find an association between uptake and gender.6 In a review of factors associated with participation in colorectal cancer screening published in 2002, Petersen stated that males generally had lower adherence to colorectal cancer screening.5 In their 2010 review of factors associated with colorectal cancer screening in people aged 65 years of age and over, Guessous et al. reported that in 14 of 26 studies (54%) ‘female sex’ was a significant barrier to participation, while two studies (8%) reported it to be a significant facilitator. Although not stated, it is assumed that the remaining 10 studies (38%) found no association between gender and the uptake of colorectal cancer screening.3 As shown in Table 6, very small gender differences in the uptake of FOBT were reported in the USA for 2005; the rate for females was slightly lower than for males. In the second round of the English pilot of bowel cancer screening, uptake was higher among females than males (OR 1.42, 95% CI 1.36–2.48) (Table 6). The gender differences in the UK appeared to be greater in the younger age groups (Table 7). There was also a statistically significant difference in participation VOL 2– 16 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening between the sexes in the National Bowel Cancer Screening Program in Australia – males 36.0% (95% CI 35.5‐35.8), females 42.6% (95% CI 42.0‐42.4) (Table 6). Table 6 Differences in male and female FOBT participation rates: USA, UK and AUS Country FOBT participation Male Female Female – male difference USA FOBT within past year (2005) 12.7% 11.7% –1.0% UK Uptake in second round of English Pilot of Bowel Cancer Screening (Feb 2003–Nov 2004) 47.7% 56.1% +8.4%(1) AUST Uptake by people invited to participate in NBCSP (2008) 36.0% 42.6% +6.6%(2) Sources: USA Cokkinides et al. (2007)8 UK Source: Weller et al (2006)11 Downloaded from the NHS website on 26 November 2009 http://www.cancerscreening.nhs.uk/bowel/pilot‐2nd‐round‐evaluation.pdf Notes: FOBT: Faecal occult blood testing (1) Difference between female and male participation rates was statistically significant: OR 1.42, 95% CI 1.36 – 1.48, p < 0.001 (2) At 52 weeks since invitation. Table 7 Female–male differences in uptake of FOBT screening in the second round of the English Pilot of Bowel Cancer Screening (Feb 2003–Nov 2004) Males Females Female–male difference <55 41.3 50.2 +8.9% 55–59 46.9 56.2 +9.3% 60–64 51.0 60.0 +9.0% 56.2 60.6 +4.4% Age (years) 65–69 11
Source: Weller et al (2006) Downloaded from the NHS website on 26 November 2009 http://www.cancerscreening.nhs.uk/bowel/pilot‐2nd‐round‐evaluation.pdf Gender differences in other health‐seeking behaviour A review of 124 key gender‐comparative health‐seeking studies published between 1966 and 2004 did not fully support the hypothesis that men are less likely than women to seek help when they experience ill health. Many studies noted the relative under‐use of health services and symptom reporting by men compared with women, while others found an increase in help‐
seeking in men compared with women, or no significant difference in help‐seeking behaviour between genders.13 According to Galdas et al. (2005), the contradictory nature of the findings and the ‘sex‐
differences’ approach in these investigations fail to provide an explanation for the actual differences between men and women in help‐seeking patterns. Neither do they address the issue of within‐group variability. The authors identified one review of gender‐specific research that revealed a clearer picture of potential factors influencing help‐seeking behaviour in men that supported the proposition that males delay seeking help.13 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 17 Galdas et al. (2005) also identified an increasing body of research literature pointing towards cultural stereotypes (e.g. ‘traditional masculinity’ and/or ‘masculinity beliefs’) as a significant variable influencing the help‐seeking behaviour and health‐risk appraisal of men when they become ill. Theories prevalent in the international men’s health literature contend that men: (i) are not permitted to be expressive in their illness behaviour; or (ii) are unable to be expressive because of the ‘construction of traditional masculinity’ and efforts to conform to a ‘socially prescribed male role’ where weakness and need for help are not believed to be masculine. As a result of these beliefs it is likely that men will react differently to health care services and health promotion messages, both in comparison with women and between men of differing age, social and ethnic groups.13 The role of cultural factors influencing men’s utilisation of health care services was espoused by Woods (2001) as one of the explanations accounting for the lower usage rate of GP services by Australian men. This explanation suggests that our culture conveys different values regarding health to each gender, and that men have not been encouraged to place the same premium on health than women do. To support this explanation, Woods (2001) quoted a study of men in rural Queensland which indicated that health only becomes a priority for men once it is under threat from illness or injury. These men equated health with ‘being able to work’.14 An alternative hypothesis locates the problem of under‐utilisation of health services by men in the structure of the health care system including: the location (availability), convenience (accessibility), ‘male friendliness’ (appropriateness) and affordability of health care services. A survey of men in Western Australia regarding the factors that militated against their use of GP services indicated that the main reason why men were reluctant to access GP services was the amount of time spent in waiting rooms. Lesser reasons included: negative perceptions of GP knowledge and skills; feeling ‘uncomfortable’; cost; and restricted surgery hours. These findings are supported in a further Australian study by Aoun and Johnson (2000).14‐16 Finally, Smith et al. (2006) have pointed out that it is the pattern of help‐seeking and health service utilisation by men that differs; for men the initial approach tends to be indirect. Men tend to see their partners and friends as primary resources for help.17 3.3 Ethnicity Monitoring data from the USA indicate that there are differences in participation in cancer screening by ethnicity. White (non‐Hispanic) groups are more likely to participate in screening than other ethnic and Indigenous groups (Table 8). Table 8 Cancer screening rates by ethnicity: USA (2005) Mammogram past 2 years Pap test past 3 years White (non‐Hispanic) 68.1% 81.4% 12.6% American Indian, Alaskan Native 66.6% 75.8% 5.8% African American (non‐Hispanic) 64.9% 80.2% 10.3% Hispanic/Latin American 59.6% 74.5% 9.4% 54.2% 65.8% 10.8% Ethnicity Asian FOBT within past year 8
Source: Cokkinides et al. (2007) VOL 2– 18 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening As shown in Table 9, very few of the studies included in the Jepson et al. review published in 2000 indicated a significant association between race or ethnicity and the uptake of mammograms, Pap tests and FOBT.6 The evidence in relation to each of these screening tests is considered in more detail below. Table 9 Uptake of cancer screening tests and race/ethnicity: Jepson et al. (2000) review Factor Screening mammograms Pap tests FOBT Studies with statistically significant effects (p < 0.05) Black: 3/15 studies (20%) African American: 1/15 studies (7%) White: 1/15 studies (7%) Native American: 1/15 (7%) Race: 0/2 Nationality: 0/1 0/5 studies Authors’ conclusions The majority of studies did not find a significant association between uptake and ethnic origin but several studies reported higher uptake among black women. Insufficient studies (i.e. determinants not investigated in three or more studies) to draw a conclusion None of the studies found a significant association between uptake and ethnicity. Source: Jepson et al. (2000)6 Notes: FOBT = Faecal occult blood testing  Screening mammography Monitoring data indicate that the age‐standardised participation rate in the BreastScreen Australia program for non‐Indigenous women aged 50 to 69 years was 57% (95% CI 56.6‐
56.8) in the 200502006 compared with 38% (95% CI 37.3‐38.9) for Indigenous women. Women who mainly spoke English at home were more likely than other women to have participated in the BreastScreen Australia program in the 2005‐20006 period (59.1%, 95% CI 59.0‐59.2 vs 44.8, 95% CI 44.6‐45.0).7 Table 10 shows the participation rates for Victorian women aged 50–69 years in BreastScreen Australia program for the period 2002‐2033 to 2006‐2007. A number of observations can be made about these data: o participation rates for Indigenous women and those from non‐English speaking backgrounds (NESBs) are lower than the rate for all women o during the period rates declined across all groups o the decline for Indigenous women (–14.8%) was more marked than for women from NESBs (‐3.3%); the reasons for the marked decrease among Indigenous women are not clear. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 19 Table 10 Age standardised participation rates in BreastScreen Australia: Victorian women aged 50–69 years Period All Indigenous NESB 2002–2003 58.0% 50.9% 39.8% 2003–2004 58.6% 45.1% 39.7% 2004–2005 57.7% 37.3% 38.4% 2005–2006 56.8% 35.9% 37.5% 2006–2007 55.9% 36.1% 36.5% Change: 2002–03 to 2006–07 –2.1% –14.8% –3.3% th
Source: Productivity Commission Excel spreadsheet downloaded 15 October 2009, available at http://www.pc.gov.au/__data/assets/excel_doc/0004/85414/53‐chapter12‐attachment‐only.xls Notes: Residents of Victorian postcodes allocated to the Albury/Wodonga catchment (NSW jurisdiction) are included in Victoria’s population estimate, accounting for the slight decrease in participation rates compared to those published by BreastScreen Victoria NESB = Non‐English speaking background Lower screening mammography participation rates for Indigenous populations are also evident in the monitoring data for New Zealand. Information supplied by BreastScreen Aotearoa, indicated that coverage for the 24 months to the end of April 2009 for eligible women aged 50–69 was: (i) 51.5% for Mãori women; (ii) 54.7% for Pacific women; and (iii) 67.0% for ‘other’ women.18 In a review of Canadian and US studies, Purc‐Stephenson et al. (2008) found evidence that ethnic minority‐screening differences exist but the differences were ‘impacted’ by socioeconomic status. As shown in Table 11, when the odds ratios (ORs) were adjusted for socioeconomic status, African Americans, Hispanic and Asian/Pacific Islander women were just as likely as non‐Hispanic White women to obtain a mammogram.19 Table 11 Mammogram uptake by ethnic and non‐Hispanic White participants: Purc‐
Stephenson et al. (2008) review Adjusted OR (1) Unadjusted OR Number of studies OR 95% CI Number of Studies OR 95% CI African Americans 28 0.87 0.75–1.00 7 1.05 0.71–1.76 Hispanic 18 0.65 0.50–0.85 7 1.08 0.64–1.93 Asian/Pacific Islander 10 0.63 0.40–0.99 8 0.57 0.31–1.27 Source: Purc‐Stephenson and Gorey (2008)19 Notes: (1) Adjusted for socioeconomic status; OR: Odds ratio These results confirm the conclusion of the review by Sadler et al. (2001), which identified several studies reporting that when confounding factors are taken into account, ethnicity was not a significant factor in determining screening behaviour. However, other factors that VOL 2– 20 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening are thought to predict low screening adherence, such as not having an identified medical provider, having limited education and having a low income level, are closely aligned with many minority groups. Therefore a focus on ethnic groupings will reach groups most likely to experience those other factors. The authors also cautioned that it was important to be mindful of the diversity within diversity and, that it was not always possible to generalise results from one ethnic or Indigenous groups to another.20 Australians from culturally and linguistically diverse (CALD) backgrounds make up approximately 13% of the population and are from more than one hundred countries of origin.21 Data from BreastScreen South Australia show that participation rates differ by ethnic group (Table 12).22 Table 12 Participation rates in BreastScreen South Australia by ethnic groups BreastScreen South Australia participation rate (%) Ethnic group Greek 58% Chinese 58% Spanish‐speaking 57% Italian 54% Polish 53% German 52% Former Yugoslavia (Bosnia, Croatia, Herzegovina, Serbia) 50% Dutch 48% Filipino 45% Khmer 41% Russian 40% (1)
Pacific Islander: Melanesia (Age‐standardised rate) 36% (1)
Pacific Islander: Polynesia (Age‐standardized rate) 28% 22
Source: Cancer Council SA, SA Cervix Screening Program, BreastScreen SA (2008) Notes: (1) The data in this table are presented as they appeared in the original reference. It is unclear if all the rates shown are age‐standardised or only these two.  Pap tests According to Mullins (2006), ‘the Victorian Cervical Cytology Registry (VCCR) does not have information about the country of birth or the language spoken by women who have Pap tests. However, data from the National Health Survey (ABS 2001) indicated that women who speak a language other than English at home are less likely to have regular Pap tests (43%) than women who speak English at home (57%)’.23 Akers et al. (2007), state that studies have 'consistently shown' that screening rates are lower among immigrant groups compared with the general population in the USA. They concluded that for cervical cancer specifically, women from immigrant groups appear to be less likely to undergo screening.24 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 21 
Faecal occult blood testing In the Vernon (1997) review, three studies reported associations between race or ethnicity and adherence. However, data from national surveys of self‐reported past screening behaviour showed minimal or no differences between Whites and African Americans, as did data from four other studies. The conclusion in the Vernon 91997) review was that the majority of studies did not find an association between ethnicity and FOBT uptake. 2 The 2009 report of the Australian NBCSP indicated that the crude participation rate for FOBT was 2.2 times higher for non‐Indigenous invitees compared to Indigenous invitees and the result was statistically significant (23.5%, 95% CI 23.4‐23.6 vs 10.5%, 95% CI 9.9‐
11.2). However, a close reading of the data tables on which this result was based indicated that the 'data in the table … is considered unreliable'. The crude participation rate for people speaking English at home was 41.1% (95% CI 40.9‐41.2) and for those who speak a language other than English at home the crude participation rate was 14.0% (95%CI 13.8‐14.2). There are no cautions attached to these data.12 In 2002, Petersen (2002) stated that lower colorectal cancer screening adherence was generally seen among minority race or ethnic groups.5 The latest review of colorectal cancer screening published in 2010, reported that among people aged 65 years and over: o African American race was a significant barrier to screening in 14 of 26 (54%) studies, while 2 studies (8%) reported it to be a facilitator; although unstated, it appears that 10 studies (38%) found no association o Hispanic ethnicity was reported to be a barrier in 13 of 19 studies (68%); no studies found it to be a facilitator and, presumably, 6 studies (32%) found no association between Hispanic ethnicity and colorectal cancer screening uptake.3 The conclusion from the 2010 review was that the majority of studies found that ethnicity decreased the uptake of colorectal cancer screening among people aged 65 years of age and over.3 3.4 Socioeconomic status According to Fiscella et al. (2000) race/ethnicity and socioeconomic position in the United States are closely intertwined, and it is difficult to isolate racial/ethnic disparities in health care due primarily to socioeconomic disparities. They did however conclude that socioeconomic position appears to be the more powerful determinant of primary health care use in the United States.25 As shown in Table 13, Jepson et al. (2000) concluded that there were no significant associations between markers of socioeconomic status and the uptake of mammograms and FOBT, and the association with the uptake of Pap tests was not clear.6 The evidence for each test is considered in more detail below. VOL 2– 22 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Table 13 Uptake of cancer screening tests and socioeconomic status: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with statistically significant effects (p ≤ 0.05) No studies No studies reported No studies reported Authors' conclusions No significant association with uptake:  income  home ownership  employment status  occupation  partner's occupation Association with uptake unclear:  employment status  level of income No significant association with uptake:  occupation No significant association with uptake:  employment status  income Source: Jepson et al. (2000)6  Screening mammography In a review published in 2007, of the factors associated with the uptake of screening mammography among Latin American women in the USA, 4 studies reported that income was a significant predictor of screening mammography in multiple logistic regression analysis (adjusted OR 1.02–1.69). However, another 4 studies did not find this association.10 In the same review 5 studies evaluated the association between employment and the uptake of mammograms, but only 3 reported adjusted ORs. One of the three (33%) found that full‐time employment was associated with higher odds of screening mammography (adjusted OR 1.27). The other two (66%) did not find a statistically significant association.10 As shown in Table 14, a meta‐analysis by Schueler et al. (2008) found that having a low income or money worries reduced the likelihood of having a mammogram, but employment status did not have a statistically significant effect.26 Table 14 Relationship between financial factors and uptake of screening mammography: Schueler et al. (2008) review Number of studies Adjusted OR 95% CI Low income/money concerns 49 0.74 0.67–0.82 Not employed 17 0.89 0.77–1.03 Source: Schueler et al. (2008)26 Notes: OR: Odds ratio; CI: Confidence interval In Australia, monitoring data for BreastScreen Australia for the 2005‐2006 period indicated that women living in 'middle' socioeconomic status areas (59.4%) were more likely to participate than those living in low (57.1%‐57.6%) and high socioeconomic status areas (55.0%‐55.4%) (Table 15).7 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 23 Table 15 Age‐standardised participation rates in BreastScreen Australia by socioeconomic status of area of residence (2005‐2006) Socioeconomic Status of Area Age Standardised Rate 95% CI 1 (Lowest) 57.1% 56.9‐57.4(b) 2 57.6% 57.4‐57.8(b) 3 59.4% 59.2‐59.6(a) 4 55.4% 55.2‐55.7(c) 5 (Highest) 55.0% 54.8‐55.2(c) Source: Australian Institute of Health and Welfare and Department of Health and Ageing (2009)7 Notes: CI: Confidence interval (a), (b), (c) indicate overlapping confidence intervals  Pap tests According to Akers et al. (2007), the most common measures of socioeconomic position that have been examined for their association with cervical cancer screening rates include income, poverty level, educational status, and residence in socioeconomically disadvantaged areas. The authors state that ‘in general, these studies have found that higher socioeconomic status correlates with higher cervical cancer screening rates’. They also state that, in a 'number' of the studies, it has been noted that socioeconomic position explained differences in cervical cancer screening rates better than race/ethnicity.24 In a review of the factors associated with regular cervical screening among African American and Hispanic women it was noted that ‘several’ studies reported that the lower the income, the less likely it was that a woman would obtain a Pap test.27 Monitoring data for the Australian National Cervical Screening Program for the 2005‐2006 period indicated that women living in the highest socioeconomic areas were more likely to obtain a Pap test than those living in other areas. Women living in the lowest socioeconomic status areas were the least likely to obtain a Pap test (Table 16).28 VOL 2– 24 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Table 16 Age standardized participation rates of women aged 20–69 years in the Australian National Cervical Screening Program, by socioeconomic status of area of residence (2005–2006) Age Standardised Rate(1) 95% CI 5 (Lowest) 57.3 % 57.1–57.4 4 60.2 % 60.0–60.3 3 61.8 % 61.7–62.0 2 68.3 % 68.2–68.5 1 (Highest) 71.5 % 71.3–71.7 Socioeconomic Status of Area Source: Australian Institute of Health and Welfare and Department of Health and Ageing (2008)28 Notes: CI: Confidence interval (1) Tasmanian data were not included in the report  Faecal occult blood testing Data from the Australian NBCSP indicated that people living in the lowest socioeconomic status areas were less likely than those living in other areas to participate in the program. But there did not appear to be a consistent pattern in the association between uptake of FOBT and the socioeconomic status of the area in which the invitees lived (Table 17).7 Table 17 Crude participation in Australian NBCSP by socioeconomic status of area of residence (2008) Socioeconomic Status of Area Crude Participation Rate 95% CI 1 (Lowest) 34.6% 34.3‐34.8(a) 2 37.4% 37.1‐37.6(b) 3 36.8% 36.5‐37.0(c) 4 37.3% 37.1‐37.6(b) 5 (Highest) 37.4% 37.1‐37.7(b) Source: Australian Institute of Health and Welfare (2009)7 Notes: (a), (b), (c) indicate overlapping confidence intervals The evaluation of the first round of the UK Pilot reported that the uptake of the FOBT was close to its target of 60%. However, people living in the most deprived areas were less than half as likely to participate as those living in the least deprived areas (Table 18). Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 25 Table 18 Second round of the English Pilot of Bowel Cancer Screening: Uptake of FOBT screening by deprivation category of area of residence (Feb 2003–Nov 2004) Deprivation category (IMD) (1) Adjusted OR (95% CI) 5 Most deprived 0.41 (0.39–0.43) 4 0.60 (0.57–0.62) 3 0.74 (0.72–0.77) 2 0.86 (0.83–0.90) 1 Least deprived 1 11
Source: Weller et al (2006) Downloaded from the NHS website on 26 November 2009 http://www.cancerscreening.nhs.uk/bowel/pilot‐2nd‐round‐evaluation.pdf Notes: OR: Odds ratio; CI: Confidence interval; Adjusted for all other factors (2) IMD – The Index of Multiple Deprivation (IMD) 2004 is a measure of multiple deprivations at the small area level. The index contains seven domains of deprivation: income, employment, health and disability, education, skills and training, barriers to housing and service, living environment and crime. The overall IMD is conceptualised as a weighted area level aggregation of these specific dimensions of deprivation. In the Vernon review published in 1997, 5 of the 7 studies (71%) found a positive association between income and FOBT completion; higher income was associated with higher completion rates. Vernon (1997) noted that in the 2 studies that did not report an association, the income range was 'truncated'.2 In 2002, Petersen concluded that lower adherence to colorectal cancer screening was generally associated with lower income.5 In the latest review published in 2010 of the factors associated with colorectal cancer screening among older people (≥65), low socioeconomic status was reported as a significant barrier in only 7 of the 19 studies (37%). Socioeconomic status was not reported as a facilitator in any studies and, presumably, in the other 63% of studies there was no association.3 Time orientation According to Fiscella et al. (2000), the ‘pathways through which socioeconomic position and race/ethnicity affect health care are complex’. The authors indicate that the pathways are likely to include health care affordability, geographic access, transportation, education, knowledge, literacy, health beliefs, racial concordance between physician and patient, patient attitudes and preferences, competing demands including work and child care, and provider bias. The significance of any one of these factors is seen as varying by patient and physician.25 One of the factors that appears to have received some attention is the time orientation of low socioeconomic status groups. According to Deshpande et al. (2009), time orientation reflects a person's tendency to think and act according to consequences that are primarily immediate (i.e. present) or more distal (i.e. future). Deshpande et al. (2009) state that, by circumstance and necessity, people living in poverty tend to be present rather than future‐oriented.29 According to Wolff et al. (2003) African American underserved patients have difficulty prioritising preventive and screening activities in the context of more immediate survival needs such as food, shelter, safety and clothing. The stress associated with living in less‐than‐secure environments and existing on insufficient and unreliable resources makes focusing on preventive health care difficult. They state that ‘many’ underserved patients report feeling VOL 2– 26 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening isolated and alone, and heightened concerns about daily safety and well‐being are perceived as more prominent needs than worries about a disease they do not believe they have.30 In their review Deshpande et al. (2009) concluded that present time orientation was negatively associated with uptake of mammograms. One study reported that, overall, lower income black women tended to be more future‐ than present‐oriented, but high present‐orientation scores were negatively associated with the use of screening mammography even after adjusting for education and income.29 In the Australian context, Newman et al. (2008) have also pointed out the importance of looking at the context of people's lives in understanding Australian Aboriginal experiences of cancer. The authors stat that ‘health is often of little importance to those living with more immediate issues such as poverty, unemployment, poor housing, or violence’.31 3.5 Education As shown in Table 19, monitoring data indicate higher levels of education appear to be related to higher levels of participation in cancer screening activities in the USA.8 Table 19 Screening rates and education: USA (2005) Mammogram past 2 years Pap test past 3 years FOBT within past year 11 or fewer 53.0 68.0 8.9 12 64.4 77.0 11.2 13–15 69.1 83.7 13.8 76.8 88.1 15.3 Years of education ≥ 16 8
Source: Cokkinides et al. (2007) Notes: FOBT: Faecal occult blood testing Although the review by Jepson et al. (2000) concluded that the majority of studies did not find a significant association between education and uptake of screening tests (Table 20), later reviews have come to a different conclusion and these are discussed in more detail. Table 20 Relationship between education and screening uptake: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with statistically significant effects (p ≤ 0.05) 3/18 (17%) 1/7 (14%) 1/4 (25%) (1) Authors' conclusions The majority of studies did not find a significant association between uptake and level of education The majority of studies did not find a significant association between uptake and level of education The majority of studies did not find a significant association between uptake and the level of education Source: Jepson et al. (2000)6 Notes: (1) Only found to be significant in older women (> 65 years) FOBT: Faecal occult blood testing Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 27  Screening mammography In a review of studies involving Latin American women, Wells et al. (2007) concluded that ‘for the most part’, the studies reviewed reported that Latin American women with at least some college education were more likely to receive a screening mammogram (adjusted OR range 1.36–3.69). However, the authors noted that the evidence for the relationship between high school education and the uptake of mammograms was equivocal. Two studies found that a high school education was associated with higher odds of screening mammography compared with less than a high school education, but two other studies did not find that relationship.10 In the studies included in the Wells et al. (2007) review that reported an association between education and uptake of mammograms, education was entered into the multiple logistic regressions as a continuous variable or was categorised across several categories. In the studies that did not report an association, 4 did not report the adjusted OR, 3 dichotomised education near or at the completion of high school, and the remaining study sampled an immigrant population.10 The Scheuler et al. meta‐analysis published in 2008 included 52 studies that examined the relationship between education and the uptake of mammograms. Schueler et al. (2008) reported that, despite variability in the magnitude of effects, studies consistently demonstrated that low education levels correlated with a low uptake of mammograms (adjusted OR 0.78, 95% CI 0.73–0.83).26  Pap tests In the Ackerson et al. review published in 2007, it was concluded that ‘most’ studies found that lower levels of education were associated with decreased likelihood of Pap testing.27  Faecal occult blood testing In a review of the factors associated with the uptake of FOBT published in 1997, Vernon et al. reported that, with one exception, both prospective and retrospective studies found a positive association between higher education and completion of FOBT.2 In 2002, Petersen reported that lower participation in colorectal cancer screening tests were associated with less educational attainment.5 In the latest review of factors associated with colorectal cancer screening in older adults (≥ 65 years), published in 2010, 19 of 27 studies (70%) found that low education was a significant barrier to screening uptake and none found it was a significant facilitator.3 Presumably the other 30% of studies did not find an association Summary  Age There is evidence to suggest that age is associated with the uptake of screening but the patterns vary by screening type. For screening mammography and Pap tests, it appears that participation is lower in younger age groups (i.e. among those who become eligible to join population‐based programs). Participation in Pap testing appears to decline as women move towards the upper age limits for population‐based screening. Available VOL 2– 28 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening evidence indicates that the uptake of FOBT will increase with age up until about 70 years of age.  Gender The strongest evidence in relation to uptake of FOBT comes from an evaluation of the second phase of the UK pilot program. This study indicated that uptake was higher among women than men, particularly in younger age groups. Findings from peer‐
reviewed literature and a brief review of health‐seeking behaviour are less consistent. However, this literature did warn against treating men as a homogeneous group. On balance, it seems likely that the uptake of FOBT will be higher among women than among men, although the differences may be smaller among older age groups.  Ethnicity Monitoring data clearly indicate lower participation among ethnic and Indigenous groups compared with other groups. However, the reviews indicate that when the analyses control for other factors (e.g. socioeconomic status), ethnicity may not be a significant factor in the uptake of screening.  Socioeconomic status When socioeconomic status is operationalised in terms of income, there appears to be sufficient evidence to conclude that lower income is associated with lower screening rates. There is some research to indicate that the underlying issue with low socioeconomic status may not be a lack of money as such but a focus on the more pressing problems of everyday life which works against the uptake of screening. The Australian monitoring data indicate lower uptake in the areas ranked lowest on a socioeconomic area index. The relationship between uptake and area of residence socioeconomic status for other rankings varies across tests.  Education
On balance, the evidence indicates that low levels of education are associated with lower uptake of mammographic screening, Pap tests and FOBT. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 29 4 COGNITIVE FACTORS The term cognition (Latin: cognoscere, to know or to recognise) refers to the processing of information, applying of knowledge, and changing of preferences. Within psychology the concept is closely related to concepts such as the mind, reasoning, perception, intelligence, learning, and other concepts describing ‘capabilities of the mind’. 6 Cognition can therefore be regarded as a term referring to the mental processes involved in gaining knowledge and understanding. 7 Five factors have been included under the umbrella of cognitive factors: literacy, health literacy (HL), cancer related knowledge, perceived risk and the perceived ambiguity of messages relating to cancer screening.  Literacy In 1991, the US Congress defined literacy as ‘the ability to read, write and speak the language in order to compute and solve problems at levels of required to function on the job and in society, achieve one's goals and develop one's knowledge and potential’.32, 33  Health literacy (HL) According to Ishikawa et al. (2008) there are numerous HL definitions. For example: o the American Medical Association (AMA) defines HL as a constellation of skills required for functioning in the health environment o the National Library of Medicine (NLM) defines HL as the degree which a person has the capacity to obtain, process and understand basic health information and services needed to make appropriate health decisions o the World Health Organization (WHO) defines HL as the cognitive and social skills which determine the motivation and ability of individual to gain access to, understand and use information in ways that promote and maintain good health.33 Regardless of how it is defined, HL affects: (i) a person’s knowledge about health and health care; (ii) the ability to find and communicate health information; and (iii) the skills required for making health decisions.34  Cancer‐related knowledge In the literature, cancer‐related knowledge covers a range of issues related to screening: (i) knowledge of the risk factors; (ii) knowledge of the screening process itself; and (iii) knowledge of screening guidelines.  Perceived risk According to Katapodi et al. (2004), education interventions that aim to improve the uptake of mammograms have generally been based on theoretical models that attempt to explain how and why individuals adopt health‐protective behaviours. The majority of these models adopt a decision‐making perspective focused on a cost–benefit analysis of outcomes (e.g. Health Belief Model, Theory of Planned Behaviour). 8 In these models one 6
http://en.wikipedia.org/wiki/Cognition http://psychology.about.com/od/cindex/g/def_cognition.htm 8
An overview of these models is contained in Vol 3. 7
VOL 2– 30 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening of the major components of the cost–benefit analysis is a person’s perceived susceptibility to the disease. Perceived susceptibility or perceived risk refers to a person's belief about the likelihood that a health problem will be experienced if no precautions or behavioural changes are undertaken.35  Perceived ambiguity Han et al. (2006) have defined ambiguity as uncertainty about the ‘reliability, credibility, or adequacy’ of one's information concerning the likelihood of outcomes. Ambiguity arises from incomplete or conflicting evidence or expert opinion and is high when ‘expert knowledge’ is contested.36 4.1 Literacy The Federation of Ethnic Communities' Councils of Australia (2006) has pointed out that low literacy levels among some national groups in Australia have implications for the success of health promotion programs with new and emerging communities.37 The relationships between the host country's language and literacy is shown in two reports from the Organisation of Economic Co‐operation and Development (OECD). In one report, non‐native language status is a significant predictor of literacy in all English‐
speaking countries (including Australia, Canada, New Zealand, the United Kingdom and the United States) and exerts an effect in the smaller European countries with large immigrant populations such as Finland (Table 21). As shown in Table 21, there is a negative relationship between age and literacy; in both Australia and Finland, older age is associated with lower literacy levels.38 Table 21 Percentage of variance (R sq) in literacy proficiency accounted for by 12 predicator variables: Australia and Finland Variable Australia Variable Finland Native vs foreign language 0.299 Respondent's education 0.318 Respondent's education 0.294 Parents' education 0.159 Occupational category 0.164 Occupational category 0.138 Labour force participation 0.112 Native vs foreign language 0.113 Reading at home 0.093 Labour force participation 0.103 Participation in adult education 0.090 Participation in adult education 0.091 Participation in voluntary activities 0.083 Reading at work ‐0.042 Parents' education 0.052 Participation in voluntary activities 0.038 Industrial sector 0.033 Reading at home 0.019 Reading at work 0.033 Industrial sector 0.019 Gender 0.017 Gender 0.007 Age –0.133 Age –0.175 Explained variance 0.488 Explained variance 0.454 38
Source: Organisation for Economic Co‐operation and Development (2000) Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 31 The second OECD report related to the International Adult Literacy Survey (IALS), which measured literacy proficiency in the domains of prose literacy, document literacy and quantitative literacy.38 Each domain is divided into five levels.  Level 1 Persons at this literacy level have very poor literacy skills. For example, a person at this level may be unable to determine the correct amount of medicine to give a child from information printed on the package.  Level 2 People at this level can deal only with material that is simple, clearly laid out, and in which the tasks involved are not too complex. It identifies people who can read and they may have developed coping skills to manage everyday literacy demands, but their low level of proficiency makes it difficult for them to face novel demands.  Level 3 This is considered a suitable minimum for coping with the demands of everyday life and work in a complex, advanced society. It corresponds roughly to the skill level required for successful secondary school completion and college entry. It requires the ability to integrate several sources of information and solve more complex problems.  Levels 4 and 5 People at these levels demonstrate command of higher‐order information processing skills.39 Document literacy is described as ‘the knowledge and skills required to locate and use information contained in various formats’.38 It has been found that, compared to people born in Australia (native‐born), those born in a foreign country and whose ‘mother tongue’ is not English (second‐language, foreign‐born) are four times more likely to be in the lowest document literacy level (Figure 4). A review of studies conducted in the ‘developed world’ which was published in 2004, contained two studies which evaluated the relationship between reading ability and health promotion.  Scott et al. (2002) found that after controlling for age, gender, race, education and income, Medicare enrolees in the USA with lower literacy had greater odds of never having had a Pap test (OR 1.7, 95% CI 1.0–3.1) and not having had a mammogram in the past 2 years (OR 1.5, 95% CI 1.0–2.2).40, 41  In a study of cervical cancer screening in India, researchers interviewed women residing in rural areas and urban indigent areas who had been invited to a cervical screening program. They found that there were no statistically significant differences in the age distribution, marital status and socioeconomic status of the women who did and did not attend screening. However, there was a lower literacy rate among the non‐attenders (49.0%) compared with attenders (59.5%), and the difference was statistically significant (OR 2.25; 95% CI 1.23–41.3).42 VOL 2– 32 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Figure 4 Document literacy: percentage of Australian‐born (native‐born) and second language, foreign‐born Australians aged 16–65 years in each literacy level 100%
Levels 4&5: 6%
Levels 4&5:
19%
Level 3: 25%
75%
Level 3: 40%
Level 2: 22%
50%
25%
Level 2: 29%
Level 1: 48%
Level 1: 12%
0%
NATIVE BORN
SECOND-LANGUAGE
FOREIGN-BORN
Source: Organisation for Economic Co‐operation and Development (2000)38 4.2 Health literacy HL has been described as ‘literacy in the context of health and health care’. According to Ishikawa et al. (2008), some studies have reported significant associations between measure of literacy and measures of functional HL. However, it has also been noted that even individuals with adequate literacy might not have adequate HL.33 According to Ackerson et al. (2007), people with poor HL skills have difficulty understanding information containing unfamiliar vocabulary and concepts. They point out that low HL rates are more common in certain populations, such as the elderly, ethnic minorities, immigrants, non‐active English speakers, and those with limited education and low income. Low HL has consistently been associated with inadequate knowledge about disease and reduced health screening behaviours.24  Pap tests According to Giordano et al. (2008), HL is recognised as a critical factor affecting the individual’s ability to understand the importance of health care and the risks and benefits of cervical cancer screening.43 One study published in 1991 found that HL (in English) was the only factor independently associated with cervical cancer screening knowledge. A second study published in 2004 found that low HL among Spanish speakers in the USA also correlated with low cervical cancer screening rates.24  Faecal occult blood testing According to Friedman et al. (2008), the majority of cancer studies relating to HL involve people with breast cancer, even though information for other cancers (e.g. colorectal cancer) is more difficult for people to understand. For example, interviews with older Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 33 adults on their comprehension of breast, prostate and colorectal cancer information found that individuals had significantly greater difficulty comprehending resources about colorectal cancer compared to other cancers, especially regarding screening procedures such as flexible sigmoidoscopy, colonoscopy and FOBT. Friedman et al. (2008) state that both printed and web‐based colorectal cancer material materials are written at difficult readability levels.34 Donelle et al. (2008) undertook a study to describe the impact of HL and level of education on participants’ understanding of ‘common’ and ‘uncommon’ internet‐based colorectal cancer prevention information. As shown in Table 22, education was not a statistically significant predictor, but HL and particularly health numeracy were significant.44 Table 22 Multiple regression modelling of comprehension of internet‐based colorectal cancer information Explanatory variable Common internet information Uncommon internet information Combined information Beta coefficient Beta coefficient Beta coefficient Gender 0.118 0.30 0.00 Age –0.26* –0.32 –0.678* Education 0.15 0.21 0.363 STOFHLA prose 0.014 0.007** 0.067** STOFHLA numeracy 0.087** 0.043 0.10 Health numeracy 0.401** 0.38** 0.838** General numeracy –0.13 0.58** 0.466 Math anxiety –0.004 Model Statistics R sq –0.006 0.377** –0.01 0.562** 0.598** 44
Source: Donnelle et al. (2008) Notes: STOFHLA: Short Test of Functional Health Literacy for Adults * & ** represent statistically significant coefficients Univariate analyses in a study undertaken by Guerra et al. and published in 2005 indicated that HL was associated with awareness of colorectal cancer and FOBT but not with knowledge of the screening guidelines for FOBT. However, in the multivariate analyses there was no statistically significant relationship between awareness and HL (Table 23). VOL 2– 34 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Table 23 Association between functional health literacy and cancer‐related knowledge Inadequate/marginal HL Adequate HL Number Percent Number Percent (a) Yes: Heard of colorectal cancer 34 61% 75 95% 0.00 0.87 (b) Yes: Heard of FOBT 28 50% 66 84% 0.00 0.10 46 72% 0.24 p value Un‐
Adjusted (1)
adjusted If Yes to (b) asked How often should a person be tested? Correct response 16 59% 45
Source: Guerra et al. (2005) Notes: (1) Adjusted for ethnicity, Medicaid status, insurance status, education and income HL: Health literacy; FOBT: Faecal occult blood testing Reading and information‐seeking According to von Wagner et al. (2009), the evidence indicates that adults with limited HL are less likely to seek out or engage with printed health communication material and that limited HL also causes difficulties once information has been accessed. This has implications for cancer awareness and knowledge. For example, limited HL has been associated with lower knowledge about, and more negative attitudes towards, colorectal cancer screening. Participants with limited HL have expressed more negative attitudes towards colorectal cancer screening using FOBT, including worries that FOBT may be messy and inconvenient. Men with lower HL skills were found to be four times more likely to refuse the offer for FOBT screening, even if it was recommended by their doctor.46 In the review published in 2002 of the factors associated with colorectal cancer screening, Petersen concluded that lower adherence with colorectal cancer screening guidelines was generally associated with lower tendency to be an information seeker.5 However, it has also been found that ‘many’ patients, regardless of their HL levels, have reported difficulties with health information leaflets. For example, in a study undertaken by Shaw et al. (2009), 8% of the participants with adequate HL indicated they had some difficulty understanding the medical leaflets used in the study (Table 24).46, 47 Table 24 Relationship between health literacy and understanding of medical leaflets (n = 321) Adequate level of HL(1) Number Difficulty No difficulty Total Percent Low level of HL Number Percent 19 8% 26 37% 232 92% 44 63% 251 100% 70 100% 47
Source: Shaw et al. (2009) Notes: HL = Health literacy Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 35 Self‐efficacy Wolf et al. (2007) have pointed out that, although the association between HL, health‐related knowledge and attitudes are significant, studies have found that these associations only partially account for the link between HL and health outcomes. These authors believe that it is important to establish connections between HL and variables that are ‘proximal’ to behaviour such as self‐efficacy.48 Self‐efficacy is an individual’s perception of his/her ability to perform a specified behaviour/set of behaviours.49 It has been used extensively to predict the likelihood of enacting behaviours in various contexts including colorectal cancer screening.48 In a workplace study of employees aged 40 years and older, self‐efficacy was a significant and independent predictor of colonoscopy attendance.50 In their review published in 1997, Vernon et al. found that in self‐efficacy was positively associated with adherence to FOBT in 3 of 3 studies.2 However in a 2007 study, there were no differences in reported self‐efficacy to complete FOBT between participants with adequate HL and those with limited HL (3.93 vs 3.87, p = 0.44, adjusted for age, sex, race, insurance status).51 4.3 Cancer‐related knowledge In their systematic review published in 2000, Jepson et al. concluded that there was little evidence of a statistically significant positive relationship between mammogram uptake and breast cancer‐related knowledge. They did not include any studies relating to the relationship between knowledge and Pap tests or FOBT (Table 25). Table 25 Relationship between cancer related knowledge and screening uptake: Jepson et al. (2000) review Screening mammograms Studies with statistically significant effects (p ≤ 0.05) Breast cancer: 0% (0/4) Mammography and screening guidelines: 20% (1/5) Authors' conclusions The majority of studies did not find a significant association between the uptake of mammograms and knowledge about breast cancer or screening mammography Source: Jepson et al. (2000)6  Screening mammography In 2006, Parsa et al. published a review that included several studies examining the factors affecting screening practices among women of Asian descent. According to the authors, these studies demonstrated that lower screening rates are associated with lower levels of cancer‐related knowledge. Studies in Korea and Turkey showed knowledge of breast cancer screening guidelines was a major predictor of regular screening women; women who had knowledge of mammography guidelines were ten times more likely to have regular mammograms.52 A study in Hong Kong that was included in the Parsa et al. (2006) review reported that 42% of the surveyed population had heard of screening mammography. In this study, education level had no impact on the awareness. Full‐time housewives were significantly more likely to have heard of screening mammography compared to non‐housewives (49% VOL 2– 36 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening vs 73%, p = 0.0001).52 In the Wells et al. (2007) review of factors associated with the uptake of screening among Latin American women in the USA, 2 studies evaluated knowledge of breast cancer and attitudes towards breast cancer and participation in breast cancer screening. One found that a higher level of knowledge about breast cancer was related to higher odds of ever having a mammogram (adjusted OR 2.8) but another found that knowledge about risk factors for breast cancer and symptoms of breast cancer was not related to screening mammography.10 As shown in Table 26, the Schueler et al. (2008) meta‐analysis of factors associated with screening mammography uptake indicated a statistically significant association between poor screening knowledge and lower levels of uptake. Moreover, beliefs that mammograms are harmful and are only needed if symptoms are present were associated with reduced uptake.26 Table 26 Relationship between breast cancer knowledge on screening mammography uptake: Schueler et al. (2008) review Number of studies Adj OR 95% CI screening 11 0.46 0.35–0.60 risk factors 6 0.76 Not available Mammogram ineffective/ inaccurate 6 0.48 Not available Mammogram harmful 6 0.54 0.43–0.67 Mammogram only needed if symptoms present 5 0.56 0.43–0.72 Poor knowledge of: Source: Schueler et al. (2008)26 Notes: Adj OR: Adjusted odds ratio; CI: Confidence interval  Pap tests According to Ackers et al. (2007) lack of knowledge about the importance of cervical cancer screening, early detection, and treatment has been extensively studied, particularly among ethnic minorities and women from low socioeconomic groups. These studies have shown that women who are younger, non‐white, and those with low educational attainment or low income are more likely to be unaware of the purpose of Pap testing.24 According to Ackerson et al. (2007), few studies have been published that describe the beliefs and perceptions of African‐American women regarding cervical cancer and the impact these have upon screening uptake.27  Faecal occult blood testing In 1995, a population study was conducted of Australians' colorectal screening attitudes, intentions, beliefs and knowledge. Based on a sample of 1776 people, the study found that although community awareness of FOBT screening was ‘reasonable’ there was a lack of uptake of FOBT. The potential barriers to screening on a mass scale identified included a lack of knowledge about the causes and disease trajectory of colorectal cancer. The study noted that participants born in Australia, the UK and Ireland were more likely to be Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 37 aware of colorectal cancer screening tests.21 In the 1997 Vernon review, 3 of 4 studies (75%) reported a positive association between knowledge of cancer risk factors and the uptake of FOBT. The other study reported no association.2 In 2002, Petersen reported that having ‘less knowledge of colorectal cancer and cancer screening’ was generally associated with lower adherence to colorectal cancer screening guidelines.5 According to Wolff et al. (2003), African‐American underserved populations are not as likely to be informed about the current guidelines recommended for cancer and screening.30 One study found that the American Cancer Society recommendations for early detection in asymptomatic people were not well known by a study sample of African‐Americans. Although 75% of the men participating in the study knew that annual physical examinations were recommended for cancer detection in men over 40 years of age, only 25% of the men identified colorectal examinations as part of this annual examination.30 In 2004, the baseline knowledge, attitudes and practices survey conducted at the commencement of the Australian Pilot Program for colorectal cancer screening showed that awareness among Australians of FOBT screening was ‘relatively low’. Only 41% of those surveyed had heard of FOBT and only 26% had ever been tested. In the follow‐up survey undertaken near the completion of the pilot, 81% of those surveyed had heard of FOBT with 43% having undertaken a test in the last 12 months. However, in these surveys, people from CALD backgrounds were under‐represented; people were excluded from participating if their English language skills were not adequate for them to participate in the survey.21 In the 2010 review by Guessous et al., of the factors associated with colorectal cancer screening among people aged 65 years of age and over, a lack of awareness of colorectal cancer screening was reported as a significant barrier in 100% of studies (11 of 11).3 4.4 Perceived risk In the Jepson et al. review published in 2000, the majority of studies failed to find an association between perceived risk or susceptibility and the uptake of mammograms and FOBT (Table 27). There were no studies in this review that examined the relationship between perceived risk and the uptake of Pap tests. Table 27 Relationship between perceptions of risk and screening uptake: Jepson et al. (2000) review Screening mammograms FOBT Studies with statistically significant effects (p ≤ 0.05) 1/8 (12.5%) 1/3 (33%) Authors’ conclusions The majority of studies did not find an association between the uptake of mammograms and perceived vulnerability or susceptibility (risk) The majority of studies did not find an association between the uptake of FOBT and perceived vulnerability or susceptibility (risk) Source: Jepson et al. (2000)6 Notes: FOBT: Faecal occult blood testing VOL 2– 38 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening  Screening mammography In an analysis of the 1990 US National Health Interview Survey data, Austin et al. (2002) found that the response ‘not needed’ was the largest barrier to screening mammography among older Hispanic women. According to Austin et al. (2002), it appears that many women understand that mammography successfully detects breast cancer early, but they do not perceive their own vulnerability to breast cancer and do not see themselves at risk if they asymptomatic or have no family history of the disease.53 In 1996, McCaul et al. undertook a meta‐analysis of the relationship between perceived breast cancer risk and the uptake of mammograms. Their results showed that perceived risk was positively associated with screening mammography in 95% of studies (18 of 19). Most, but not all, of these studies were of women at average risk of breast cancer. The average effect size (adjusted for sample size) was r = 0.16, and was smaller for prospective (r = 0.10) compared with cross‐sectional studies (r = 0.19). The authors concluded that there was no support for the hypothesis of a curvilinear relationship between perceived risk and screening (i.e. that high and low perceived risks are negatively associated with screening).4 In 1999, Vernon et al. concluded that studies had consistently found a positive association between perceived risk and screening mammography in women at average risk of breast cancer although the magnitude of the overall effect size was small.54 Katapodi et al. (2004) built on the McCaul et al. (1996) meta‐analysis and included 13 extra studies published between 1993 and 2002. Four of the extra 13 studies (31%) did not demonstrate a positive association between perceived risk and screening mammography. The average effect size for the extra studies weighted by sample size was 0.20 (95% CI 0.18–0.23). When the 13 studies were added to the McCaul et al. (1996) data, the average effect size was 0.19. The authors concluded that the results suggest that perceived risk has a small but significant positive effect on adherence to screening mammography.35 Overall, 42% of the 42 studies included in Katapodi et al. (2004) meta‐analysis included women of diverse racial/cultural backgrounds in percentages ranging from 14% to 100%. The remaining 58% of the studies reviewed included mostly, or exclusively, white women. Five studies examined the relationship between race/culture and perceived risk of breast cancer in samples consisting of 14–49% minority women. In these studies, white women were more likely to perceive themselves as being at increased risk for developing breast cancer compared with other women. Black women were more likely to be unaware that diagnosis of a first‐degree relative with breast cancer increased their risk of developing the disease (ES = 0.38, 95% CI 0.28–0.47). However, two studies with an over‐
representation (> 60%) of women from diverse racial/cultural backgrounds reported no significant differences perceived risk of breast cancer among women of diverse ethnic/cultural groups and white women but there were ‘insufficient data to calculate effect sizes’.35 In 2007, Magai et al. concluded that it appears that perceptions of risk do have an impact on screening adherence but effect sizes appear to be ‘comparatively’ small.55. In a meta‐
analysis published in 2008, Schueler et al. concluded that the belief one was less Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 39 personally susceptible did not have a statistically significant impact on the uptake of mammograms (6 studies, Adj OR=0.75, 95% CI 0.45–1.24).26  Pap tests The Vernon et al. (1999) review contained 3 studies with a multivariate analysis of a number of cognitive and attitudinal variables, including perceived risk, and cervical cancer screening. After controlling for other variables, 1 study found a positive association with cervical cancer screening and 2 studies found no association. 54 According to Ackerson et al. (2007) minority women were less likely to participate in Pap tests if they did not perceive cervical cancer as a possibility. Hispanic women believed that physical trauma related to abortion and rough sex, an infected partner, and lack of feminine hygiene made an individual vulnerable to cervical cancer. If Hispanic women did not feel that they were personally vulnerable, they were less likely to obtain testing.27  Faecal occult blood testing In a review of the literature on colorectal cancer screening adherence, Vernon et al. found that 2 of 8 studies (25%) reported a positive association between perceived risk and completion of FOBT while 6 studies (75%) reported no association. The authors concluded that, at that time, there were insufficient data to draw firm conclusions about the pattern or magnitude of the association between perceived risk and uptake of FOBT.2 In 2002 Petersen concluded that lower perceived risk of colorectal cancer and lower perceived susceptibility to colorectal cancer were associated with lower screening adherence.5 In a 2009 study, Weinberg et al. found that, in a population of ‘average‐risk, non‐
compliant women’ for whom the barriers regarding access and insurance were almost eliminated, inaccurate risk perception about colorectal cancer and the belief that screening might be painful were among the key factors that contributed to reduced intention to participate in colorectal cancer screening.56 4.5 Perceived ambiguity Perceived ambiguity appears to be a relatively ‘new’ line of research in relation to screening behaviour. It has been included because, as Han et al. (2007) state, there are ‘conflicting expert recommendations regarding cancer screening and prevention’ that are ‘growing in number, visibility and importance’.57 Mullins (2005) has also pointed out that ‘recommendations on the frequency of cervical cancer screening vary from country to country’.58 Moreover, a country’s recommendations may change over time and these changes are often highlighted in the media (Box 3). According to Han et al. (2006) one of the consequences of publicity about health care (including cancer screening) is a belief among the scientific community that it leads to an increase in ‘public confusion and scepticism about health recommendations’. Added to this publicity, the promotion of ‘informed consent’ and ‘shared decision making’ as normative ideals in health care has meant that doctors have an obligation to increase patients’ awareness of scientific uncertainties so that they can understand the implications of their decisions in relation to health care.36 VOL 2– 40 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Box 3 USA: Changes in screening mammography and Pap testing guidelines New cervical cancer screening guidelines released By Saundra Young, CNN Medical Producer November 20, 2009 10:51 a.m. EST The new mammogram recommendations out earlier this week caused quite an uproar. Now comes another change in screening tests for women ‐‐ this one for cervical cancer. The American College of Obstetricians and Gynecologists (ACOG) releases new guidelines Friday, saying women don’t need their first cervical cancer screening ‐‐ or Pap test ‐‐ until they’re 21 years old. And, they don’t need follow‐up examinations as often as previously recommended. According to the guidelines, women younger than 30 should be screened every two years, instead of annually. Women 30 or older can be examined once every three years. "The tradition of doing a Pap test every year has not been supported by recent scientific evidence," said Dr. Alan G. Waxman, who developed the document for ACOG’s Committee on Practice Bulletins‐Gynecology. "A review of the evidence to date shows that screening at less frequent intervals prevents cervical cancer just as well, has decreased costs, and avoids unnecessary interventions that could be harmful." The tradition of doing a Pap test every year has not been supported by recent scientific evidence. The current guidelines, from 2003, recommend that women get a Pap test three years after they begin having sexual intercourse, but no later than age 21. And that women younger than 30 have an annual exam. For women 30 or older, the recommendation was every two to three years, if they’d had three consecutive negative Pap tests. The American Cancer Society (ACS) supports the guidelines and said it is reviewing new data and updating its own recommendations. "There’s good data since the last guidelines in 2003 that show that screening teens or before age 21 is not having an impact on reducing cervical cancer," said Debbie Saslow of the Cancer Society. And, Saslow added, this is completely different from the new, hotly debated mammogram recommendations. "Getting an annual Pap test is the equivalent to getting a mammogram every four months. Breast cancer on average is growing at a point where, if you get a mammogram every two years, you will miss a lot of deadly cancers that you would have caught if you’re having them every year. This is not true for cervical cancer; we are detecting pre‐cancers that are taking 10 to 20 years to develop into cancer." According to the ACS, there are about 10,000 new cases of cervical cancer each year, and more than 4,000 deaths. Over half were found in women who never had a Pap test. Most cases are in women younger than 50, and rarely occur in females younger than 20. The risk simply is not there, though the human papillomavirus (HPV), which is responsible for 70 percent or more cervical cancers, is high among sexually active teens, said Dr. David Soper of the Medical University of South Carolina. The vast majority of those infections will resolve and not cause any significant pre‐cancerous lesions, according to Soper. Females, particularly adolescents, develop immunity to HPV and can resolve the infection without treatment. Source: Downloaded on 14th January 2009 from http://www.cnn.com/2009/HEALTH/11/20/cervical.cancer.guidelines/index.html Han et al. (2006) reported that studies have provided ‘strong’ support for the proposition that cancer worry will ‘predispose’ people to interpret ambiguity relating to cancer prevention recommendations more pessimistically; it is likely to reduce people’s beliefs that cancer can be prevented. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 41 In a 2006 study, Han et al. found increased perceived ambiguity was associated with an increase in:  perceived cancer risk (Adj OR 1.46, 95% CI 1.18–1.80, p < 0.01)  cancer‐related worry (Adj OR 1.53, 95% CI 1.01–2.32, p =0.04).36 In a study reported in 2007, Han et al. conducted a secondary analysis of longitudinal data from the 1995 Maximizing Mammography Participation Trial in the USA.57 The results indicated that, compared with women reporting the lowest level of perceived ambiguity, women with higher levels of perceived ambiguity had progressively:  lower odds of intending to obtain a future mammogram; for women indicating the highest level of perceived ambiguity OR = 0.34 (95% CI 0.20–0.55);  lower odds of actually obtaining a subsequent mammogram; for women indicating the highest level of perceived ambiguity (OR 0.43, 95% CI 0.22–0.84). Contrary to predictions arising from the earlier research, no significant relationship was found between baseline perceived ambiguity and perceived breast cancer risk at the 12‐month follow‐
up, controlling for demographic variables and baseline perceived risk. The authors concluded that, among women aged 50–79 years, high perceived ambiguity about mammography recommendations is associated with both diminished uptake of screening mammography over time, and lower intentions for future mammography. High perceived ambiguity also predicted greater mammography‐related worry. Summary  Literacy There is evidence to indicate lower literacy levels among Australians for whom English is not their ‘native’ language. However, there is very little evidence relating to the association between literacy and screening uptake. Two studies indicated that lower literacy was associated with lower uptake of mammograms among Medicare enrollees in the USA and Pap tests among women residing in rural and urban indigent areas in India. However, it is not clear that these results would generalise to other groups.  Health literacy HL is, to some extent, associated with literacy but not entirely. HL has been found to be lower among the elderly, ethnic minorities, immigrants, non‐active English speakers and those with limited education and low incomes. Low HL has been associated with lower uptake of Pap tests when other factors are controlled. There is some evidence that low HL is associated with proximal measures of participation in colorectal cancer screening (e.g. knowledge, comprehension, information seeking, and difficulty in understanding information) and negative attitudes towards colorectal cancer screening and refusal to participate in FOBT screening.  Cancer‐related knowledge The evidence for the relationship between breast cancer knowledge and uptake of screening mammograms is conflicted. However, the latest review published in 2008 indicates that poor knowledge of screening, the belief that mammograms are harmful and that mammograms are not needed unless symptoms are present are associated with a VOL 2– 42 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening reduction in the uptake of mammograms. There is insufficient evidence to draw any conclusions about the relationship between cancer‐related knowledge and Pap tests. The evidence for FOBT indicates that knowledge may be lower in ethnic groups, and there is some evidence that poorer knowledge of colorectal cancer is associated with a lower uptake in colorectal cancer screening.  Perceived risk On balance, the evidence indicates that perceiving oneself to be at a heightened risk of breast cancer is likely to increase the uptake of mammograms; however, the effect is small. There is little evidence to support this association in relation of Pap tests and FOBT.  Perceived ambiguity There was some evidence that higher levels of perceived ambiguity are related to lower participation in the uptake of mammograms. There was no evidence for the impact of this factor on the uptake of Pap tests and FOBT. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 43 5 PSYCHOSOCIAL FACTORS In 2003, Magai et al. wrote that the literature on screening adherence had moved away from emphasising demographic and structural variables as predictors to include cognitive and emotion variables. At that time, the authors indicated that studies suggested that emotion variables could add ‘significant’ variance to the prediction of screening in multiple regression and regression models. They used the example of the Bowen et al. (2003) study, in which emotion variables (cancer worry, anxiety, and depression) added an additional 4% of the variance to screening mammography.55, 59 This section of the VCS review examines the relationship between screening and the three following psychosocial factors.  Fear/anxiety/worry Definitions of cancer worry and anxiety are conceptually and empirically distinct from perceived cancer risk. Perceived risk and cancer worry are only moderately correlated (0.30–0.40) and both factors have been found to be independently related to cancer screening.60 A clinical diagnosis of anxiety disorder also differs from the measurement of anxiety symptoms.61 Spielberger (1996) differentiated between state and trait anxiety. Trait anxiety was defined as a person’s predisposition to experience anxiety in a stressful situation and tends to be stable over time; its test–retest reliability is high (range 0.73–
0.86). State anxiety was defined as a short‐lived emotional response characterised by unpleasant feelings of tension and apprehensive thoughts 61‐63 In the cancer prevention literature, there does not appear to be a distinction between these two types of anxiety. However, the small amount of literature relating to trait anxiety has been included in the section of the review that focuses on health and lifestyle factors rather than in this section.  Emotion regulation (coping styles) Emotion, such as fear, anxiety and worry, have well‐documented, strong motivational properties, and people often feel impelled to regulate them. It has been hypothesised that the manner in which the individual regulates, or fails to regulate, these emotions is likely to exert an important influence on health behaviour.1, 55 According to Magai et al. (2007), people engage in a range of regulatory strategies, including active problem solving and seeking social support, as well as by activating more defensive coping styles, such as repression and denial.55  Social networks According to Kinney et al. (2005), research and theory suggest that social networks play a role in a range of health‐related outcomes. In the literature, social networks and social support have been operationalised in two ways: o the structural approach focuses upon the aspects of social networks such as social integration, group membership, marital status and the number of social contacts o the functional approach focuses on the nature of the support provided, such as emotional support, the provision of tangible support (e.g. the provision of material and VOL 2– 44 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening financial support) or the provision of information (e.g. contact with someone who has experienced something similar). 64 5.1 Fear/anxiety/worry In 2005, Hay et al. wrote that ‘the question of whether cancer worry … facilitates or deters cancer screening has been pursued for over 30 years. Yet … we currently lack consensus concerning whether cancer worry motivates of inhibits screening’.60 This review was accepted for publication in 2004, the same year in which a review by Consedine et al. was published.1 Consedine et al. (2004) also noted that the relationship between screening behaviour and ‘the construct variously called anxiety, fear, or worry’ had been studied extensively, but it was unclear whether fear was positively or negatively associated with screening behaviour. The authors noted that part of the difficulty was due to the fact that ‘fear’ had been operationalised in a variety of ways: fear of cancer, fear of the screening process, or fear of the outcome. Consedine et al. (2004) proposed a model in which:  fear of the screening components is likely to result in a decrease in screening behaviour,  an undifferentiated fear of cancer is likely to lead to an increase in screening behaviour, and  the result for a fear of the screening outcome was not clear (Figure 5).1 Magai et al. (2007) in their review of the psychosocial and cognitive variables associated with breast cancer screening indicated that the few papers that had been published on this subject since the Consedine et al. (2004) review were ‘not inconsistent’ with the model.55 Figure 5 Model of the relationship between fear and breast cancer screening behaviour: Consedine et al. (2004) review
Emotion A Fear of screening components  B Fear of screening outcome  C Undifferentiated fear of cancer    Mixed effects, may depend on relations with A and C   Impact on Screening Behaviour Decreased likelihood of screening (avoidance) Increased likelihood of screening (endorsement) Source: Consedine et al. (2004)1 Hay et al. (2006), in their review of the role of worry in breast cervical and colorectal cancer screening, pointed out that the conflicting empirical findings prior to the production of the Consedine et al. (2004) model could have been due to factors including:  the ‘multiple, un‐reconciled methods of defining and measuring cancer worry’ Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 45  the preponderance of cross‐sectional studies that measured cancer fear/worry/anxiety and screening adherence simultaneously. Hay et al. (2006) limited their review to studies in which worry assessment preceded measures of screening uptake. They found that 92% of the prospective studies (11 of 12) reported a positive relationship between cancer worry and screening behaviour. The un‐weighted effect size for screening mammography was 0.10 (p < 0.001). The authors also concluded that the data suggested that high levels of cancer worry were not common and that higher worry levels were not associated with reduced screening.65 5.2 Coping style (emotion regulation) In 2007, Magai et al. pointed out that there was ‘slim but important literature’ on how coping styles relate to breast cancer screening.55 The two styles that appear to have been researched are ‘denial’ and ‘repression’. 9 These are psychological mechanisms that people use to protect themselves from anxiety and are generally regarded as maladaptive from a psychoanalytical perspective. However, more recent and everyday understandings of the concepts indicate that whether or not they are maladaptive will depend on the costs and benefits associated with their use.66 Based on one study, Consedine et al. (2004) stated that research has shown that beliefs indicative of realism and willingness to directly face breast problems – attributes that appear antithetical to denial – were more prominent in women who screened compared with those who did not. They also reported that 2 studies showed that the emotion regulatory style of repression may be associated with increases in screening.1 One study compared 210 women who self‐referred for screening with 210 non‐attenders and found that those who self‐referred evinced higher repressiveness than non‐attenders.67 The other study, involving 1364 women in six ethnic groups, reported that greater repression was associated with increased uptake of screening mammography even when demographics and background characteristics were controlled.68 Magai et al. (2007) have also pointed out the different effects of denial and repression. In an attempt to reconcile these contradictory results, the authors speculated that: (i) repression operates more broadly, such that cancer threat is pre‐emptively excluded from consciousness; whereas (ii) denial comes into play after threat has entered consciousness and generates avoidance behaviours aimed at reducing anxiety. 55 5.3 Social networks Structural approach Social integration There was little evidence in the reviews of the impact of social integration on the uptake of mammograms, Pap tests and FOBT. However, the database used for the review contained 2 9
Denial occurs when a person is faced with a fact that is too uncomfortable to accept and rejects
it instead, insisting that it is not true despite what may be overwhelming evidence to the contrary.
http://en.wikipedia.org/wiki/Denial Repression occurs when a person excludes thoughts which may
cause fear and anxiety from consciousness. http://en.wikipedia.org/wiki/Psychological_repression VOL 2– 46 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening primary studies showing the impact of the size of the participants’ networks on screening uptake; these studies have been included. 
Screening mammography Tejeda et al. (2009) used the Social Network Index (SNI) to determine the size of the social network of the participants. The results of the study found few differences between the size of the network and the uptake of mammograms, although a greater proportion of women who reported a having a recent mammogram were categorised as having large or medium social networks (Figure 6).69 However, the small number of participants in each group precluded the use of tests of statistical significance (e.g. Chi2). Figure 6 Association between social network size and screening mammogram status 100%
80%
Large Network , 7
60%
40%
Medium Network, 5
20%
Small Network, 2
0%
Mammogram Within
Previous 2 yrs
(N=14)
69
Source: Tejeda et al. (2009) Large Network , 2
Large Network , 4
Medium Network, 6
Medium Network, 4
Small Network, 5
Mammogram Over
2 yrs Ago
(N=13)
Small Network, 4
Never Had
Mammogram
(N=12)
 Faecal occult blood testing The Jepson et al. (2000) review included 4 studies relating to social influences on the uptake of FOBT. One used a version of the SNI to measure network characteristics. This study did not find any significant relationship between network characteristics and the uptake of FOBT.6 A primary study undertaken by Kinney et al. (2005) suggests that the impact of social integration may differ across ethnic groups. The results of this study indicated that participants who were ‘most socially connected’ (measured using the SNI) were more likely to report recent use of colorectal cancer screening tests 10 (OR 3.2, 95% CI 1.7–6.2) and that this association was stronger among blacks (OR 3.8, 95% CI 1.3–10.7) than whites (OR 2.9, 95% CI 1.2–6.9; p value for the interaction –0.006).64 10
Recent use was indicated by having at least one of: (i) FOBT within previous 12 months, (b)
sigmoidoscopy within the past 5 years, colonoscopy within the past 10 years, or (iii) double contrast barium
enema within the previous 5 years.
Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 47 Source of social support  Screening mammography Jepson et al. (2000) reviewed studies relating to the sources of social influence that might have affected women’s decisions to have a screening mammogram. They included health care professionals (2 studies), other household members (1 study), significant others (2 studies), family and friends (3 studies) and general social support (1 study). None of the studies found social support from any of these sources to be significantly associated with the uptake of mammograms. However, Jepson et al. (2000) reported that the studies involved primarily Caucasian populations and did not investigate women from ethnic minorities or low socioeconomic areas.6 Wu et al. (2005) reviewed the factors associated with breast screening practices among Chinese, Korean, Filipino, and Asian Indian American women and concluded that women whose friends and relatives favoured mammogram screening were more likely to be screened.70 Similarly in a study of the factors associated with the uptake of mammograms among Hispanic and non‐Hispanic white women living in a rural area, Tejeda et al. (2009) found that women who reported having a recent mammogram were more likely to mention female friends and daughters as people they trusted with their personal issues was positively associated with participation.69  Faecal occult blood testing The Jepson et al. (2000) review included 1 study that examined the impact of having two members of the same household taking part in the study. There was a significantly higher rate of FOBT uptake among study participants who lived with other participants.6 Marital status As shown in Table 28, Jepson et al. (2000) concluded that the majority of studies in their review did not find a significant relationship between marital status and the uptake of mammograms or Pap tests. The review also concluded that marital status was not found to be significantly associated with FOBT uptake, although the basis for this conclusion is not clear.6 Table 28 Relationship between marital status and screening uptake: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with 3/11 (27%) statistically significant effects (p ≤ 0.05) 2/5 studies (40%) Not included Authors’ conclusions The majority of studies did not find a significant association between the uptake of Pap tests and marital status. However, there was some evidence to suggest that being single was associated with non‐attendance Marital status was not found to be significantly associated with the uptake of FOBT The majority of studies did not find a significant association between marital status and the uptake of mammograms Source: Jepson et al. (2000)6 VOL 2– 48 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening  Screening mammography In a review published in 2007 of the factors associated with the uptake of mammograms among Latin American women in the USA, Wells et al. found that marital status was not consistently related to the uptake of mammograms in multiple logistic regression analyses. Two studies reported a relationship between being widowed (adj OR 0.66) or unmarried (adj OR 0.74) and the lack of recent screening mammography. Four studies reported no association between marital status and uptake of screening mammography based on multiple logistic regression analysis.10 In the Australian context, the Blue Moon Research and Planning (2008) review included 4 studies that indicated lower uptake of mammograms among single, widowed, or divorced women than among women in long‐term relationships.71  Faecal occult blood testing In their 1997 review, Vernon et al. reported that there were few studies of the associations between marital status and uptake of FOBT and that among those studies, results were inconsistent. Four studies reported that married persons were more likely to complete FOBT and 3 reported no relationship.2 Functional approach In the introduction to their study, Kinney et al. (2005) stated that most of the literature on the association between social networks and health has been concerned with structural analyses. However, researchers have hypothesised that it may be the type of support that influences health outcomes. Types of support have been classified as emotional (e.g. offering reassurance that one is loved and cared for), tangible (e.g. giving material or financial assistance), or informative (e.g. someone who has experienced something similar providing information). Vernon et al. (1997) included 1 study that examined the relationship between ‘social support’ and the uptake of FOBT. That study did not find a statistically significant association between the two factors.2 Similarly in their study, Kinney et al. (2005) found that neither emotional reassurance nor instrumental support was associated with colorectal cancer screening behaviour.64 Summary  Fear/anxiety/worry The work of Consedine et al. (2004) has been instrumental in understanding about the impact of fear/anxiety/worry on cancer screening behaviour. The model indicates that: (i) an undifferentiated fear of cancer is associated with an increase in screening uptake; and (ii) fear of the screening process itself is associated with a decrease in screening. However, the relationship between fear of the outcome of the screening process is not definitive. Other research has indicated that, in the community, high levels of cancer worry are uncommon, and that higher levels of cancer worry do not reduce screening uptake.  Coping style There is little research on the impact of coping styles on screening uptake. The two styles that have been investigated appear to have different effects – denial leading to a Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 49 reduction in screening activity and repression leading to an increase in the uptake of mammograms.  Social networks There appears to be little evidence about the impact of different types of social support (emotional, tangible, and informative) on screening uptake. Therefore, it is not possible to draw any firm conclusions. The structural aspects of social networks have been more thoroughly researched, but there is little firm evidence that social networks have an impact. There is some evidence that family and friends may have an influence on screening among ethnic groups. There is also some evidence to suggest that being single, widowed or divorced is associated with a decrease in the uptake of mammograms. VOL 2– 50 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 6 HEALTH AND LIFESTYLE Three sets of factors have been considered under the heading of health and lifestyle. These are: (i) physical and mental health; (ii) tobacco smoking and alcohol consumption (drug use); and (iii) previous participation in health screening. 6.1 Physical/mental health As shown in Table 29, the Jepson et al. (2000) review reported little evidence to support an association between health status and the uptake of screening mammography or Pap tests. However, there was some evidence to support the hypothesis of a positive association between the capacity to perform activities of daily living (ADL) and the uptake of FOBT.6 Table 29 Association between perceived health status and the uptake of screening: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with statistically significant effects (p ≤ 0.05) Perceived health status: 1/4 (25%) Chronic illness: 1/2 (50%) Capable of performing ADL: 2/3 (67%) Authors’ conclusions The majority of studies did not find an association between health status and the uptake of mammograms The relationship between health status and the uptake of Pap tests is unclear The majority of studies found that being able to perform ADL was associated with the uptake of FOBT. More able individuals were more likely to participate in screening Source: Jepson et al. (2000)6 Notes: FOBT: Faecal occult blood testing; ADL: Activities of daily living  Screening mammography In 2001 Sadler et al. wrote that women who are blind, deaf, have physical or mental disabilities and who are chronically ill may experience particular challenges in relation to receiving mammography services and that the ‘literature is particularly silent on the issue of screening mammography among these vulnerable populations’.20 This silence seems to have continued. Only the Schueler et al. (2008) review examined the relationship between physical health and uptake of screening mammography. As shown in Table 30, the only factor that was statistically significant was ‘no personal history of benign breast disease’, and this was associated with a decrease in the uptake of mammograms.26 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 51 Table 30 Impact of breast disease on screening mammography uptake: Schueler et al. (2008) review Number of studies Adj OR 95% CI Class 1 obesity 4 0.96 0.89–1.94 Disability/impairment in ADL 6 0.84 0.66–1.07 Self‐assessed poor health 15 0.95 0.78–1.15 13 0.51 0.42–0.62 (1)
No personal history of benign breast disease 26
Source: Schueler et al. (2008) Notes: (1) Class 1 obesity = body mass index of 30–34.9 kg/m2 Adj OR: Adjusted odds ratio; CI: Confidence interval; ADL: Activities of daily living In relation to mental health: o Consedine et al. (2004) reported that comparisons of screeners and non‐screeners have suggested that trait anxiety 11 is significantly more pronounced among women who do not attend a screening in response to an invitation than among those who do. They also state that ‘other research’ has indicated no relationship between screening behaviour and trait anxiety. The two prospective studies in the review are equally conflicted. One suggested improved screening at intermediate levels of worry and the other suggested a negative relation between worry and mammography.1 o According to Magai et al. (2007), most studies of depression and screening mammography have failed to find an association. The single exception was based on a sample of 364 African‐American women recruited from churches and low‐income housing projects; but it is unclear whether this association will generalise across other groups.55.  Pap tests There were no data in the literature reviews relating to the association between health status and the uptake of cervical screening. In the grey literature Johnson et al. (2002) provided some evidence that women with intellectual disabilities are less likely to be screened than women without disabilities (Box 4).72 11
Trait anxiety is defined as a person's predisposition to experience anxiety in a stressful situation and tends to be 62, 63
stable over time.
VOL 2– 52 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Box 4 Impact of intellectual disabilities on the uptake of Pap tests Research with women with intellectual disabilities in the UK and Australia has revealed that they are less likely than women without disabilities to have regular cervical screening. For example Stein and Allen (1999) audited cervical screening for women with an intellectual disability in one English health district. Out of 389 women with an intellectual disability identified … as eligible for cervical screening, only 13% had a record or a smear test in the previous five years73 … Similar research by Mencap(1) revealed that only 8% of women with intellectual disabilities underwent cervical screening compared with 85% of women who did not have disabilities.74 A state‐wide database is currently being developed by the Centre for Development Disabilities Health in Melbourne (2002 Sutherland, unpublished data). The preliminary database involved 243 women … results … indicated that 10% of women with an intellectual disability were involved in cervical screening during the previous 12 month period … Only 8% of the youngest age group (18‐35 years) had participated in cervical screening. 14% of women aged 36‐50 and 13% of women aged over 50 years indicated that they had participated. Sutherland (2002) commented that the older subset of women was much smaller and therefore results should be viewed with caution.75 Source: Johnson et al. (2002)72 Notes: (1) Mencap is a UK‐based organization which is committed to ‘valuing and supporting people with a learning disability and their families and carers’ (http://www.mencap.org.uk/landing.asp?id=6)  Faecal occult blood testing As shown in Table 31, Petersen (2002) found that more anxious persons, those with no concurrent or post illness and those with no prior tumour diagnosis were less likely to adherent with colorectal cancer screening guidelines.5 Table 31 Health factors associated with lower uptake of colorectal cancer screening tests: Petersen (2002) review Health factor Lower adherence generally seen with Anxiety More anxious persons Illness No concurrent or past illness Tumour No prior tumour diagnosis 5
Source: Petersen (2002) In a review of the factors associated with colorectal cancer screening among people aged 65 years of age and older published in 2010, Guessous et al. reported that only 3 of 13 studies (23%) reported the presence of chronic/co‐morbid conditions as a significant barrier to screening uptake. Seven of the 13 studies (54%) reported this factor as a significant facilitator of uptake.3 The Australian monitoring data for the NBCSP shows that males and females with severe or profound activity limitation were more likely to return FOBT kits than those who did not have a limitation (Table 32).12 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 53 Table 32 Australian National Bowel Cancer Screening Program (NBCSP): crude participation rate by disability status (2008) Severe or profound activity limitation Yes No Rate 95% CI Rate 95% CI Females 48.0 47.2‐48.9 37.2 37.0‐37.4 Males 38.4 37.6‐39.2 30.9 30.7‐31.0 Total 43.0 42.4‐43.6 34.0 33.9‐34.1 Source: Australian Institute of Health and Welfare and Australian Government Department of Health and Ageing (2009) 12 Notes: CI: Confidence interval 6.1 Drug use: alcohol and tobacco consumption The data about these lifestyle factors related to the uptake of mammograms. None of the reviews for Pap tests or colorectal cancer screening examined these factors.  Smoking The Jepson et al. (2000) review included 3 studies that investigated whether smoking was predictive of screening uptake. Only 1 study (33%) found a significant effect: women who currently smoked were less likely to adhere to screening than those who did not.6 Wells et al. (2007) reported that among Latin American women in the USA, not being a current smoker reduced the likelihood of participating in screening mammography (Adj OR 0.71).10 The 2008 review by Schueler et al. found that women who smoke cigarettes consistently demonstrated lower rates of mammography use (Adj OR 0.69, 95% CI 0.60–
0.80).26  Alcohol consumption The Jepson et al. review (2000) included 2 studies that examined the relationship between alcohol consumption and the uptake of mammograms. Only 1 study found this factor to be statistically significant: within an inner city population of women in the UK, women who consumed alcohol at least once a month were more likely to obtain a mammogram than those who did not.6 Schueler et al. (2008) concluded that drinking alcohol in any amount showed ‘consistent, modest effects’ on the uptake of mammograms (Adj OR 1.30, 95% CI 1.09–1.54).26 6.3 Previous participation in screening As shown in Table 33, the Jepson et al. (2000) review indicated that the majority of studies indicated previous participation in screening was associated in an increase in uptake of mammograms and FOBT but not Pap tests.6 VOL 2– 54 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Table 33 Association between previous screening behaviour and the uptake of screening: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with statistically significant effects (p ≤ 0.05) Previous mammograms: 13/20 (65%) Previous Pap tests: 1/4 (25%) Previous CRC screening: 4/5 (80%) Authors’ conclusions The majority of studies found that women who had previously attended for screening (compared with those who had not) were significantly more likely to attend for further mammograms The majority of studies found no significant association between the uptake of Pap tests and previous screening history Participation in previous colorectal screening was found to be significantly associated with the uptake of FOBT Source: Jepson et al. (2000)6 Notes: FOBT: Faecal occult blood testing; CRC: Colorectal cancer  Screening mammography A review published by Bankhead et al. (2003) included 22 papers comparing breast screening behaviours in women who had or had not had a mammogram previously: o 50% (11 studies) showed a positive relationship indicating that women who undergo screening mammography are significantly more likely to be screened in the future than women who have not been screened o 27% (6 studies) showed no differences in screening mammography uptake between previous attenders and non‐attenders o 23% (5 studies) showed an association between lower rates of attendance and having had a mammogram previously.76 In the 2000 review by Jepson et al., 5 studies examined whether previous screening behaviour (including attendance for Pap tests, clinical breast examinations and dental checks) was a significant factor in determining mammogram uptake. The authors concluded it was not clear whether attendance for other screening tests was associated with the uptake of mammograms.6 The Bankhead et al. review (2003) contained 26 studies investigating the effect of cervical screening on subsequent breast screening practices. Nineteen studies (73%) reported a positive association and 7 (27%) reported no association. Eight studies in the Bankhead et al. (2003) review examined the effect of the recency of the Pap test on screening mammography behaviour and all indicated that women with more recent Pap tests were more likely to attend for breast screening.76 In the Wells et al. review published in 2007, 2 studies examined the relationship between other health behaviours and uptake of screening mammography. The reported ORs Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 55 showed a significant association between screening mammography and having received a Pap test within the past 3 years (Adj OR 2.65).10  Pap tests Three studies in the Bankhead et al. (2003) review examined the association between attendance at previous screening and follow‐up screening. Two found no association with prior use of breast or cervical screening.76 Twelve studies investigated the effects of breast screening on cervical screening. Ten of the papers (83%) showed a significant positive association between the two screening behaviours: women who had participated in breast screening were more likely also to attend for cervical screening. The remaining 2 studies (17%) showed no association.76  Faecal occult blood testing In 1997, Vernon et al. concluded that health motivation or a preventive health orientation (e.g. engaging in other health‐promoting behaviours, such as regular medical or dental check‐ups) showed the ‘most consistent positive association with FOBT completion’. The authors found that this factor differentiated persistent compliers (persons who reported completing five consecutive FOBTs) from persistent refusers (persons who refused three consecutive FOBTs). Seven of 9 studies (78%) showed a positive association between health motivation/prevention orientation and FOBT uptake, while 2 studies (22%) showed no association.2 The review by Vernon et al. (1997) reported high rates of adherence (56% – 93%) when repeat screening was offered to those groups that had participated previously. The authors cited findings that low uptake rates (6%) characterised initial refusers who were subsequently reoffered FOBT screening.2 Petersen (2002) concluded that less participation in other screening behaviours (e.g. mammography) was associated with lower adherence to colorectal cancer screening.5 The 2010 review by Guessous et al. did not examine this factor.3 There is some evidence in the Australian context that previous participation in other cancer screening activities is associated with an increase in the uptake of FOBT. Women who indicated that they had a Pap test or a mammogram in the past 2 years and men who had participated in prostate cancer screening were more likely to return their FOBT kits than those who had not participated (Figure 7). However, these associations are based on univariate analyses and there is no way of determining if the associations would be statistically significant in a multivariate analysis which controlled for other factors associated with the uptake of screening. VOL 2– 56 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Figure 7 Association between previous cancer screening activities and the uptake of faecal occult blood testing Non-Returners, 45%
20%
Returners, 62%
40%
Non-Returners, 57%
60%
Returners, 67%
80%
Non-Returners, 71%
Returners, 89%
100%
0%
Pap Test
(p=0.033)
Mammogram
(p=0.000)
Females
Prostate Screening
(p=0.000)
Males
Source: Department of Health and Ageing (2004)77 Notes: Calculation of the Pap test p value excludes respondents who indicated that they did not need to have a Pap test in the past 2 years, and therefore differs from the p value shown in the report Summary  Physical/mental health The assumption that people living with physical and mental health problems or disabilities would be less likely to participate in screening has face validity (i.e. seems reasonable). However, the evidence to support this assumption is not strong. In part, this appears to be due to: (i) the fact that the issue that has received very little attention in the literature; and (ii) the broad range of physical and mental health problems included under the umbrella of ‘physical and mental health’. The most compelling (but not strong) evidence in the literature concerns the uptake of FOBT. For this test, increased uptake appears to be associated with: (i) the ability to perform ADL; and (ii) having a chronic/co‐morbid condition. However monitoring data for the uptake of FOBT in the Australian NBCSP indicates the reverse; people with severe limitations are more likely to participate than those who do not.  Drug and alcohol use A small amount of evidence was identified relating to the association between smoking and alcohol consumption on the uptake of screening mammography, but no evidence was found in relation to Pap tests and FOBT. The evidence in relation to smoking is conflicted. However, the most recent meta‐analysis indicated lower rates of uptake among women Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 57 who smoke cigarettes. The evidence in relation to alcohol consumption has consistently shown that alcohol consumption is associated with an increase in mammogram uptake.  Previous participation in screening The screening behaviours considered in this section related to: (i) previous dissimilar screening (e.g. the association between participation in prostate cancer screening on participation in FOBT, or of participation in screening mammography on having a Pap test and vice versa); and (ii) repeat screening (e.g. the impact of having a mammogram on future uptake). On balance, the evidence appears to indicate that people who participate in screening mammography, Pap test and FOBT screening are more likely to participate in repeat screenings than those who have not participated or refused to participate in earlier rounds. There is also evidence to support the proposition that participation in other types of screening is associated with the increase in the uptake of mammograms, Pap tests and FOBT. VOL 2– 58 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 7 CULTURAL FACTORS Many of the reports and reviews included in this literature review point to the importance of considering cultural factors in relation to cancer screening. 21, 31 However, it is also clear that many of the factors that inhibit cancer screening among Indigenous and ethnic groups are also relevant to Caucasian groups. For example, in a review of the factors associated with cancer screening among Hispanic women in the USA published in 2002, the authors pointed out that that low socioeconomic status, poverty, low levels of education, lack of knowledge, cost of screening as well as acculturation have been established as reasons for low screening rates among this group of women.53 Similarly, Condron and Fernbach (2002) undertook an evaluation of the 2001 PapScreen Victoria multicultural media campaign targeting Arabic‐, Mandarin‐, Polish‐, Russian‐ and Serbian‐speaking groups. They concluded that some factors that act as barriers for women from these ethnic groups to having a Pap test were also relevant for English‐
speaking women.78 The factors included in this section of the review are those that appear to have been most consistently associated with cancer screening among ethnic and Indigenous groups, namely: acculturation, fatalism, modesty/embarrassment/shame, medical mistrust, collectivism/communalism, and spirituality/religiosity. In reading this section it is also important to bear in mind that there is diversity within and between Indigenous and ethnic groups.79 As Newman et al. (2008) point out: ‘An understanding of the diversity of conceptions and beliefs about health and illness that may be held by Aboriginal people is essential for developing more effective cancer prevention … While some issues may be broadly relevant to others of the many ethnic and cultural groups that make up the Australian population, many issues will be differently experienced or expressed by Aboriginal people.’31 7.1 Acculturation Acculturation is the process by which individuals whose primary learning has been in one culture, adopt attitudes, values and behaviours from another culture.80 Acculturation is often assessed by taking into account the length of residency in the host country or proficiency in the host language.24 As shown in Table 34, those not born in the USA are less likely to participate in cancer screening and those who have been in the USA for less than ten years have the lowest participation rates.8 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 59 Table 34 Cancer screening rates by immigration status: USA (2005) Mammogram past 2 years Immigration Pap test past 3 years FOBT within past year Born in USA 67.2% 81.4% 12.5% In USA≥ 10 years 63.3% 73.0% 9.1% In USA< 10 years 50.0% 66.8% 2.6% Source: Cokkinides et al. (2007)8 Notes: FOBT: Faecal occult blood testing  Screening mammography Two studies included in the Tejeda et al. (2009) citation indicate that, among Hispanic women in the USA, acculturation is negatively associated with the uptake of mammograms (2 studies: published 1999 and 2005). However, other studies described in the same citation show that the effect attenuates after adjusting for socioeconomic factors (2 studies: published in 1987 and 2005). In a review of the literature about breast screening among Chinese, Korean, Filipino, and Asian‐Indian American women published in 2005, Wu et al. reported that: o 40% of studies (4 of 10) showed that, among Chinese women, the inability to speak English was a significant barrier; women who spoke English fluently were about four times more likely to have had a mammogram than those with poor English fluency o 33% of studies (1 of 3) showed that Korean American women who had resided longer in the United States were more likely to be screened o Filipino women who had spent more time in the United States were more likely to be adherent to breast screening procedures. 70 In a review of the literature relating to Latin American women in the USA, Wells et al. (2007) concluded that the relationship between immigration status and adherence to screening mammography is ‘unclear and appears to be associated with the population sampled’. The review included five studies that evaluated the relationship between acculturation and screening status: o 1 study found that, among immigrant women, naturalised citizens were more likely to be screened than non‐citizens (adjusted OR 1.57) o 1 study found that foreign‐born Hispanic women were more likely to participate in screening mammography than non‐Hispanic white women (adjusted OR 2.15) o 1 study found that Hispanic women born in the USA were as likely to participate in screening mammography as non‐Hispanic white women o 2 studies that reported ORs did not find an association between foreign birth and screening mammography. Schueler et al. (2008) found that ‘poor English language skills’ and ‘foreign birth’ showed only modest clinical relevance in the multivariate model and the results were not VOL 2– 60 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening statistically significant. However, being a recent immigrant did appear to have an important negative impact on the uptake of mammograms (Table 35).26 Table 35 Impact of acculturation on the mammogram uptake: Schueler et al. (2008) review Number of studies Adj OR 95% CI Recent immigrant 6 0.54 0.37–0.79 Poor English 7 0.77 0.58–1.02 7 0.84 0.64–1.12 Foreign born 26
Source: Schueler et al. (2008) Notes: Adj OR: Adjusted odds ratio; CI: Confidence interval  Pap tests Austin et al. (2002) found that in Canada and the United States: (i) the ability to speak English correlated positively with adherence to cancer screening guidelines, especially among older Hispanics; and (ii) women speaking only, or mostly, Spanish were consistently less likely to be screened for cervical cancer.53 In a single study of the factors associated with Pap testing, Ackerson et al. (2007) found that the inability to speak English, living in the USA for less than 5 years, preferring to speak only Spanish and filling out the questionnaire in their native language were associated with lack of cervical cancer screening in the Hispanic population.27  Faecal occult blood testing There was no information in the reviews on the impact of acculturation (however measured) on the uptake of colorectal cancer screening. 7.2 Fatalism The literature includes two definitions of fatalism. The first is the belief that events (such as those related to health) are beyond the control of the individual; they depend on fate or luck.29, 81
The second, which is quite common in cancer research, defines cancer fatalism as the belief that death is inevitable when cancer is present.21, 29 According to the National Bowel Cancer Screening Pilot Program’s Multicultural Working Group (2004), exploration of these concepts has ‘moved beyond anecdotal evidence that fatalism influences screening behaviour to establishing it as an independent predictor of participation in breast, cervical, colorectal and skin cancer screening’. This report indicated that fatalism is a ‘strong theme’ of Australian qualitative cancer research among ethnic communities. In the three research studies documented in the report, a common attitude among participants was the hope that cancer was not one’s personal destiny.21 The report for the Department of Health and Ageing prepared by Blue Moon Research and Planning (2008) as part of the evaluation of the BreastScreen Australia program indicated that fatalism inhibited screening among Greek participants in the qualitative research but fatalism may not be a factor among other participating ethnic groups (e.g. Chinese).71 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 61 Among the literature reviews used as the basis for preparing this review, very little quantitative empirical research was presented that examined the relationship between fatalism and screening behaviour.  Screening mammography The Powe et al. (2003) review included 4 articles focused on breast cancer screening. Only 1 of the studies reported that women with more fatalistic beliefs were less likely to engage in screening. However, it was not clear whether this was based on screening intentions or actual participation in screening.82 The Schueler et al. (2008) review included 14 studies containing a univariate analysis of fatalism and uptake of mammograms and 7 studies containing a multivariate analysis. Higher fatalism scores decreased the uptake of screening: univariate OR 0.74, multivariate OR 0.61 (95% CI 0.40–
0.92) 26  Pap tests The Powe et al. (2003) review included only 1 article focused on cervical cancer screening. However, the authors report that the article had methodological problems in relation to the determination of a fatalism score and that there did not appear to have been any attempt to look at the relationship between fatalism and the uptake of cervical cancer screening.82  Faecal occult blood testing In the Petersen (2002) review, the tendency to be more fatalistic and the attitude that one is not in control of one’s health were reported to be associated with lower adherence to colorectal cancer screening.5 The Powe et al. (2003) review included 3 studies that investigated the relationship between fatalism and colorectal cancer screening. However, only 1 study examined the relationship between fatalism and uptake of FOBT among African‐American and Caucasian participants. The results indicated that African‐American women had significantly higher levels of cancer fatalism than other participants, and cancer fatalism was ‘the only statistically significant predictor of FOBT in the study’.82 7.3 Modesty/embarrassment/shame According to Austin et al. (2002), culturally based embarrassment and similar emotions are a major barrier to breast and cervical cancer screening among Hispanic women in the USA. In one study, embarrassment was found to be a stronger predictor of screening than perceived susceptibility and perceived benefits of early detection. The authors’ hypothesised that embarrassment about discussion of private body parts and embarrassment about exposing private body parts during a physical examination may pose a barrier for some Hispanics, especially if examined by a male physician.53 The review by Magai et al. (2007) contained 1 study that looked at the relationship between embarrassment and breast cancer screening among women from six ethnic groups. It found that greater embarrassment predicted poorer screening when other variables were controlled (i.e. age, socioeconomic status, physician recommendation, anxiety and emotion regulation style): women with high embarrassment scores were 29% less likely to screen.55 The 2008 review by Schueler et al. contained 4 univariate and 3 multivariate studies that examined the relationship between modesty/embarrassment and uptake of mammograms. VOL 2– 62 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Modesty/embarrassment was associated with a decrease in screening rates (univariate OR 0.54; multivariate OR 0.55; 95% CI 0.39–0.76). However, there is no way of telling from the review whether or not the studies related specifically to ethnic groups. If they do not then this is a factor that is generalisable across groups.26 In the Australian context, there is some evidence from a qualitative study undertaken as part of the evaluation of the BreastScreen Australia program that ‘embarrassment may be a more difficult barrier to overcome in relation to screening mammography for some women from non‐
English speaking backgrounds’.71 The report stated that: ‘Iraqi and Lebanese women associated mammograms with discomfort, humiliation and invasive handling of their breasts. This was most acute for one Lebanese participant, who felt that mammograms were against her Muslim religion as they required a woman to reveal her breasts to a stranger.’71 However, the report did also note that this was not an issue raised by all Muslim women.71 Shahid et al. (2009) noted that within Indigenous communities, there are beliefs that cancer is related to feelings of shame. No studies were cited examining the relationship between shame (either because cancer is consider a ‘white man’s disease’ or because cancer is seen as a ‘curse’ for wrongdoing) and screening uptake. However, it can be hypothesised that such shame may deter Indigenous people from accessing screening.79 7.4 Medical mistrust According to Deshpande et al. (2009), trust in the medical profession involves the belief that individuals and institutions will act appropriately and will perform competently, responsibly, and in a manner consistent with patients’ interests.29 According to Wolff et al. (2003), there is considerable evidence that many African‐Americans believe this standard is not met regarding their care.30 This mistrust has been linked to a variety of health‐related decisions, behaviours, and interactions, and has been associated with negative attitudes about cancer screening and decreased use of mammography.29 This issue is also relevant to the Australian health care system. The Federation of Ethnic Communities’ Councils of Australia cite lack of trust as a major issue that may make people from ethnic communities reticent to access services, including health services. One of the antecedents for this could be experiences in a person’s country of origin and this may be a particularly important consideration when trying to engage with humanitarian entrants into Australia.3712 12
It should be noted, however, that medical mistrust is not confined to members of ethnic and Indigenous groups. Blue Moon Research and Planning (2008) classified women who never screened and who had active reasons for not participating as women who ‘tended to be from higher socioeconomic backgrounds and/or were relatively highly educated… They were most likely to be Anglo‐Australians, although the literature review showed some Chinese women had similar views’. These women were ‘often sceptical about information provided by health professionals, BreastScreen Australia and the Department of Health and Ageing'. 71 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 63 7.5 Collectivism/communalism Collectivism is the tendency to subordinate personal goals to those of the group. In this strategy, the basic unit of society is the family, community, or tribe rather than the individual. Collectivism is expressed in values of group survival and security, concern and responsibility for others, and respect for traditions and elders. According to Deshpande et al. (2009), studies suggest African Americans are more likely than European‐Americans to hold collectivist beliefs.29 This issue is also relevant to Indigenous communities in Australia and elsewhere.31, 79 Newman et al. (2008) pointed out that ‘it is widely acknowledge that an Aboriginal conception of health encompasses the physical, social, emotional, cultural and spiritual wellbeing of not only the individual but of the community and the environment as a whole’. Such beliefs ‘emphasise the capacity of Aboriginal people to perform social and domestic roles as criteria for wellness’ and suggest that health‐related treatments that interrupt the individual’s capacity to engage in normal social activities would be viewed as ‘unhealthy’.31 Although this review focuses on the factors associated with cancer screening, it can be argued that, Aboriginal people may be reluctant to undergo screening for the same reasons that they are reluctant to take up cancer treatment: because of the disruption it would cause to performing daily activities and fulfilling family roles. Deshpande et al. (2009) stated that no studies have clearly linked collectivist beliefs and ethnic identity to health behaviours.29 Shahid et al. (2009) pointed out that further community‐based research is needed to understand Indigenous perspectives and needs, and the way that these affect care‐seeking behaviour. They also pointed out that there has been no exploration of the differences, if any, between urban, rural and remote populations.79 Finally, Newman et al. (2008) pointed out that that they could find no research focusing on the experiences of Aboriginal men in relation to cancer. 31 7.6 Spirituality/religiosity In 2009, Deshpande et al. defined spirituality as the ‘internal manifestation of belief in a higher power and the genuine, consistent commitment to its attendant values’. They pointed out that it contains, but is not limited to, adherence to doctrines, beliefs, and the ritual practices of religious institutions or a person’s level of religious organisation involvement. Religiosity has been shown to be particularly high among older African‐American women.29 According to Lende et al. (2009), religiosity has been shown to have an impact on breast cancer screening among African‐American women.83 However, Deshpande et al. (2009) point out that in some studies, religiosity has been negatively associated with health‐promoting behaviour, while others have found spirituality to be positively associated with health behaviours.29 Summary This section has focused on factors that have most often been associated with Indigenous and ethnic groups in relation to screening activity. Interpretation of the results should take account of the fact that: (i) ethnic and Indigenous groups are heterogeneous; and (ii) some of the factors examined in this section may also apply to non‐ethnic and non‐Indigenous groups. VOL 2– 64 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening  Acculturation This factor has generally been operationalised as ‘length of residence in the host country’ and/or ‘ability to speak English’. No evidence was found in relation to the impact of acculturation (however defined) on the uptake of FOBT. Evidence for a relationship between acculturation and uptake of screening mammography was mixed, with the clearest evidence suggesting lower uptake among those who have most recently arrived in the host country. There is less evidence of a relationship between acculturation and uptake of Pap tests, but what evidence there was appears to point to lower uptake among those with lower English language skills and among the most recent immigrants.  Fatalism Fatalism has been defined as the belief that either there are events that are beyond the control of the individual and/or the that death is inevitable when cancer is detected. Little quantitative research was found relating to the relationship between fatalism and screening uptake. What evidence there was indicated that higher levels of fatalism (however, defined) were associated with decreases in the uptake of mammograms and FOBT. There was no evidence in relation to the association between fatalism and uptake of Pap tests.  Modesty/embarrassment/shame On balance, the evidence indicated that modesty, embarrassment and shame inhibit breast and cervical cancer screening among ethnic and Indigenous groups.  Medical mistrust; collectivism/communalism; spirituality/religiosity Whilst there is research which suggests that these factors have an impact on screening uptake, only a small amount of literature exists that specifically examines the relationships. One review concluded that medical mistrust had a negative impact on the uptake of screening mammography but there was insufficient evidence to draw any conclusions about collectivism/communalism. The evidence for spirituality/religiosity was inconclusive. There was no evidence in relation to the relationship of these factors with uptake of Pap tests and FOBT. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 65 8 HEALTH SYSTEM FACTORS According to Hausmann‐Muela et al. (2003), it has become popular among researchers to use four categories to group key health system factors that affect health‐seeking behaviour. 84 In this review, these four categories have been used to structure the evidence in relation to health system factors that may impact on the uptake of screening mammography, Pap tests and FOBT. 1. Availability: refers to the geographic distribution of health services and facilities. Having a ‘usual source of care’ and ‘utilisation of health care services’ have been included as factors relevant to this category. 2. Accessibility: refers to how easy or difficult it is to use the services that are available and includes such things as transport, roads and opening hours. Accessibility can be understood in terms of convenience for potential users. Included in the discussion for this category is ‘provider recommendation’ of a screening test. 3. Affordability: refers to the costs incurred by individuals, households or families in using services. Health insurance reduces out‐of‐pocket costs where screening services are not provided free of charge and has been included in this category. 4. Acceptability: refers to the cultural and social distance between services and users. The focus is mainly on the characteristics of the health providers, excessive bureaucracy, and cultural safety. Each of these categories is discussed independently, although they are interrelated. As Wolff et al. (2003) state: ‘Many medically underserved African‐Americans lack a primary care physician and access the health care system only in an urgent or emergent care situation … Primary care clinics and doctors may not be conveniently available in the geographic areas where these underserved population live, tend to be overburdened, and may not schedule evening and weekend hours. The underserved may also have to travel to different facilities to obtain screening tests and endure long waiting times. The cost and inconvenience of public transportation may also deter regular visits to a doctor.’30 8.1 Availability In their 2008 review, Schueler et al. found a negative relationship between rurality and uptake of screening mammography (Adj OR 0.75, 95% CI 0.63–0.90).26 However, in the 2010 review by Guessous et al., only 30% of studies (3 of 10) found that rural residence was a significant barrier to the uptake of colorectal cancer screening.3 There is ample evidence in the monitoring data of geographic differences in screening rates in countries that do have population based screening programs (eg Australia and the UK) and those that do not (eg USA) (Table 36, Table 37, Table 43). In Australia, the patterns within the geographic variations differ across tests and in Victoria at least, there is evidence of marked changes in geographic participation rates over time (Table 39) VOL 2– 66 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Table 36 Geographic variations in screening rates in Australia Major cities Inner regional Outer regional Remote Very remote 54.6% 61.1% 62.2% 60.1% 49.5% 54.5‐54.7 (a) 60.9‐61.3 (b) 61.9‐62.5 (c) 59.2‐61.0 (b) 48.3‐50.8 (d) 62.5% 61.2% 58.9% 53.6% 54.0% 62.4‐62.5 (a) 61.0‐61.3 (b) 58.7‐59.1 (c) 53.1‐54.1 (d) 53.3‐54.7 (d) 35.2% 40.1% 39.1% 34.6% 25.0% 35.1‐35.4 (a) 39.9‐40.4 (b) 38.8‐39.5 (c) 33.7‐35.4 (a) 23.9‐26.1 (e) Mammograms ASR 2005‐2006 95% CI Pap tests ASR 2006‐2007 95% CI FOBT Crude Rate 2008 95% CI Sources: Mammograms: Australian Institute of Health and Welfare (2009)85; Pap tests: Australian Institute of Health and Welfare (2009)9; FOBT: Australian Institute of Health and Welfare and Australian Government Department of Health and Ageing (2009)12 Notes: Geographic classification based on the Australian Bureau of Statistics (ABS) Australian Standard Geographic Classification (ASGC) ASR: Age standardized rate CI: Confidence interval FOBT: Faecal occult blood test (a), (b), (c), (d), (e) denote overlapping confidence intervals within each screening test. Table 37 Geographic variation in screening rates in the USA Age range (years) Average Mammogram 40–64 59.7% 45.7% Utah – 71.0% Rhode island Pap test 18–64 85.6% 74.4% Utah – 91.9% Maine FOBT 50–64 13.4% 5.1% Utah – 19.6% District of Columbia Test Uptake Range 8
Source: Cokkinides et al. (2007) Notes: FOBT: Faecal occult blood testing Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 67 Table 38 Five‐year screening rates for NHS Breast Screening Programme: uptake of invitations to screen for women aged 50–70 years in selected geographic areas 2003–04 2007–08 Strategic Health Authority Rate Rank Rank Rate East Midlands 80.2% 1 1 78.6% South West 77.7% 4 4 75.6% North West 74.5% 9 9 72.3% London 62.5% 10 10 60.6% Difference: Rank 1 – Rank 10 17.7% 18.0% Source: Downloaded from the NHS website on 24 November 2009 from: http://www.ic.nhs.uk/webfiles/publications/breastscreening0708/2007‐08%20Breast%20Screening%20 Table 39 Geographic variation in BreastScreen Victoria 2‐year participation rates (50–69 age standardised rates) Major cities Inner regional Outer regional Remote 2004–2005 58.0% 56.2% 59.9% 72.5% 2005–2006 56.9% 56.5% 60.9% 49.7% Source: Productivity Commission Excel spreadsheet downloaded 15th October 2009, available at http://www.pc.gov.au/__data/assets/excel_doc/0004/85414/53‐chapter12‐attachment‐only.xls Table 12A.12 Notes: Residents of Victorian postcodes allocated to the Albury/Wodonga catchment (NSW jurisdiction) are included in Victoria’s population estimate, accounting for the slight decrease in participation rates compared to those published by BreastScreen Victoria. The reasons for these geographic variations have not been clearly articulated. Akers et al. (2007) noted that, in terms of health system infrastructure, the availability of hospitals, primary care, sub‐specialty providers and quality laboratory services is relatively limited in rural areas of the USA and this is likely to impact negatively on women’s ability to obtain cervical cancer screening services.24 Whilst this argument has face validity, the authors did not present any evidence to support their hypothesis. The following analysis in relation to the uptake of screening mammography, Pap tests and FOBT concentrates on the impact of having, or not having, a regular source of care and, where data are available, the utilisation of a doctor’s services.  Screening mammography Jepson et al. (2000) reviewed five studies of the association between the number of previous visits to a health care provider on uptake of screening mammography. The authors concluded that the majority of studies did not a find a significant association between uptake and the number of visits women make to their health care provider.6 In a review of the uptake of screening mammography by Latin American women in the USA, Wells et al. (2007) found that: VOL 2– 68 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening o having a usual source of health care was strongly associated with the uptake of mammograms (adjusted ORs ranged from 1.24 to 3.4); 1 study investigated different types of usual care providers (primary care physicians, other physicians, other facilities and non‐physician providers) and found that all were positively associated with screening uptake o recently receiving health care from a doctor was associated with higher adjusted OR of receiving a screening mammogram (adjusted OR 1.15–1.47); 2 studies evaluated the type of recent visit and found that: ‒ outpatient, office or primary care visits were associated with a higher odds of receiving screening mammography ‒ visits to the emergency room were associated with lower or the same odds as not being seen in the emergency room ‒ being hospitalised was associated with lower odds of receiving screening mammography.10 As shown in Table 40, Schueler et al. (2008) found that (i) not visiting a doctor within the past year, and (ii) not having a primary care doctor or usual source of care were associated with a reduction in the uptake of mammograms.26 Table 40 Relationship between availability and the uptake of mammograms: Schueler et al. (2008) review Number of studies Adj OR 95% CI No physician visit within year 17 0.34 0.25‐0.47 No primary care physician/usual source of care 33 0.41 0.32‐0.53 26
Source: Schueler et al. (2008) Notes: Adj OR: adjusted odds ratio; CI: Confidence interval  Pap tests Condron and Fernbach (2001) undertook an evaluation of the PapScreen Victoria Neighbourhood Grants Program to examine which steps in the process of getting a Pap test were most difficult for women. As shown in Table 41, compared to well‐screened women, other women were more likely to find choosing a health care practitioner more difficult. However other steps such as making an appointment and turning up for it were not more difficult for the non well‐screened women.86 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 69 Table 41 Comparison of difficulty in the steps necessary to obtain a Pap test with well‐
screened women Decision point Compared to well‐screened women OR 95% CI Deciding to have a Pap test 0.24 0.11–0.53 Choosing a practitioner 0.29 0.13–0.61 0.73 0.34–1.59 0.42 0.16–1.11 Statistically significant difference No statistically significant difference Making an appointment Turning up to the appointment 86
Source: Condron and Fernbach (2001) Notes: OR: Odds ratio; CI: Confidence interval Ackerson et al. (2007) found that lack of an established usual source of health care influenced Pap testing in 3 of 7 studies (43%). For example, women without a usual source of health care were not aware of the available services within their community to obtain low‐cost Pap smears and were less likely to obtain regular cervical cancer screening. However, one study found that having a usual source of health care did not influence whether African‐American women obtained Pap tests.27  Faecal occult blood testing The Jepson et al. (2000) review included one study that looked at the association between uptake of FOBT and having a regular source of care for older black Americans. In this study, having a regular source of care was found not to be a significant factor in the uptake of FOBT.6 In 2002, Petersen concluded that not having a doctor was associated with lower adherence to colorectal cancer screening in the USA.5 The review by Guessous et al. (2010) reported that among people aged 65 years of age and older: o 77% of studies (17 of 22) found that having a usual source of care/personal physician significantly improved the uptake of colorectal cancer screening o 94% of studies (17 of 18) found that routine preventive care/primary care/doctor visits in the recent past significantly improved the uptake of colorectal cancer screening.3 8.2
Accessibility  Screening mammography In the Jepson et al. (2000) review, 50% of studies (2 of 4) found a positive association between uptake of mammograms and the recommendation of health care providers. The authors concluded that it was not clear whether recommendations from health care providers were associated significantly with uptake of screening mammography. VOL 2– 70 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening According to Austin et al. (2002), recommendation by a doctor is one of the most important cues to cancer screening and is more important than the availability of services. However, this review only included studies undertaken in the USA and, as the authors point out, the majority of women in the USA who have mammograms are referred by their doctor.53 The Austin et al. (2002) review reported on 2 studies of Hispanic women in the USA in which close to 100% of the women had visited a doctor in the last year, over three‐
quarters of whom indicated that nobody had ever suggested getting a mammogram. The authors indicated that ‘respect for authority (respect) is an important characteristic of Hispanic culture. Latin Americans regard doctors as powerful authority figures and have a tendency to listen to what doctors say, but rarely show self‐initiated health care behaviours’.53 Wu et al. (2005) reviewed the factors associated with breast screening practices among Chinese, Korean, Filipino, and Asian‐Indian American women and concluded that a doctor’s recommendation influenced screening mammography participation.70 In the Australian context, BreastScreen Australia is a self‐referral service. However, Blue Moon Research and Planning (2008) indicated in their review that: ‘Numerous studies point to the positive influence of a doctor’s recommendation or endorsement on breast cancer screening behaviour. 13 In particular, one article highlighted that studies have shown that between 68% and 91% of women will attend for screening after a recommendation from their GP. Another study indicated that there is a link between a doctor’s involvement and re‐screening activity. While a doctor’s recommendation or referral seems to be influential across audiences, one study suggests this can be particularly beneficial among women from non‐English speaking backgrounds.’71 In a review of 23 studies, Schueler et al. (2008) found that the lack of a doctor’s recommendation had a negative impact on the uptake of mammograms (Table 42).26 Table 42 Relationship between accessibility factors and uptake of screening mammography: Schueler et al. (2008) review Number of studies Adj OR 95% CI No physician recommendation 23 0.16 0.08‐0.33 Poor healthcare access 6 0.76 Not available 26
Source: Schueler et al. (2008) Notes: Adj OR: Adjusted odds ratio; CI: Confidence interval  Pap tests In a review published in 2007, Ackerson et al. concluded that African‐American and Hispanic women who did not receive a recommendation to obtain cervical screening from 13
11 references, see the Blue Moon report for the primary sources. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 71 a health care provider were less likely to obtain Pap tests. One of the studies included in the review indicated that participation among Hispanic and African‐American women who reported not receiving a recommendation for a Pap test was 24.3% compared to 75.7% among those who did report receiving a recommendation. However, none of the studies reported an actual chart audit to evaluate whether or not documented provider recommendations coincided with the participants’ recollections.27 Ackers et al. (2007) have pointed out that transportation problems, such as large distances between women’s residence and health facilities and the absence of mass transit systems, may also present barriers to regular screening in rural areas.24  Faecal occult blood testing In the Vernon et al. (1997) review, 75% of studies (3 of 4) showed a positive association between doctor advice/ability and FOBT and one (25%) found no association.2 The Jepson et al. (2000) review included 2 studies that examined the influence of health care professionals on the uptake of FOBT. In neither was there a statistically significant effect for the influence of health care professionals.6 Petersen (2002) concluded that: (i) absence of a doctor’s recommendation; (ii) lack of information from health care providers; and (iii) distance or lack of transport to health care providers, were all associated with lower adherence to colorectal cancer screening.5 The Guessous et al. (2010) review included 7 studies that considered the impact of a doctor’s recommendation. All 7 studies found that lack of a recommendation was a significant barrier to the uptake of colorectal cancer screening in people aged 65 years and older.3 8.3 Affordability In the literature, affordability has most often been operationalised as ‘having health insurance’. As shown in Table 43, just over half of the studies included in the Jepson et al. (2000) review reported a significant positive association between having insurance and the uptake of screening mammography. However, few studies examined this association in relation to Pap tests and FOBT.6 Table 43 Uptake of cancer screening tests and insurance status: Jepson et al. (2000) review Screening mammograms Pap tests FOBT Studies with statistically significant effects (p ≤ 0.05) 7/12 (58%) 2/4 (50%) One study found an association between the type of insurance and uptake Authors’ conclusions The majority of studies found a significant association between uptake and insurance status. Those who had some form of insurance were, in general more likely to attend than those who did not Unclear whether insurance status is associated with uptake No conclusions reached Source: Jepson et al. (2000)6 VOL 2– 72 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening As shown in Table 44, monitoring data in the USA appear to indicate that having health insurance is associated with an increase in the uptake of screening mammography, Pap tests and FOBT in that country.8 Table 44 Association between health insurance and the uptake of cancer screening in the USA (2005) Mammogram past 2 years Pap test past 3 years 69.8% 82.4% 12.7% 33.2% 61.4% 3.1% Yes: Have insurance No: Do not have insurance FOBT within past year 8
Source: Cokkinides et al. (2007)  Screening mammograms In the Wu et al. (2005) review, 50% of studies (5 of 10) found that that lack of money and health insurance influenced mammography use in Chinese women. The review also found that a ‘significant correlate’ of Korean Americans having insurance coverage was an annual income greater than $25,000.70 In the Wells et al. (2007) review of factors associated with the uptake of mammograms among Latin American women in the USA, 11 studies reported health insurance as a predictor of screening mammography in multiple logistic regression analysis. Adjusted ORs ranged from 1.7 to 8.5. Compared to having no insurance, both fee‐for‐service and Health Maintenance Organisation (HMO) insurance were associated with higher odds of obtaining a mammogram.10 Schueler et al. (2008) concluded that, although there was variability in the magnitude of the results, nearly all studies included in their review showed that having no insurance was a strong predictor of not undergoing screening mammography (Table 45). However, the impact varied across ethnic groups. The effect was stronger among white and Chinese women than among African‐American and Latin American women (p < 0.001).26 Table 45 Relationship between financial factors and uptake of screening mammography: Schueler et al. (2008) review No insurance Public vs private insurance Number of studies Adj OR 95% CI 40 0.47 0.39–0.57 12 0.74 0.63–0.86 26
Source: Schueler et al. (2008) Notes: Adj OR: Adjusted odds ratio; CI: Confidence interval  Pap tests In the Ackerson et al. (2007) review, out‐of‐pocket expenses were found to influence the uptake of Pap tests: women having to pay out‐of‐pocket costs were less likely to be screened. This included women with insurance plus a co‐pay charge or the expense of an office visit or laboratory testing.27 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 73 The 2007 Akers et al. review indicated that having insurance increased the uptake of Pap tests, but they also concluded that the type of insurance had an impact. For example: o in a study to evaluate a culturally appropriate cervical cancer intervention, women with private health insurance were more likely to be screened compared with those who were uninsured and those who were covered by the public Medicaid or Medicare insurance o an evaluation of three interventions to increase cervical cancer screening rates in a multi‐ethnic sample found that having private insurance was one of the strongest predictors of cervical cancer screening behaviour.24 However, Akers et al. (2007) point out that several studies have demonstrated that cervical cancer screening may not occur among insured women even when they have made multiple health care visits. They concluded that this indicates that insurance coverage alone does not ensure the uptake of Pap tests.24  Faecal occult blood testing Petersen (2002) concluded that a lack of health insurance is associated with lower adherence to colorectal cancer screening.5 Guessous et al. (2010) reported that in 71% of studies (5 of 7), having private/supplemental insurance (including Medigap) was a significant facilitator of the uptake of colorectal cancer screening in people aged 65 years and over.3 In 2006, Wong et al. reported on a telephone survey in Hong Kong of attitudes towards colorectal cancer screening among 18–55 year olds in Hong Kong. As shown in Table 46, younger and older respondents were less likely to indicate a willingness to join a paid program than a free program.87 Table 46 Effect of cost on willingness to join a colorectal cancer screening program in Hong Kong (young and old age groups only) Type of screening program Willingness to join the program Definitely Free Paid Probably Undecided Definitely not 29.5% 23.9% 9.6% 37.0% 7.9% 21.7% 13.0% 57.4% 87
Source: Wong et al. (2006) 8.3
Acceptability According to the Federation of Ethnic Communities’ Councils of Australia (2006), ‘newly arrived migrants’ lack of knowledge about existing health and social services, and lack of transportation may be compounded by services providers’ inability or lack of interest in providing culturally sensitive services that address clients’ needs’.37 There is a paucity of data relating to the cultural safety of services and the cancer screening uptake. The issue most often discussed relates to the gender of the service provider in relation to mammograms and Pap tests. VOL 2– 74 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 
Screening mammography The Jepson et al. (2000) review contained 3 studies that examined the influence of healthcare provider gender on uptake of screening mammography. Only one of the studies found a significant association.6 One study in the Wu et al. (2005) review examined language concordance between doctors and Chinese women and found that women who had a female physician and Chinese language ‘concordance’ had the greatest likelihood of prior and recent screening.70 In the Australian context, qualitative research by Blue Moon Research and Planning (2008) indicated that: ‘Across non‐English speaking audiences, there was an expectation that access to interpreters would be limited. Lapsed users in these groups confirmed this was their experience of using BreastScreen Australia services. Both lapsed and never participants felt that having access to either an interpreter or bilingual health worker would increase comfort levels and encourage them to screen regularly. Assistance of this kind was seen as particularly important during a first visit to help explain unknown factors and procedures. This issue was most crucial for recent arrivals to Australia.’71  Pap tests According to Ackerson et al. (2007), the gender of the doctor providing screening services is correlated with the provision of cervical cancer screening, with female doctors ‘consistently noted’ to screen more women than male doctors.24 It has been hypothesised that ethnic minorities often prefer to have a physician of the same ethnicity to minimise cultural and linguistic barriers to care. However, Ackerson reported ‘several’ studies which noted that having a physician of the same ethnicity may be associated with lower rates of cervical cancer screening. Unfortunately, none of the studies examined other doctor characteristics (e.g. gender, physician type, or being trained outside the USA), so it is not clear whether the findings are mediated by other factors.24 Summary  Availability Monitoring data provide clear evidence of geographic differences in uptake of screening. This was also apparent in the most recent literature review relating to the uptake of screening mammography. There is less evidence in the literature relating to Pap tests and FOBT. Reasons for geographic differences in uptake have been proposed but have not been examined in the literature. Evidence indicates that people with a ‘usual source of care’ are more likely to be screened (especially for breast and colorectal cancer) and that recent visits to a health care provider are associated with higher screening rates. However, the results for uptake of screening mammography indicate that the type of health care provider has an influence; the association does not appear to hold for visits to emergency rooms or being Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 75 hospitalised. There is little evidence that the number of visits during a particular time period impacts on the uptake of screening.  Accessibility The major issue discussed in the literature in relation to accessibility is a recommendation for screening by a doctor. Recent evidence indicates that recommendation by a doctor is associated with uptake of breast and colorectal cancer screening. For mammograms, this result appears to be pertinent even in Australia where BreastScreen Australia is a self‐
referral service. It is also consistently pointed out that this factor may have particular relevance for people from some ethnic groups.  Affordability In the USA, having health insurance lowers the cost of screening services and having insurance has been associated with an increase in uptake of screening mammography, Pap tests and colorectal cancer screening including, but not limited to, FOBT. Having health insurance is likely to be of little relevance in those health systems that provide cancer screening services free of charge to eligible individuals within the target population.  Acceptability There is some evidence that female health care providers are associated with the greater uptake of breast and cervical cancer screening. It also appears that, in Australia at least, the provision of interpreters may increase the uptake of mammograms by NESB women. VOL 2– 76 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 9. DISCUSSION AND IMPLICATIONS 9.1 Discussion Literature on screening adherence has moved away from a reliance on demographic and structural variables as sole predictors of adherence to include cognitive and, more recently, emotional variables as well as health and lifestyle factors.55 It is clear that numerous, interacting factors are associated with screening uptake; 27 such factors have been examined in this review. Before presenting a summary of the evidence in relation to these factors, a number of caveats need to be outlined.  The data used in relation to associations between demographic variables and uptake of screening tests came primarily from two sources: (i) monitoring data for population‐based screening programs from various countries; and (ii) reviews of factors associated with screening contained in peer‐reviewed literature. In many instances there are discrepancies between these two sets of data, due in part to the fact that the conclusions in literature reviews were often based on multivariate analyses that controlled for other variables, whereas the monitoring data were not. 
Even within the literature reviews, there were differences of opinion, with earlier conclusions being contradicted in more recent reviews. These differences may be due in part to: (i) the inclusion of a larger number of studies in the later reviews; (ii) differences in review methodologies; and (iii) differences in the populations included in the reviews. 
Because of the lack of evidence in relation to FOBT screening, the reviews by Petersen (2002) and Guessous et al. (2010) were included. These reviews examined factors related to the uptake of colorectal cancer screening and did not focus specifically on FOBT. 
The analysis for each of the factors is based primarily on literature reviews in both the peer‐reviewed and grey literature. The amount of evidence for each factor was dependent on the evidence contained in those reviews. In some instances, there is inadequate evidence on which to base firm conclusions. Stronger conclusions may have been reached if a review of primary sources had been conducted for each factor. This was, however, precluded by resources and time constraints. 
Most of the evidence in this volume relates to breast cancer screening. It cannot be, and has not been, automatically assumed that the results for screening mammography will generalise to other cancer screening tests. Bearing these caveats in mind, a brief summary of the results for each factor is shown in Table 47. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 77 Table 47 Summary of the evidence Domain Factor Association with screening uptake Demographic Age 


For screening mammography and Pap tests, participation appears to be lower in groups at the younger end of the target age for population‐
based screening programs. For Pap tests, there appears to be a decline in participation as women move towards the upper limits of the target age. For FOBT, it appears that participation is likely to increase up to about 70 years of age when it will decline. Gender 
On balance, the evidence suggests that uptake of FOBT will be higher among women than men, although gender differences may be smaller in older age groups. Ethnicity 
Monitoring data indicate lower participation among ethnic and Indigenous groups. It appears that ethnicity may not be a significant factor in the uptake of screening once other factors have been controlled for (e.g. socioeconomic status). 
Socioeconomic status 


There is some evidence that low income may be associated with lower screening rates; however, the latest colorectal cancer review indicated that this association may not hold among older people (≥ 65 years). No evidence was found to suggest that employment status has an impact on screening participation. There is some evidence to suggest that the association of lower screening uptake with low socioeconomic status may not be due to a lack of money but rather a result of the pressures of everyday life, which work against the uptake of screening. Education 
On balance, evidence indicates that low levels of education are associated with a lower uptake of screening mammography, Pap tests and FOBT. Literacy 
Australian literature suggests that literacy levels are lower among groups for whom English is not the first language. There is insufficient evidence to draw conclusions about the impact of low literacy levels on the uptake of screening mammography, Pap tests and FOBT. 
Health literacy (HL) 


HL is related to, but not synonymous with, literacy. There is some evidence that HL is lower among the elderly, ethnic minorities, immigrants, non‐active English speakers, those with limited education and low incomes. There is some evidence that lower HL is associated with: o a lower uptake of Pap tests (this association holds for within ethnic‐
group analyses) o lower scores on proximal measures of participation in colorectal cancer screening (e.g. knowledge, comprehension, information seeking), negative attitudes towards colorectal cancer screening and an increased refusal to participate in FOBT screening o a reduced willingness to seek out and engage with health‐related information. VOL 2– 78 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Domain Factor Association with screening uptake Cancer‐related knowledge 



Perceived risk 

Perceived ambiguity 

Psycho‐social Evidence indicates that perceiving oneself to be at a heightened risk of breast cancer is associated with an increase in uptake of screening mammography, but the effect is small. There is little evidence of an association between perceived risk and uptake of screening in relation to Pap tests and FOBT. A small amount of evidence suggests that higher levels of perceived ambiguity (e.g. in relation to cancer screening) are associated with lower uptake of screening mammography. There is no evidence about the association between perceived ambiguity and uptake of Pap tests and FOBT. Fear/anxiety/ worry 
There is relatively strong evidence to suggest that: (i) fear of screening decreases screening participation, (ii) cancer worry increases screening participation, and (iii) fear of the screening outcome has mixed effects on participation. Coping style 
Little research has been conducted into the impact of coping styles on participation in screening. There is a small amount of evidence to indicate that denial is associated with an uptake in screening and repression with a decrease in screening. 
Social networks 


Health and lifestyle There is some evidence that poor knowledge of screening, the belief that mammograms are harmful and the belief that mammograms are not needed in the absence of symptoms is associated with lower screening mammography rates. There is insufficient evidence to draw any conclusions in relation to Pap tests. There is some evidence that poorer knowledge of colorectal cancer may be associated with lower levels of colorectal cancer screening. There is some evidence that colorectal cancer knowledge is lower in some ethnic groups. Physical/mental Health 

There is some evidence that family and friends may have an influence on the uptake of screening, especially among ethnic groups. There is evidence to suggest that being single, widowed or divorced may be associated with a decrease in the uptake of mammograms. There is little evidence about the impact of different types of social support (emotional, tangible, informative) on screening uptake. The notion that people living with physical and mental health problems or disabilities would be less likely to participate in screening seems reasonable. However the evidence is not strong. In part this is due to: (i) the fact that the issue has not received much attention in the literature reviews (ii) the broad range of physical and mental health problems that could be included in this research that makes it difficult to generalise The most compelling (but not strong) evidence concerns the uptake of FOBT; evidence indicates that increased uptake is associated with: (i) the ability to perform ADL (ii) having a chronic/co‐morbid condition. Australian monitoring data indicates increase in FOBT uptake by people with severe or profound physical limitation. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 79 Domain Factor Association with screening uptake Drug and alcohol use 

Cultural Screening participation 
There is evidence that people who participate in: (i) an earlier screening round are more likely to respond to future invitations to participate (ii) one type of screening are more likely to participate in a different type. Acculturation 
Acculturation has generally been measured as ‘length of residence in the host country’ and/or ‘ability to speak English’. The strongest evidence was for the association between recent arrival in the host country and lower uptake of mammograms and Pap tests. There is some evidence that lower English language skills are associated with a lower uptake of Pap tests. 

Fatalism 



Fatalism has been defined as either the belief that events are beyond the control of the individual and/or the belief that death is inevitable when cancer is detected. There is little quantitative research relating to the relationship between fatalism and screening uptake. Available evidence indicates that higher levels of fatalism are associated with lower levels of uptake of screening mammography and FOBT. No evidence was found in relation to the relationship between fatalism and uptake of Pap tests . Modesty/ embarrassment/ shame 
On balance, the evidence indicates that modesty, embarrassment and shame inhibits breast and cervical cancer screening among ethnic and Indigenous groups. Medical mistrust 
One review concluded that medical mistrust has a negative impact on the uptake of mammograms. (Note: There is some evidence that medical mistrust/scepticism is not confined to ethnic and Indigenous groups but was associated with a refusal to screen among well‐educated Australian Caucasian women). Collectivism/ communalism 
There is some indication that collectivism/communalism has a negative impact on screening uptake, but there was little research was that examined these factors specifically. Spirituality/ religiosity 
Evidence around an association between spirituality/religiosity and uptake of screening mammography is inconclusive. No evidence was identified in relation to the effect of spirituality/religiosity and uptake of Pap tests or FOBT. 
Health system The most recent evidence indicates an association between smoking and a reduction in uptake of screening mammography and between alcohol consumption and an increase in uptake of screening mammography. No evidence was found in relation to the relationship between drug and alcohol use and uptake of Pap tests and FOBT. Availability 



Monitoring data provide clear evidence of geographic differences in screening uptake. The literature provides some hypotheses as to why this may be (e.g. lack of services) but evidence Is not definitive. There is evidence to suggest that people with a ‘usual source of care’ are more likely to be screened (especially for breast and colorectal cancer). There is evidence that recent visits to a health care provider are associated with higher screening rates; however, this association does not appear to hold for visits to an emergency room or admittance to hospital. VOL 2– 80 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Domain Factor Association with screening uptake Accessibility 


Affordability 

Acceptability 

9.2 There is strong evidence of a positive association between recommendation by a doctor and uptake of screening mammography and colorectal cancer screening. The literature consistently suggests the relevance of accessibility for people from some, but not all ethnic groups (see also medical mistrust). The association between a doctor’s recommendation and screening uptake is evident in Australia, where BreastScreen Australia is a self‐
referral service There is strong evidence of a positive association between having insurance and increased uptake of screening in the USA. The issue of affordability may be of little relevance in health systems that provide free screening to the target group. There is some evidence that female health care providers are associated with increased uptake of breast and cervical cancer screening. There is some evidence that, in Australia at least, provision of interpreters may increase the uptake of screening mammography among NESB women. Implications The implications of the findings from this review for population‐based screening programs in Victoria are outlined below.  Demographic factors Where monitoring data for population‐based screening programs include reliable demographic data, they can be used to identify, in broad terms, groups who are non‐
adherent to screening guidelines, thereby identifying where efforts to increase screening participation should be focused. However, monitoring data offer little information about underlying reasons for non‐adherence. The factors underlying some participation differences require further exploration. For example, evidence suggests that lower screening rates among low socioeconomic status groups may be related to lack of income or money worries as well as a focus on present and competing priorities. The review also indicates that identified groups (e.g. men, older, rural, ethnic, Indigenous, recent immigrant) are likely to be heterogeneous and that within‐group differences are likely to affect screening uptake. These differences need to be considered when designing strategies to improve screening uptake.  Cognitive factors A major finding of this review of factors influencing cancer screening participation is the importance of the way in which information is presented to target populations. There is evidence that literacy is likely to be low among some non‐adherent groups. Literacy is related to, but not entirely synonymous with, HL. Even those with good literacy skills may have reduced HL levels because of the specialised nature of health‐related information. There is evidence that individuals with reduced HL are less likely to seek out and/or engage with health‐related literature and information. In such instances, improving cancer‐related knowledge about screening, cancer and risk as a way of increasing screening participation rates will prove challenging. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 81  Psychosocial factors It is likely that addressing issues of fear/worry/anxiety that reduce people’s generalised cancer worries as well as fear of the screening process itself would have a positive impact on screening uptake. However, strategies for achieving this will need to take account of cognitive and cultural factors that may impact on these feelings. Evidence of a relationship between coping style and screening uptake was not strong and the implication for population‐based screening programs is unclear. Similarly, there is limited evidence in relation to social networks. If family and friends are able to influence the uptake of screening, they can presumably do so for better or for worse. Programs would need to ensure that family and friends have ‘good’ experiences and positive attitudes if they are to exert a positive pressure on individuals who are reluctant to participate in screening programs.  Health and lifestyle factors Little evidence was found of a relationship between physical and mental health and screening uptake. The issue here is the lack of evidence in the database to underpin common sense notions. This lack of evidence does not mean that physical health and mental health are not important factors in the uptake of screening. Screening programs should be cognisant of the difficulties people with physical and mental issues face when undergoing screening and make efforts to accommodate them. Concordance between people who have been screened previously and those who will participate in future screening or take part in other screening programs is not definitive. However, the implication is that engaging people at the time of the first screen and ensuring that they have a positive experience is likely to enhance repeat screening within programs and screening across programs.  Cultural factors Some evidence suggests that recent immigrants are less likely to participate in screening mammography and Pap tests. This suggests that screening programs should consider how to engage this group in culturally appropriate ways. If this could be achieved, it may have positive implications for later screening and participation in other screening programs. Engagement would need to address the issues of fatalism, modesty/embarrassment/ shame and medical mistrust.  Health system factors Monitoring data highlight geographic differences in screening participation rates but the reasons underlying these differences have not been explored in detail. Clear associations were apparent between: o having a usual source of care and increased screening uptake o recent visits to an outpatient or primary care doctor and screening uptake o recommendations by a doctor and screening uptake o female health care providers and increased uptake o provision of interpreters at screening facilities and increased uptake. VOL 2– 82 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening These findings point to the importance of doctors and other health professionals being involved in the promotion of screening programs to their patients. Whilst it will not be practicable to increase the number of female doctors, the results imply that other female health practitioners could be used to improve screening rates. The evidence also implies that screening programs need to ensure that their facilities can respond to the needs of non‐English speaking participants. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 83 REFERENCES 1. Consedine NS, Magai C, Krivoshekova YS, Ryzewicz L, Neugut AI. Fear, anxiety, worry, and breast cancer screening behavior: A critical review. Cancer Epidemiology Biomarkers and Prevention. 2004;13(4):501‐510. 2. Vernon SW. Participation in colorectal cancer screening: a review. J Natl Cancer Inst 1997;89(19):1406‐1422. 3. Guessous I, Dash C, Lapin P, Doroshenk M, Smith RA, Klabunde CN. Colorectal cancer screening barriers and facilitators in older persons. Preventive Medicine. 2010/2// 2010;50(1‐2):3‐10. 4. McCaul KD, Branstetter AD, Glasgow RD, Schroeder DM. What is the relationship between breast cancer risk and mammography screening? A meta‐analytic review. Health Psychol. 1996;15:423‐
429. 5. Petersen GM. Barriers to preventive intervention. Gastroenterology Clinics of North America. Dec 2002;31(4):1061‐1068. 6. Jepson R, Clegg A, Forbes C, Lewis R, Sowden A, Kleijnen J. The determinants of screening uptake and interventions for increasing uptake: A systematic review. Health Technology Assessment. 2000;4(14). 7. Australian Institute of Health and Welfare. Breast cancer in Australia. Canberra: AIHW; 2009. Cancer series no. 50, Cat no. 46. 8. Cokkinides V, Bandi P, Seigel R, Ward EM, Thun MJ. Cancer Prevention & Early Detection Facts & Figures 2008. Atlanta, GA: American Cancer Society; 2007. 9. Australian Institute of Health and Welfare. Cervical screening in Australia 2006‐2007. Canberra: Australian Institute of Health and Welfare; 2009. Cancer series no. 47, Cat no. CAN 43. 10. Wells KJ, Roetzheim RG. Health Disparities in Receipt of Screening Mammography in Latinas: A Critical Review of Recent Literature. Cancer Control. 2007;14(4):369‐395. 11. Weller D, Moss S, Butler P, Campbell C, Coleman D, Melia J, Robertson R. English Pilot of Bowel Cancer Screening: an evaluation of the second round. Edinburgh: The Institute of Cancer Research, The University of Edinburgh; 2006. 12. Australian Institute of Health and Welfare, Australian Government Department of Health and Ageing. National Bowel Cancer Screening Program: annual monitoring report 2009. Canberra: AIHW; 2009. Cancer series 49, Cat no. 45. 13. Galdas PM, Cheater F, Marshall P. Men and health help‐seeking behaviour: literature review. J Adv Nurs. Mar 2005;49(6):616‐623. 14. Woods M. Men's use of general practitioner services. NSW Public Health Bulletin. 2001;12(12):334‐335. 15. Wright A. What puts men off visiting their GP. Paper presented at: 3rd National Men's Health Conference, 1999; Alice Springs. 16. Aoun S, Johnson L. What motivates rural men to improve their health status? Canberra: Rural Industries Research and Development Corporation; 2000. 00/157 2000. 17. Smith JA, Braunack‐Mayer A, Witters G. What do we know about men's help‐seeking and health service use? Medical Journal Australia. 2006;184(2):81‐83. VOL 2– 84 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 18. Aotearoa B. BreastScreen Aotearoa: Coverage for the 24 months to end April 2009. Aukland: BreastScreen Aotearoa; 2009. 19. Purc‐Stephenson RJ, Gorey KM. Lower adherence to screening mammography guidelines among ethnic minority women in America: A meta‐analytic review. Preventive Medicine. 2008;46(6):479‐
488. 20. Sadler GR, Fullerton JT. Effective strategies that enhance adherence to breast cancer screening guidelines. Breast Disease. 2001;13:3‐12. 21. National Bowel Cancer Screening Pilot Program's Multicultural Working Group. Addressing the needs of culturally and linguistically diverse communities in a national screening program for bowel cancer: Discussion Paper. Carlton, VIC: National Bowel Cancer Screening Pilot Program's Multicultural Working Group; 2004. 22. Cancer Council SA, SA Cervix Screening Program, BreastScreen SA. Small Grants Scheme ‐ Business Plan 2008/09. In: Victoria B, ed. Adelaide: Cancer Council SA, SA Cervix Screening program, BreastScreen SA; 2008. 23. Mullins R. Evaluation of the impact of PapScreen's campaign on Culturally and Linguistically Diverse (CALD) women. Melbourne, Victoria: The Cancer Council Victoria; 2006. 24. Akers AY, Newmann SJ, Smith JS. Factors Underlying Disparities in Cervical Cancer Incidence, Screening, and Treatment in the United States. Current Problems in Cancer. 2007;31(3):157‐181. 25. Fiscella K, Franks P, Gold MR, Clancy CM. Inequality in quality: Addressing socioeconomic, racial and ethnic disparities in health care. Journal of the American Medical Association. 2000;283(19):2579‐2584. 26. Schueler KM, Chu PW, Smith‐Bindman R. Factors associated with mammography utilization: A systematic quantitative review of the literature. Journal of Women's Health. 2008;17(9):1477‐
1498. 27. Ackerson K, Gretebeck K. Factors influencing cancer screening practices of underserved women. Journal of the American Academy of Nurse Practitioners. 2007;19(11):591‐601. 28. Australian Institute of Health and Welfare. Cervical screening in Australia 2005‐2006. Canberra: AIHW; 2008. Cancer series no. 41, Cat no. CAN 36. 29. Deshpande AD, Sanders Thompson VL, Vaughn KP, Kreuter MW. The use of sociocultural constructs in cancer screening research among African Americans. Cancer Control. 2009;16(3):256‐265. 30. Wolff M, Bates T, Beck B, Young S, Ahmed SM, Maurana C. Cancer prevention in underserved African American Communities: Barriers and effective strategies ‐ A review of the literature. Wisconsin Medical Journal. 2003;102(5):36‐40. 31. Newman C, Butow P, Knight R, McMillan K, Treloar C, Kippax S, Eades S. Cancer and Aboriginal people in Australia: A review of the literature. Critical Public Health. 2008;18(1):65‐75. 32. US Congress. National Literacy Act of 1991. USA; 1991. 33. Ishikawa H, Yano E. Patient health literacy and participation in the health‐care process. Health Expectations. 2008;11(2):113‐122. 34. Friedman DB, Hoffman‐Goetz L. Literacy and health literacy as defined in cancer education research: A systematic review. Health Education Journal. 2008;67(4):285‐304. 35. Katapodi MC, Lee KA, Facione NC, Dodd MJ. Predictors of perceived breast cancer risk and the relation between perceived risk and breast cancer screening: A meta‐analytic review. Preventive Medicine. 2004;38(4):388‐402. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 85 36. Han PKJ, Moser RP, Klein WMP. Perceived ambiguity about cancer prevention recommendations: Relationship to perceptions of cancer preventability, risk, and worry. Journal of Health Communication. 2006;11(SUPPL. 1):51‐69. 37. Federation of Ethnic Communities' Councils of Australia. Access and Equity Report: Feedback from FECCA's consultation regarding access and equity issues for people from culturally and linguistically diverse backgrounds, conducting during November and December 2006. Canberra: FECCA; 2006. 38. Organisation for Economic Co‐operation and Development. Literacy in the Information Age. Final Report of the International Adult Literacy Survey. Cedex, France: OECD; 2000. 39. Entwistle V. Offering choices: Editorial. Health Expectations. 2008;11(1):1‐2. 40. De Walt DA, Berkman ND, Sheridan S, Lohr KN, Pignone MP. Literacy and Health Outcomes. Journal of General Internal Medicine. 2004;19(12):1228‐1239. 41. Scott TL, Gazmararian JA, Williams MV, Baker DW. Health literacy and preventive health care use among Medicare enrollees in a managed care organization. Med Care. 2002;40:395‐404. 42. Basu P, Sarkar S, Mukherjee S, Ghoshal M, Mittal S, Biswas S, Mandal R, Sankaranarayanan R. Women's perceptions and social barriers determine compliance to cervical screening: Results from a population based study in India. Cancer Detection and Prevention. 2006;30(4):369‐374. 43. Giordano L, Webster P, Anthony C, Szarewski A, Davies P, Arbyn M, Segnan N, Austoker J. Improving the quality of communication in organised cervical cancer screening programmes. Patient Education and Counseling. 2008;72(1):130‐136. 44. Donelle L, Arocha JF, Hoffman‐Goetz L. Health literacy and numeracy: Key factors in cancer risk comprehension. Chronic Diseases in Canada. 2008;29(1):1‐8. 45. Guerra CE, Dominguez F, Shea JA. Literacy and knowledge, attitudes, and behavior about colorectal cancer screening. Journal of Health Communication. Oct‐Nov 2005;10(7):651‐663. 46. von Wagner C, Semmler C, Good A, Wardle J. Health literacy and self‐efficacy for participating in colorectal cancer screening: The role of information processing. Patient Education and Counseling. 2009;75(3):352‐357. 47. Shaw A, Ibrahim S, Reid F, Ussher M, Rowlands G. Patients' perspectives of the doctor‐patient relationship and information giving across a range of literacy levels. Patient Education and Counseling. 2009;75(1):114‐120. 48. Wolf MS, Davis TC, Osborn CY, Skripkauskas S, Bennett CL, Makoul G. Literacy, self‐efficacy, and HIV medication adherence. Patient Education and Counseling. 2007;65(2):253‐260. 49. Bandura A. Self‐efficacy: toward a unifying theory of behavioral change. Psychol Rev. Mar 1977;84(2):191‐215. 50. McQueen A, Vernon SW, Myers RE, Watts BG, Eun SL, Tilley BC. Correlates and predictors of colorectal cancer screening among male automotive workers. Cancer Epidemiology Biomarkers and Prevention. 2007;16(3):500‐509. 51. Peterson NB, Dwyer KA, Mulvaney SA, Dietrich MS, Rothman RL. The influence of health literacy on colorectal cancer screening knowledge, beliefs and behavior. Journal of the National Medical Association. 2007;99(10):1105‐1112. 52. Parsa P, Kandiah M, Abdul Rahman H, Zulkefli NM. Barriers for breast cancer screening among Asian women: a mini literature review. Asian Pacific journal of cancer prevention : APJCP. 2006;7(4):509‐514. VOL 2– 86 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 53. Austin LT, McNally M, Stewart DE. Breast and cervical cancer screening in Hispanic women: a literature review using the health belief model. Women's Health Issues. 2002;12(3):122‐128. 54. Vernon SW. Risk perception and risk communication for cancer screening behaviors: a review. J Natl Cancer Inst Monogr. 1999(25):101‐119. 55. Magai C, Consedine N, Neugut AI, Hershman DL. Common psychosocial factors underlying breast cancer screening and breast cancer treatment adherence: A conceptual review and synthesis. Journal of Women's Health. 2007;16(1):11‐23. 56. Weinberg DS, Miller S, Rodoletz M, Egleston B, Fleisher L, Buzaglo J, Keenan E, Marks J, Bieber E. Colorectal Cancer Knowledge is not Associated with Screening Compliance or Intention. Journal of Cancer Education. 2009;24(3):225‐232. 57. Han PKJ, Kobrin SC, Klein WMP, Davis WW, Stefanek M, Taplin SH. Perceived ambiguity about screening mammography recommendations: Association with future mammography uptake and perceptions. Cancer Epidemiology Biomarkers and Prevention. 2007;16(3):458‐466. 58. Mullins R. Evaluation of a strategy to increase cervical screening through the use of personalised letters. Melbourne, Victoria: The Cancer Council Victoria; 2005. 59. Bowen DJ, Helmes A, Powers D, Andersen MR, Burke W, McTiernan A, Durfy S. Predicting breast cancer screening intentions and behavior with emotion and cognition. Journal of Social and Clinical Psychology. 2003;22(2):213‐232. 60. Hay JL, Buckley TR, Ostroff JS. The role of cancer worry in cancer screening: A theoretical and empirical review of the literature. Psycho‐Oncology. 2005;14(7):517‐534. 61. Hosseini SM. Maternal Anxiety During Pregnancy: Its relations to birth outcomes and to offspring depression during late childhood and adolescence [Doctor of Philosophy]. Pittsburgh: Graduate School of Public Health, University of Pittsburgh; 2006. 62. Spielberger CD. The effects of anxiety on complex learning and academic achievement. In: Spielberger CD, Ed. Anxiety and behavior. New York: Academic Press; 1966. 63. Spielberger CD, Gorsuch RL, Lushene RE. Manual for the State‐Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press; 1970. 64. Kinney AY, Bloor LE, Martin C, Sandler RS. Social ties and colorectal cancer screening among blacks and whites in North Carolina. Cancer Epidemiology, Biomarkers & Prevention. 2005;14(1):182‐189. 65. Hay JL, McCaul KD, Magnan RE. Does worry about breast cancer predict screening behaviors? A meta‐analysis of the prospective evidence. Preventive Medicine. 2006;42(6):401‐408. 66. Vos MS, de Haes JCJM. Denial in cancer patients, an explorative review. Psycho‐Oncology. 2007;16:12‐25. 67. Kreitler SC, S., Kreitler H. The psychological profile of women attending breast‐screening tests. Soc Sci Med. 1990;31:1177‐1185. 68. Consedine NS, Magai C, Neugut AI. The contribution of emotional characteristics to breast cancer screening among women from six ethnic groups. Preventive Medicine. 2004;38(1):64‐77. 69. Tejeda S, Thompson B, Coronado GD, Martin DP. Barriers and facilitators related to mammography use among lower educated Mexican women in the USA. Social Science & Medicine. Mar 2009;68(5):832‐839. 70. Wu TY, Bancroft J, Guthrie B. An Integrative Review on Breast Cancer Screening Practice and Correlates Among Chinese, Korean, Filipino, and Asian Indian American Women. Health Care for Women International. 2005;26:225‐246. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 87 71. Blue Moon Research and Planning. BreastScreen Australia Evaluation; Participation Qualitative Study. Canberra: Department of Health and Ageing; 2008. Screening Monograph No. 3/2009. 72. Johnson K, Strong R, Hillier L, Pitts M. Screened Out!: Women with disabilities and cervical screening. Melbourne, Victoria: La Trobe University; 2002. 73. Stein K, Allen N. Cross sectional survey of cervical cancer screening in women with learning disabilities. British Journal of Learning Disability. 1999;23:56‐58. 74. Cambridgeshire County Council. Joint Investment Plan for Adults with Learning Disabilities; 2002. 75. Sutherland G. Preliminary Results ‐ Living Well Research Project. In: (2002) Jea, ed. Melbourne: Centre for Developmental Disability Health Victoria (CDDHV); 2002. 76. Bankhead CR, Brett J, Bukach C, Webster P, Stewart‐Brown S, Munafo M, Austoker J. The impact of screening on future health‐promoting behaviours and health beliefs: A systematic review. Health Technology Assessment. 2003;7(42). 77. Department of Health and Ageing. Bowel Cancer Knowledge, Perceptions and Screening Behaviours: Knowledge, Attitudes and Practices, Pre‐and Post‐Intervention Surveys (2002 and 2004). Canberra: Australian Government, Department of Health and Ageing; 2004. Screening Monography No. 4/2005. 78. Condron J, Remedios C, Fernbach M. Evaluation of the 2001 PapScreen Victoria multicultural media campaign. Melbourne, Victoria: Centre for Behavioural Research in Cancer; 2002. 79. Shahid S, Thompson SC. An overview of cancer and beliefs about the disease in Indigenous people of Australia, Canada, New Zealand and the US. Australian and New Zealand Journal of Public Health. Apr 2009;33(2):109‐118. 80. Negy C, Woods DJ. Mexican‐Americans' performance on the Psychological Screening Inventory as a function of acculturation level. Journal of Clinical Psychology. 1992;48(3):315‐319. 81. Straughan PT, Seow A. Fatalism reconceptualized: a concept to predict health screening behavior. Journal of Gender and Cultural Health. 1998;3(2):85‐100. 82. Powe BD, Finnie R. Cancer Fatalism: The State of the Science. Cancer Nursing. 2003;26(6):454‐
467. 83. Lende DH, Lachiondo A. Embodiment and breast cancer among African American Women. Qualitative Health Research. 2009;19(2):216‐228. 84. Hausmann‐Muela S, Ribera JM, Nyamongo I. Health‐seeking behaviour and the health system response. Basel, Switzerland: Swiss Tropical Institute; 2003. 85. Australian Institute of Health and Welfare. BreastScreen Australia monitoring report 2005‐2006. Canberra: AIHW; 2009. Cancer series no. 48, Cat no. CAN 44. 86. Condron J, Fernbach M. Evaluation of the 2001 PapScreen Victoria Neighbourhood House Grants Program. Melbourne: Centre for Behavioural Research in Cancer; 2001. 87. Wong BCY, Chan AO, Wong WM, Hui WM, Kung HF, Lam SK. Attitudes and knowledge of colorectal cancer and screening in Hong Kong: A population‐based study. Journal of Gastroenterology and Hepatology. Jan 2006;21(1):41‐46. 88. Ackerson K, Preston SD. A decision theory perspective on why women do or do not decide to have cancer screening: Systematic review. Journal of Advanced Nursing. 2009;65(6):1130‐1140. 89. Watts T, Merrell J, Murphy F, Williams A. Breast health information needs of women from minority ethnic groups. Journal of Advanced Nursing. 2004;47(5):526‐535. VOL 2– 88 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening 90. Weller DP, Patnick J, McIntosh HM, Dietrich AJ. Uptake in cancer screening programmes. The Lancet Oncology. 2009;10(7):693‐699. Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 89 APPENDIX: CANCER SCREENING LITERATURE REVIEWS Reference Type of cancer Number of studies included Publicatio
n years Countries in which studies located Types of studies 1.
Ackerson et al. 27
(2007) Cervical 17 1999–2005 USA All types of empirical studies with at least 10% of study participants African American and/or Hispanic women 2.
Ackerson et al. (2009)88 Breast, Cervical 19 1994–2008 English language, worldwide Qualitative analysis of quantitative studies related to decision, choice and framing of screening decisions 3.
Austin et al., 53
(2002) Breast, Cervical Unclear but <30 1966–2001 USA Studies that used the Health Belief Model to examine factors influencing screening in Hispanic women 4.
Bankhead et al. 76
(2003) Breast, Cervical Breast only: 71 Cervical only: 25 Breast & Cervical: 24 1988–2000 English language, worldwide All study designs relating to the impact of breast and cervical screening on future health‐promoting behaviours and beliefs 5.
Blue Moon Research and Planning (2008)71 Breast 76 2000–2007 English language, worldwide All study designs; included peer‐reviewed and grey literature in relation to perceptions, knowledge and beliefs about breast cancer screening 6.
Consedine et al. 1
(2004) Breast 22 1987–2003 Appears to be: English language, worldwide Studies examining the role of anxiety, fear, and worry as related to breast cancer screening behaviour 7.
Deshpande 29
(2009) Breast, Cervical, Colorectal 35 1990–2006 USA All types of empiric studies 8.
Guessous et al., 3
(2010) Colorectal 83 (59 (71%) reported on FOBT alone or in combination with other screening tests) 1995–2008 English language, worldwide Full‐length, peer‐reviewed studies relating to colorectal cancer screening in people aged 65 years and older 9.
Hay et al. (2005)60 Breast, Mammograms
Cervical Cervical 3 Colorectal Breast & Cervical Colorectal 10
1872–2003 English language, worldwide Studies with empirical findings about cancer worry as a correlate of cancer screening Breast 1872–2006 English language, Prospective studies involving a breast cancer 10. Hay et al. (2006)65 12 VOL 2– 90 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening Reference Type of cancer Number of studies included Publicatio
n years Countries in which studies located worldwide Types of studies screening measure and measures of worry or anxiety about cancer 11. Jepson et al. 6
(2000) Breast, Cervical, Colorectal Unclear but <65 relating to breast, cervical and colorectal cancer 1966–1985 English language, worldwide 82% located in USA or Canada Studies using some form of multivariate analysis. These included: randomised controlled trials (RCTs); quasi‐RCTs; cohort studies; case control studies with a prospective time barrier between collecting information and uptake of screening 12. Katapodi et al. 35
(2004) Breast Primary Studies:
Lit Reviews: 1 1985– approx. 2003 English language worldwide Studies which examined the relationship between breast cancer perceived risk and breast cancer screening 13. Magai et al. (2007)55 Breast 71 1980–2005 Unstated, written in English was the only criterion Studies of the psychosocial and cognitive variables associated with screening 14. McCaul et al. 4
(1996) Breast Unsighted but quoted in Katapodi et al. (2004)35 15. Petersen (2002) 5 Colorectal Not Stated Not Stated Not Stated Not Stated 16. Purc‐Stephenson, Breast and Gorey (2008)19 33 1975–2006 USA, Canada Empirical studies which included a white non‐
Hispanic comparison group focused on women aged 50 years and who did not have a history of breast cancer and which included findings with sufficient detail to calculate effect sizes 17. Powe and Finnie (2003)82 Breast, Cervical, Colorectal Breast 4 Cervical 1 Colorectal 3 1996–2003 USA Empirical studies on the phenomenon of fatalism and its association with cancer screening 18. Sadler and Fullerton 20
(2001) Breast Unclear but <80 Appears to be 1995–
2000 Unclear Unclear 19. Schueler et al. (2008)26 Breast 221 1988–2007 USA Studies were included if they reported original data on factors related to mammography use and provided quantitative results in a form from which contingency tables and ORs or a percentage Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOL 2– 91 Reference Type of cancer Number of studies included Publicatio
n years Countries in which studies located Types of studies could be derived 2
20. Vernon (1997) Colorectal Factors 36 Reasons 15 1973–1996 US, Canada, Europe, Other Countries (percentages not stated) All types of empiric studies addressing the: (i) correlates and predictors of adherence to FOBT; and (ii) reasons for non‐
adherence. Published between 21. Vernon (1999)54 Breast, Colorectal No studies were located for the correlates of perceived risk of cervical cancer Breast 9 Colorectal 2 1996–1998 English language, worldwide Empirical studies relating to risk perception and cancer screening 22. Watts et al. 89
(2004) Breast Unclear but authors indicated it was a ‘small number’ 1996–2002 English language, worldwide Predominantly North American Studies relating to the breast health information needs of women from minority ethnic groups 23. Wells and Roetzheim (2007)10 Breast 15 1997–2007 USA Empirical studies with population‐based samples of participants that used multiple logistic regression analysis that addressed the association between Hispanic ethnicity and receipt of screening mammography and which reported adjusted ORs 24. Weller et al., 90
(2009) Cancer Unclear Appears to be 2000–
2008 English language, worldwide Unclear 25. Wolff et al. (2003)30 Cancer 25 Appears to be 1991–
2000 USA Unclear 26. Wu et al. 70
(2005) Breast 23 1990–2003 USA Survey research (mailed and face‐to‐face) focusing on Chinese Korean, Filipino and Asian Indian American women VOL 2– 92 Identifying hard‐to‐reach groups: a review of the factors (including barriers) associated with cancer screening VOLUME 3 KNOWLEDGE TRANSLATION: A REVIEW OF STRATEGIES TO INCREASE PARTICIPATION IN CANCER SCREENING TABLE OF CONTENTS TABLE OF CONTENTS ...................................................................................................................................... 2 LIST OF TABLES ............................................................................................................................................... 4 LIST OF FIGURES ............................................................................................................................................. 6 1. INTRODUCTION ................................................................................................................................... 5 2. OVERVIEW OF CANCER SCREENING: AUSTRALIA AND OVERSEAS ...................................................... 7 2.1 Breast cancer screening: mammography ................................................................................... 7 2.2 Cervical cancer screening (Pap tests) ....................................................................................... 10 2.3 Colorectal cancer screening...................................................................................................... 13 3. METHODOLOGY ................................................................................................................................ 15 3.1 Locating the literature .............................................................................................................. 15 3.2 Inclusion criteria ....................................................................................................................... 15 3.3 Strength of the evidence .......................................................................................................... 17 3.4 Classifying the strategies .......................................................................................................... 18 4. STRATEGIES TARGETING SCREENING POPULATIONS........................................................................ 20 4.1 Invitations and reminders......................................................................................................... 20 4.2 Message framing ...................................................................................................................... 25 4.3 Education .................................................................................................................................. 28 4.4 Counselling................................................................................................................................ 32 4.5 Coaching ................................................................................................................................... 34 4.6 Community interventions ......................................................................................................... 36 4.7 Mass media campaigns............................................................................................................. 38 4.8 Worksite interventions ............................................................................................................. 41 4.9 Financial incentives/disincentives ............................................................................................ 42 4.10 Procedures................................................................................................................................ 43 4.11 Multi‐component interventions ............................................................................................... 46 5 STRATEGIES TARGETING PROVIDERS ................................................................................................ 50 5.1. Provider only strategies ............................................................................................................ 50 5.2 Provider and screening population strategies combined ......................................................... 51 5.3 Hard‐to‐reach groups ............................................................................................................... 51 5.4 Literature reviews and grey literature...................................................................................... 52 6. ENGAGING THE HARD‐TO‐REACH ..................................................................................................... 54 6.1 Engaging men ........................................................................................................................... 54 6.2 Engaging hard‐to‐reach women in cancer screening ............................................................... 57 6.2 Parallel evidence from other areas........................................................................................... 65 7. DISCUSSION AND CONCLUSIONS ...................................................................................................... 68 7.1 Overview of primary studies..................................................................................................... 68 7.2 Summary of results for cancer screening strategies................................................................. 71 7.3 Summary of results for hard‐to‐reach groups .......................................................................... 73 7.4 Conclusions ............................................................................................................................... 75 REFERENCES ................................................................................................................................................. 79 APPENDIX: ANNOTATED BIBLIOGRAPHY (154 PRIMARY STUDIES) .............................................................. 92 VOL 3– 2 Knowledge translation: a review of strategies to increase participation in cancer screening LIST OF TABLES Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 Table 14 Table 15 Table 16 Table 17 Table 18 Table 19 Table 20 Table 21 Table 22 Table 23 Table 24 Table 25 Table 26 Table 27 Table 28 Table 29 Table 30 Table 31 Table 32 Table 33 Table 34 Table 35 Table 36 Table 37 Table 38 Table 39 Table 40 Table 41 Table 42 Table 43 Table 44 Table 45 Organisation, participation, and estimated coverage in 18 European breast screening programs......................................................................................................................................9 Participation of women aged 20–69 in the National Cervical Screening Program....................10 Details of population‐based, national cervical screening programs in the European Union ....12 Participation rates in Phase 1 of the National Bowel Cancer Screening Project (30 June 2008)..........................................................................................................................................13 Inventory of colorectal cancer screening activities (May 2008) ................................................14 Number of citations scrutinised for possible inclusion in the review........................................15 Overview of the primary studies included in the analysis .........................................................16 Levels of evidence......................................................................................................................17 Classification of strategies targeted at screening populations..................................................18 Classification of strategies targeted at healthcare professionals ..............................................19 Peer‐reviewed primary studies: invitations and reminders (n = 32 studies).............................20 Impact of different forms of invitations on mammography participation at 12‐month follow‐up (odds ratios and 95% CI)............................................................................................23 Impact of reminder letters on participation by women due for Pap tests ................................23 Overview of the impact of invitations and reminders on screening uptake .............................24 Peer‐reviewed primary studies: message framing (n = 27 studies)...........................................25 Impact of message framing in screening uptake .......................................................................28 Peer‐reviewed primary studies: education (n = 24 studies) ......................................................28 Overview of the impact of educational strategies on screening uptake ...................................31 Peer‐reviewed primary studies: counselling (n = 17 studies) ....................................................32 Overview of the impact of counselling on screening uptake.....................................................33 Peer‐reviewed primary studies: coaching (n = 15 studies)........................................................34 Overview of the impact of coaching on screening uptake ........................................................36 Peer‐reviewed primary studies: community interventions (n = 13 studies) .............................36 Overview of the impact of community interventions on screening uptake ..............................38 Peer‐reviewed primary studies: mass media campaigns (n = 4 studies) ...................................38 Peer‐reviewed primary studies: worksite interventions (n = 4 studies)....................................41 Peer‐reviewed primary studies: financial incentives (n = 3 studies) .........................................42 Peer‐reviewed primary studies: procedures (n = 15 studies)....................................................43 Overview of the impact of procedures on screening uptake ....................................................46 Peer‐reviewed primary studies: multi‐component (n = 21 studies) ..........................................46 Overview of the impact of multi‐component interventions......................................................49 Peer‐reviewed primary studies: provider only strategies (n = 9 studies) ..................................50 Peer‐reviewed primary studies: provider/screening‐populationstrategies (n = 4 studies).......51 Impact of provider strategies on the uptake of mammograms and Pap tests ..........................52 Percentage of women who reported having a Pap test in the previous 12 months .................58 Changes in screening rates pre‐ and post‐intervention for Native Hawaiian women...............59 Target groups in ethno‐specific studies (27 studies, 32 intervention arms) .............................60 Impact of interventions targeting low income and non‐urban women ....................................61 Under‐served populations ......................................................................................................... 62 Impact of single strategies on the uptake of mammograms by under‐users ............................63 Overview of studies targeting indigenous women ....................................................................63 Effectiveness of intervention strategies to improve the uptake of immunisation (29 studies) ......................................................................................................................................65 Average campaign effect size by campaign topic ......................................................................66 Impact of health prevention nurse on recording of screening activities...................................67 Number of studies investigating each of the strategies (154 studies, 188 intervention arms)..........................................................................................................................................69 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 3 Table 46 Table 47 Table 48 Table 49 Table 50 Summary of the findings for the strategies ...............................................................................72 Overview of findings for strategies targeting hard‐to‐reach groups .........................................74 General population: assessment of the strategies ....................................................................76 Hard‐to‐reach groups (indigenous, ethnic, low income, non‐urban): assessment of the strategies ...................................................................................................................................77 Opportunities to combine interventions across cancer screening programs............................78 LIST OF FIGURES Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 BreastScreen two‐year coverage for women 50–69 (age‐standardised rate).............................8 Participation of women aged 20–69 in the National Cervical Screening Program by age group (1996–97 to 2004–05) .....................................................................................................11 Age‐standardised cervical screening rates 1996–97 to 2004–05 ..............................................40 Overview of effectiveness of strategies targeting specific ethnic groups .................................64 Overview of country from which primary intervention studies originated (154 studies, 183 intervention arms) .............................................................................................................. 69 Overview of type of cancer screening test targeted in primary intervention studies...............71 VOL 3– 4 Knowledge translation: a review of strategies to increase participation in cancer screening 1. INTRODUCTION Overview This is the third volume of a literature review prepared by the Centre for Health Policy, Programs and Economics at The University of Melbourne for the Victorian Cytology Service (VCS). The review was commissioned by the Victorian Department of Human Services (Department). It has included both international and Australian literature (peer‐reviewed and grey literature), and has focused on key areas that inform thinking, knowledge and approaches to engaging individuals, groups and communities to participate in organised programs, including population‐based screening programs. The review is presented as three stand‐alone documents. Each volume has a specific focus:  volume 1 provides an overview of the theoretical models underpinning thinking in relation to participation in screening programs  volume 2 focuses on the factors and barriers to participation in screening programs  volume 3 focuses on the impact of interventions designed to increase participation in organized screening/health programs. The first volume of the review concentrates on the cognitive domain. In Volume 2, a variety of factors were shown to be associated (either positively or negatively) with the uptake of screening. The factors were categorised under six domains (Table 1). Table 1 Domain Factors associated with the uptake of cancer screening Factors Demographic Age, Gender, Ethnicity, Socioeconomic status, Education Cognitive Literacy, Health literacy, Cancer‐related knowledge, Perceived risk, Perceived ambiguity Psychosocial Fear/Anxiety/Worry, Coping style (Emotion regulation), Social networks Health and lifestyle Physical/mental health, Drug use, Previous participation in screening Cultural Acculturation, Fatalism, Modesty/embarrassment/shame, Medical mistrust, Collectivism/communalism, Spirituality/religiosity Health System Availability, Accessibility, Affordability, Acceptability The aim of this final report is to present the VCS with a summary of the most up‐to‐date evidence in relation to encouraging participation in health programs, particularly among groups who have traditionally been defined as 'hard‐to‐reach'. To provide a context for the presentation of the evidence, Section 2 provides a very brief overview of cancer screening programs (rates and recruitment) in Australia and other countries. In Section 3, the methodology for locating and analysing the literature is outlined. The next two sections contain summaries of interventions relating to mammography, faecal occult blood testing (FOBT), and Pap tests aimed at the screening population (Section 4), providers, and both screening populations and providers combined (Section 5). In Section 6, evidence is presented Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 5 relating to engaging hard‐to‐reach groups including men, and parallel evidence from non‐cancer screening programs. The report concludes with a summary of the evidence and a discussion of the results (Section 7). VOL 3– 6 Knowledge translation: a review of strategies to increase participation in cancer screening 2. OVERVIEW OF CANCER SCREENING: AUSTRALIA AND OVERSEAS 2.1 Breast cancer screening (mammography)  Australia In Australia, all women who are Australian citizens and those with permanent residency status are eligible for breast screening through BreastScreen Australia. One of the objectives of BreastScreen Australia is the achievement of a 70% participation rate in the national program for women aged 50–69 years old.1 As shown in Figure 1, none of the jurisdictions achieved this target in the period 2002 to 2007. Also, except for the Northern Territory, 2‐year coverage rates during the period have converged; rates in those jurisdictions with relatively higher 2002–03 rates have declined (e.g. South Australia) and those with a relatively lower 2002–03 rates have increased (e.g. New South Wales). The rates in Victoria have declined from 58% to 56%. Other mammography for screening and diagnostic purposes is conducted outside the BreastScreen program, and is funded through the Medicare Benefits Schedule (MBS). Therefore, the coverage rates shown in Figure 1 are an underestimate of all mammography screening. However, a report prepared by the Department of Health and Ageing (2009), which examined the impact of MBS rebates for mammography, concluded that the addition of non‐diagnostic mammography funded through MBS would only minimally increase screening coverage rates.2  Europe and the UK Estimated coverage rates are shown in Table 2 for 18 screening programs. Most of the programs are organised so that the screening program is distinct from the delivery of general medical care (i.e. centrally organised). Screening tests are provided in distinct, fixed or mobile specialised units. Opportunistic screening is relatively rare, and the programs are administered at either a national or local level. In longstanding programs (e.g. in the Nordic countries, the Netherlands and the UK) the participation rates (as a percentage of invitees) are over 70%.3 Whatever the type of organisation, direct mail invitations are generally sent to women offering them free screening. Publicity campaigns through media advertising, pamphlets, newspapers, radio and television and referrals from general practitioners are frequently used with the mailings. A date and time for the appointment are always offered to the women in centralised programs. In Greece, women living in small villages in the mountains are collected by minibus and arrive at the screening unit or mobile screening bus. Once all have been screened, they return to their village.3 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 7 Figure 1 BreastScreen 2‐year coverage for women 50–69 (age‐standardised rate) 65
60
55
50
45
40
2002–2003
2003–2004
2004–2005
2005–2006
2006–2007
NSW
50.9
49.7
51.5
55.4
56.0
VIC
58.0
58.6
57.7
56.8
55.9
Qld
58.4
57.9
58.2
58.0
56.6
WA
55.7
56.6
55.8
57.3
57.5
SA
63.5
62.8
61.8
58.9
55.8
TAS
59.0
57.3
57.5
57.1
54.2
ACT
56.6
51.9
55.4
58.3
57.0
NT
43.9
42.2
41.4
41.0
40.2
Source: AIHW (2008)1 Notes: Residents of Victorian postcodes allocated to the Albury/Wodonga catchment (NSW jurisdiction) are included in Victoria's population estimate, accounting for the slight decrease in participation rates compared to those published by BreastScreen Victoria VOL 3– 8 Knowledge translation: a review of strategies to increase participation in cancer screening Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 9 1999 Switzerland Belgium Austria 50 50 50 2 2 2 70 69 69 2 2 1.5/2 2 3 2 2 2 2 2 2 2 2 2 2 Screening interval No No No Invitation specifies date and time Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes No No Notes: (1) More than one age means different regions screen different ages (2) C: Centralised; D: Decentralised; D/C: Decentralised organisation / centralised coordination (3) No screening in Austria in 2005 Source: Schopper and de Wolf (2007)3 1999 1990 Finland Netherlands Sweden Denmark UK Norway Ireland Iceland Spain Luxembourg Portugal Italy Germany Greece France Age range of screened population (1) Lower Upper 50 59/69 50 74 40/50 69/74 50 69 50 70 50 69 50 65 40 69 45/50 64/69 50 69 45 69 50 69 50 69 40/50 64 50 74 D D C C C C C C C C C C D C D D/C C D Organisation of screening (2) Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– No No Yes No No Yes No No No No No No No No No if in program Yes No if insured No Co‐
payment Organisation, participation, and estimated coverage in 18 European breast screening programs Year program started 1986 1989 1986 1991 1988 1996 1989 1987 1990 1992 1990 1985 1999 1989 1989 Table 2 9
7.3 24.6 89.3 % target population invited 2006 (a) 99.5 97.1 100.0 11.0 91.0 95.0 54.0 100.0 98.7 96.3 70.0 48.7 62.0 4.2 100.0 (3) 43.6 26.1 Participation rate as % of 2005 invitees (b) 87.4 80.8 80.0 76.6 76.4 76.2 75.2 67.0 64.6 63.8 61.8 57.0 54.4 52.5 49.3 2.2 Cervical cancer screening (Pap tests)  Australia Over the past 30 years, large increases have occurred in the number of women being screened. There was a 40% increase in screening rates from 1971 (129 smears per 1000 women aged 20–69) to 1986 (181 smears per 1000 women aged 20–69),4 and a 44 % increase from 1983 to 1992 (210/1000 'woman years' to 303/1000 woman).5 The increase in participation rates in cervical screening until the early 1990s was due to the increase in opportunistic testing, but since then various programs and initiatives have become progressively more structured, with recruitment programs to encourage high levels of participation by Australian women.6 The National Cervical Screening Program commenced in 1991. The program targets women aged 20–69 years. The proportion of women aged 20–69 years participating in cervical screening in the 2‐year period 2005–06 was 60.6%. Participation rates in Victoria were 66.7% in 1996–97 and 64% in 2005–06 and these rates compare favourably with those in other States and Territories (Table 3).7 According to Mullins (2005), PapScreen Victoria (PSV) has used mass media campaigns, community‐based activities and personalised letters to encourage women to have, and continue to have, Pap tests.8 Table 3 1996–1997 1997–1998 1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 Participation of women aged 20–69 in the National Cervical Screening Program NT 61.4 60.2 62.6 63.6 61.7 61.4 60.2 59.7 58.5 54.5 NSW 55.0 58.1 59.4 58.9 59.1 59.4 58.8 58.4 58.2 58.7 QLD 58.1 57.0 56.3 57.2 57.7 58.4 57.7 WA 64.9 64.2 63.9 61.7 61.4 60.7 60.6 59.8 60.5 60.5 TAS 63.3 65.1 64.5 63.9 65.2 65.0 63.1 62.0 62.9 62.4 ACT 63.5 65.4 65.7 63.0 62.8 63.3 62.7 62.7 65.5 63.8 SA 62.9 65.2 66.0 64.7 64.9 65.2 65.1 65.1 64.1 64.5 VIC 66.7 66.9 67.7 65.2 64.6 64.9 64.2 64.8 65.4 64.3 Source: AIHW (2008)7 and Excel spreadsheet csa04‐05‐xdi1.xls available at www.aihw.gov.au/publications/can/csa04‐05/csa04‐05‐xdi1.xls Despite the geographic differences in screening rates, the crude rate and age‐
standardised rate (ASR) of screening have been fairly stable in Australia since 1996–97 at approximately 61%.9 However, there has been a decrease in the crude rate for women under 40 during this time and an increase in the rates for women over 40 (Figure 2). In 2001, Condron and Fernbach wrote that research had shown that the Pap test rates were higher among younger women, but more recent studies have found that 2‐year participation rates are highest among women aged 50–54 years (72.5% 50–54 years, 70.7% 35–39 years, and 50.6% for 20–24 years). According to the authors, these data 'would suggest that initiatives to promote high participation among older women have been successful.’6 As pointed out by the Steering Committee, this could also suggest that women have stayed with the program and continued to be screened as they aged. The data in Figure 2 would suggest that both explanations may be true for the cohort of VOL 3– 10 Knowledge translation: a review of strategies to increase participation in cancer screening women aged 25–39 in 1996–97 (screening rates 64.5%, 66.9%, 66.4%) who would be aged 45–59 in 2006–07 (screening rates 67.5%, 64.7%, 69.1%). Screening rates have increased in the 60+ age groups between 1996–97 and 2006–07 but once women reach 60 years, there is a major decline in screening rates (Figure 2). Figure 2 Participation of women aged 20–69 in the National Cervical Screening Program by age group (1996–97 to 2004–05) 70%
60%
50%
40%
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
1996-1997
50.0%
64.5%
66.9%
66.4%
64.0%
64.3%
64.0%
62.7%
50.9%
41.2%
2006-2007
48.0%
57.5%
62.4%
64.3%
64.5%
67.5%
65.7%
69.1%
59.4%
51.7%
Source: AIHW (2009)9  Europe and the UK The Australian 3‐year participation rate for the period 2004–06 was 73.1% for women. This is reported as comparable to the 3‐year rates of 73% for New Zealand for 2003, 69.4% for England for 2007, 63.6% for Wales for 2007, and to the previously reported average for the European Union countries of 75%. The 5‐year participation rate for 2002–
06 for women aged 20–69 was 85.9%. This is higher than the rate of 79.2% reported for England for 2007, 74.6% reported for Wales for 2007, and 77% for the Netherlands, but lower than a previously estimated 5‐year participation rate of 90% for Finland.7 The majority of countries invite women to the screening program, sending a personal invitation (in some cases also stating the place and the time of the examination). Others leave the invitation responsibility to the GP or gynaecologist (Luxembourg, Portugal), or personally invite the women to make an appointment with the GP (Denmark, the Netherlands), or leave the women themselves free to arrange an examination (Germany).10 A number of countries indicate that increasing recruitment is important (Table 4). Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 11 VOL 3– 12 Knowledge translation: a review of strategies to increase participation in cancer screening 23 20/25 Sweden UK(a) 60/64 60 60 60/65 65 To 3 (25–49) 3 (23–50) 5 5 3 (23–50) Screening interval VOL 3– 12 Yes Yes Yes Yes No Invitations sent 78 NA 65 70 NA Screened/ year: % of invitees Notes Coverage rates were maintained at over 80% for a number of years but have dropped to under 80%. Younger women are now less likely to attend for screening than their counterparts in previous generations. The basic issue of recruiting women to be screened remains of the highest importance. Samples collected by midwives at antenatal centres supervised by gynaecologists. Invitations are sent to women who have not had a test registered in the database for 3 (or 5) years. In some counties a specific time and place for the test is contained in the invitation. If a woman prefers to go to a doctor, that test is registered and the next invitation is postponed. Coverage is higher in rural areas where organised screening dominates and personal invitations to screening are the rule. Samples taken by GPs or trained GP assistants. All women receive an invitation to have a test free of charge. Samples taken by trained nurses or midwives in local healthcare centres. Improving attendance and compliance into the organised program, especially among women 25–39 years of age, is a key to further prevent cervical cancer. Interventions to achieve better attendance are needed. Reminding (by letter or phone) women initially non‐responding is an option. Finland is also piloting the self‐sampling test instead of re‐invitation. In 2007, the National Board of Health issued a new program for cervical cancer screening, recommending screening every third year for women aged 23 to 50 years, and every fifth year for others if the last two smears within the last 10 years were negative. Knowledge translation: a review of strategies to increase participation in cancer screening Notes: NA: Not available (a) Excluding Scotland Sources: Ahti et al. (2009a)11,, Ahti et al. (2009b)12 30 25/30 Finland Netherlands 23 From Eligible age (years) Details of population‐based, national cervical screening programs in the European Union Denmark Table 4 2.3 Colorectal cancer screening  Australia The first phase of a National Bowel Cancer Screening Program (2006–08) using immunochemical FOBT (iFOBT), followed by colonoscopy if indicated, targeted people turning 55 or 65 years of age between 1 May 2006 and 30 June 2008, as well as all those included in an earlier pilot program. iFOBT kits were sent by direct mail; individuals with a positive test are advised to see their doctor. A national register has been established to issue invitations to screening and to follow people up to the point of diagnosis. This phase‐in of the program was evaluated in 2008 and if successful on clinical grounds, will be expanded to a wider age cohort. As shown in Table 5, the national participation rates for those who were not included in the earlier pilot program was 37% for 55 year olds and 44% for 65 year olds. When the results for the invitees to the earlier pilot program were examined, the Phase 1 participation rates were: i) 80% for those who had participated in the pilot; and ii) 21% for those who had been invited but did not participate. Table 5 Participation rates in Phase 1 of the National Bowel Cancer Screening Project (30 June 2008) NSW VIC QLD WA SA NBCSP 55 years 34.5 38.5 35.5 40.0 39.5 65 years 42.0 44.5 44.0 48.3 48.8 Age‐standardised rate 55–74 for invitees in the pilot program Invitees who participated 83.8 71.6 82.8 Invitees who did not 23.5 20.5 22.8 participate TAS ACT 40.3 46.8 NT 40.5 48.1 AUST 27.0 30.5 36.8 44.3 79.6 20.9 Source: Australian Institute of Health and Welfare & Australian Government Department of Health and Ageing (2008)13 Notes: VIC pilot site = Melbourne; QLD pilot site = Mackay; SA pilot site = Adelaide; NBCSP: National Bowel Cancer Screening Project  Overseas Screening for colorectal cancer is much less frequent than screening for breast and cervical cancer.14 This made it difficult to find screening rates for countries as many do not yet have national programs. However, as shown in Table 6, FOBT screening rates are high in Finland (but the participation rates for women are higher than the participation rates for men), moderate in Denmark and Italy, and low in Japan and Korea. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 13 Table 6 Inventory of colorectal cancer screening activities (May 2008) Country Finland Denmark Sweden United Kingdom Italy Japan Korea (Republic) Participation rate Overall 71% Women 78% Men 63% Description of CRC screening activity The program uses individual level randomisation, where half of the target population is offered screening now, and half will stay as controls for the first 6 years. Presently, the program targets women and men aged 60–69. It uses guaiac‐based FOBT, a sample collection procedure performed on 3 consecutive days by people at home. People in the target group receive and return the tests by mail. Approx 50% A pilot project of FOBT targeting people aged 50–74 years was undertaken in two counties. Its purpose was to test participation, and the stage distribution of detected cases. Participation was about 50%. The potential benefits of CRC screening have been reviewed by a new committee. Not available Population‐based service screening for CRC was introduced in Stockholm county in January 2008. The program uses FOBT and targets the age group 60–69, men and women, at 2‐year intervals. The entire population is included (no exclusion). A centrally organised call and recall system is used. Test kits are sent to participants at their home address. A reminder is sent after 4 weeks if non‐ attendance. Pilot program A program of biennial guaiac‐based FOBT in individuals aged 50–69 results only began in England in 2006 and for those 50–74 in 2007 in Scotland. The whole of England will be screening by the end of 2009 and expansion up to age 75 years will commence from 2010. Wales will 56.8% (1) begin screening some people aged 50–74 in 2009. In 2006, 69 screening programs were active in 11/21 Italian regions) 45% (target population of 6,240,000 among 14,000,000 Italians 50–69 years old) using biennial iFOBT. A national CRC screening program using annual iFOBT was 18% implemented in 1992. It is conducted through local governments and targets individuals aged 40+ who are national insurance holders. National guidelines were revised in 2005 to recommend FOBT or iFOBT, although iFOBT is stated as being the preferred test. Increasing the participation rate is an important goal. CRC screening has been part of a national cancer screening program 16% since 2004. Annual FOBT (iFOBT or Hemoquant) is offered to individuals aged 50+. The screening rate was about 16% in 2006. The screening program is free of charge. Sources: Downloaded on 8 October 2009 from the International Cancer Screening Network website at http://appliedresearch.cancer.gov/icsn/colorectal/screening.html Notes: (1) UK Colorectal Cancer Screening Pilot Project Group (2004)15 VOL 3– 14 Knowledge translation: a review of strategies to increase participation in cancer screening 3. METHODOLOGY 3.1 Locating the literature A range of databases and indexes available through the University of Melbourne library were investigated to determine search terms for the literature review that would give a broad, multi‐
disciplinary coverage, coupled with the most efficient search and retrieval capabilities. The five databases included in the search were: 1. Medline: medical research literature 2. ISI Web of Science/ISI Web of Knowledge: Science Citation Expanded®, Social Sciences Citation Index®, and Arts & Humanities Citation Index™ 3. CINAHL: nursing, biomedicine, health sciences, consumer health and seventeen allied health disciplines in journals, books and book chapters, dissertations, selected conference proceedings, standards of practice, pamphlets, educational software packages and audiovisual material 4. SCOPUS: medical and social science literature 1. Cochrane Library: systematic reviews of health care treatments and interventions. Broad searches were used to locate the screening literature, and included combinations of the following: (breast cancer OR cervical cancer OR colorectal cancer OR bowel cancer OR colon cancer); screen*; (participation OR engagement). Articles were also located by a scrutiny of the reference lists, and of grey literature databases, from the Steering Committee and through the environmental survey being undertaken by the VCS. Table 7 Number of citations scrutinised for possible inclusion in the review Peer‐reviewed literature Grey literature Strategy Database search Scrutiny of reference lists and Steering Committee Database searches and Steering Committee Total Number 1491 165 % 86 10 68 4 1724 100 3.2 Inclusion criteria The inclusion criteria for articles included in the knowledge transfer review are outlined below.  Types of screening Screening has been defined as ‘the systematic application of a test or inquiry, to identify individuals at sufficient risk of a specific disorder to warrant further investigation or direct preventive action, among persons who have not sought medical attention, on account of symptoms of that disorder.’16 By detecting disease before symptoms occur, screening programs can be an effective method of reducing morbidity and mortality. Screening can be carried out with the aim of primary prevention (e.g. screening for risk factors such as hypertension), secondary prevention (e.g. cancer screening) or tertiary prevention (e.g. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 15 screening for sensorineural deafness). We have concentrated on studies that relate to screening as secondary prevention, and excluded studies that related to diagnostic screening. However, the evidence from other programs included primary prevention as well as secondary prevention studies. The types of cancer screening included are: mammography, Pap testing and FOBT.  Endpoints We have used the model of reciprocal determinism to guide the types of studies that would be included. In this model, behaviour, personal factors (such as knowledge, attitudes, and beliefs) and environmental factors (such as availability, cost, and cultural safety) interact with one another. The complexity of this model indicates there is no guarantee that changing one of the factors (e.g. knowledge or attitudes) will result in changes in behaviour. Therefore, only studies which included some identification of the impact of the intervention on screening uptake have been included.  Other No restrictions were placed on the study designs to be included. However, only primary studies (peer‐reviewed articles) and systematic reviews published since 1998 have been included (the systematic reviews did include primary studies published prior to 1999). The citations had to be written in English and did not include editorials or letters. This resulted in a total of 154 primary studies and 43 literature reviews, and meta‐analyses for inclusion in the review of cancer screening interventions. Just over two‐thirds of the primary studies related to interventions aimed at increasing screening in the USA and fourteen (9%) related to interventions to increase screening in Australia. Twenty‐five (15%) of the primary studies related to multiple screening tests (e.g. for breast, cervical and colorectal cancer) and just under half (42%) related to screening for breast cancer alone (Table 7). Table 8 Overview of the primary studies included in the analysis Type of cancer Breast cancer only Colorectal only Cervical only Breast and cervical Breast cervical colorectal Total Studies n 64 39 28 14 9 154 % 42 25 18 9 6 100 Country USA Australia UK(1) Italy Taiwan Sweden Spain Netherlands India Finland Canada Israel Other(2) Studies n 104 14 5 4 2 2 2 2 2 2 2 2 11 154 % 68 9 3 3 1 1 1 1 1 1 1 1 7 100 Notes: (1) Includes Scotland and Wales (2) One study each from: Belgium, France, Germany, Greece, Hong Kong, Ireland, Mexico, New Zealand, Puerto Rico, South Africa, Thailand; n: Number of studies VOL 3– 16 Knowledge translation: a review of strategies to increase participation in cancer screening 3.3 Strength of the evidence All primary screening studies were graded according to their study design and the corresponding NHMRC level of evidence (Table 9). Table 9 Levels of evidence Level I II Intervention Systematic review of level II studies Randomised controlled trials III‐1 Pseudo‐randomised controlled trial III‐2 A comparative study with concurrent controls III‐3 A comparative study without concurrent controls IV Case series Designs Includes cluster randomisation and studies where the randomisation is not adequately described Allocation to groups occurs according to some algorithm but is not randomised Non‐randomised experimental trial, cohort study, case‐control study, interrupted time series with a control group Historical control study, two or more single‐arm studies, interrupted time series without a parallel control group Post‐test only or pre‐post test outcomes Source: National Health and Medical Research Council (2008)17 Statistical significance was classified according to the results presented in the studies and were grouped as outlined below.  Statistically significant results: The results presented in the studies included significance tests that clearly indicated the intervention had a statistically significant impact (p < 0.05) on screening rates compared to the control or comparison, or the 95% Confidence Intervals for the odds and relative risk ratios indicated that the results were significant.  No statistically significant results: The results presented in the studies included significance tests that indicated the intervention had not had a statistically significant impact (p ≥ 0.05) on screening rates compared to the control or comparison, or the 95% Confidence Intervals for the odds and relative risk ratios indicated that the results were not statistically significant.  Results unclear: The results presented in the studies did not include significance tests, or there was no comparison group, or the results were presented in a way that made it difficult to be sure of statistical significance of the impact on screening rates. This classification means that some studies that were just outside of statistical significance but which demonstrated a positive impact on screening have been classified as 'no statistically significant results'. Even if these studies had been included in another category of 'just outside statistical significance', it would be necessary to make a judgement about the quantitative definition of this category (e.g. p ≥ 0.050 to p < 0.100). To overcome this shortcoming, an annotated bibliography has been included in the Appendix that includes the results for each of the 154 primary studies. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 17 3.4 Classifying the strategies A broad classification for the strategies used to increase uptake was developed based on the work of Jepson et al. (2000).18 They are first classified according to the health system target (screening population, healthcare provider), and then further classified into the strategies used to influence screening behaviour (Table 10). Table 10 Classification of strategies targeted at screening populations Strategy Invitations and reminders Message framing Education Counselling Coaching Community interventions Worksite interventions Procedures Economic incentives Mass media campaigns Multi‐component interventions Definition Invitations and/or reminders sent to people to undergo screening. Includes letters with fixed or open appointments, telephone calls, verbal recommendations, prompts and follow‐up letters. Messages about screening (either verbal or written) that have been variously framed to try to increase screening. These strategies include various forms of endorsement, tailoring messages to the personal factors of individuals, different presentations of risk and loss/benefit. Educational interventions aiming to increase knowledge of the screening program or the disease being screened for. This strategy does not contain a counselling component. They include printed educational materials, audio‐
visual materials, group and individual teaching, and home visits. Counselling, either face‐to‐face or on the telephone. Must involve a discussion of barriers to screening as well as an educational component. There were two major forms of this strategy: i) Lay health workers – Interventions where community members (often from a particular ethnic group) are trained and educated, in order to provide assistance and support to people as a means of encouraging screening. ii) Patient navigation – involves having someone trained to guide patients through the healthcare system to receive appropriate services. For example, patient education, assistance with scheduling appointments, providing appointment reminders, transportation, follow‐up with patient after procedure to determine the need for further action. Interventions aimed at whole communities. They usually involve multiple strategies such as mass media campaigns and education, and involve some form of community participation. They also tend to be long‐term interventions lasting for more than a year. These are generally multi‐strategy interventions that are conducted in the workplace. Interventions to increase screening uptake by making the screening procedure easier or more acceptable to individuals undergoing screening. Includes different screening tests for the same disease, varying diets, or the length of time that the screening test takes, and opportunistic testing. These include the removal of financial barriers or the provision of economic incentives. Includes reduced or free screening tests, subsidisation of out‐of‐
pocket costs, such as transport and postage or the provision of ‘rewards’ for completion of a screening test. These are campaigns conducted through the media (e.g. television, radio, newspapers and other print media). Interventions that include a combination of strategies. For example, invitations and reminders + education. Source: Based on Jepson et al. (2000)18 VOL 3– 18 Knowledge translation: a review of strategies to increase participation in cancer screening Table 11 Classification of strategies targeted at healthcare professionals Strategy Reminders/prompts Education Audit and feedback Office systems Definition Reminders to physicians to prompt or encourage individuals to undergo screening. Include chart reminders, forms, computer generated reminders and lists of overdue patients. Includes seminars, workshops and meetings. Analysis of healthcare providers' services and provision of feedback. Assistance in design and implementation of office routines and tools. Source: Based on Jepson et al. (2000)18 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 19 4. STRATEGIES TARGETING SCREENING POPULATIONS The analysis of the evidence for each strategy consists of a synthesis of the evidence from:  primary studies in the peer‐reviewed literature;  literature reviews in peer‐reviewed journals and from the grey literature; and  other studies contained in the grey literature. The presentation of the analysis for each strategy finishes with an overall summary about the likely impact of the strategy, given the evidence presented. 4.1 Invitations and reminders Peer‐reviewed primary studies There were 32 peer‐reviewed studies relating to strategies involving invitations and reminders (Table 12). Approximately two‐thirds (63%) indicated that sending invitations and/or reminders had statistically significant positive results compared to the comparison groups (usual care or an active intervention). The studies that did not have statistically significant results included four Level II studies and three Level III studies. The unclear results included two Level III studies and three Level IV studies. Table 12 Peer‐reviewed primary studies: invitations and reminders (n = 32 studies) Country Australia (5) Spain (2) Italy (2) UK (1) USA (17) 1 each from: Belgium, Canada, Ireland, Sweden, Thailand Type of cancer Breast only (15) Cervical only (5) Colorectal only (10) Breast & cervical (2) Hard‐to‐reach Low income (4) Non‐urban (3) Ethno‐specific (1) Level of evidence II (22) III‐1 (2) III‐2 (3) III‐3 (2) IV (3) Impact Stat sig results (20) No stat sig results (7) Results unclear (5) Mailed invitations and reminders  Screening mammograms (8 studies) In two the results indicated a statistically significant positive result in favour of the intervention19, 20 and in one the invitation had a positive impact but the reminder did not.21 Four studies did not report a statistically significant positive impact on uptake (three Level II and one Level III‐1)22‐25 and in one the results were unclear (Level IV).26  Pap tests (4 studies) Three studies reported statistically significant positive impacts on the uptake of Pap tests.27‐29 In one Level II study there was no statistically significant improvement.30 VOL 3– 20 Knowledge translation: a review of strategies to increase participation in cancer screening  Combined tests: mammograms and Pap tests (2 studies) In one Level III‐2 study the results were unclear31 and in one the intervention had a statistically significant positive impact on the uptake of Pap tests but not mammograms.32  Faecal occult blood testing (7 studies) Two studies reported statistically significant positive impacts on uptake in favour of the intervention.33, 34 In three studies the results were unclear.35‐37 In one which tested both an invitation and a reminder, the invitation increased uptake, but the reminder did not increase uptake beyond the initial reminder.38 In the remaining study, an advance notification had a statistically significant positive impact on the uptake of FOBT.39 Telephone invitations and reminders  Screening mammograms (4 studies) Four Level II studies reported statistically significant positive impact on the uptake of mammograms.22, 40‐42  Pap tests (2 studies) Both were Level II studies and both reported statistically significant positive impacts on the uptake of Pap tests.28, 29  Faecal occult blood testing (1 study) One Level II study reported a statistically significant positive impact on the uptake of FOBT43 and the other did not.34 Face‐to‐face invitations and reminders  Screening mammograms (2 studies) Two Level II studies reported statistically significant positive effects on the uptake of mammograms.44, 45  Pap tests (1 study) One Level II study did not report a statistically significant impact on the uptake of Pap tests.46  Faecal occult blood testing (1 study) One Level II study reported statistically significant positive effects on the uptake of FOBT.47 Multi‐method invitations and reminders  Screening mammograms (1 study) A Level III study used an initial mailed invitation, followed by a telephone reminder and then a face‐to‐face reminder, and reported a statistically significant positive impact on the uptake on mammograms.48  Pap tests (1 study) A Level III study used an initial mailed invitation followed by either a mailed or telephone reminder, and reported a statistically significant positive impact on the uptake of Pap tests.49 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 21 Hard‐to‐reach groups Included in the above results are eight studies targeting hard‐to‐reach groups.  Ethno‐specific (1 study) 1 This Australian Level II study targeted Vietnamese women aged 18–67 years. The mailed intervention had no impact on the uptake of Pap tests.30  Non‐urban (3 studies) Two studies aimed to increase the uptake of FOBT; in one Level III study the results were unclear35 and in the other Level II study the invitations increased uptake, but the reminder did not.38 The third study reported a statistically significant positive impact on the uptake of mammograms using face‐to‐face reminders.45  Low income (4 studies) Three Level II studies aimed to increase the uptake of mammograms in low‐income women, using mail invitations and reminders. One reported a statistically significant positive impact on uptake20 and the other two did not.23, 25 The fourth study aimed to increase the uptake of mammography and Pap tests, and reported a statistically significant positive impact for the uptake of Pap tests but not for mammograms.32 Literature reviews and grey literature A number of meta‐analyses undertaken before 1999 indicated that reminders increase the likelihood of women undergoing screening for Pap tests and mammograms.50 Since 1998 there have been reviews and meta‐analyses confirming that invitations increase the uptake of Pap tests and mammograms.18, 50‐59 In 2008, Baron et al. wrote: 'There is strong evidence that client reminders increase breast and cervical cancer screening by mammography and Pap test, respectively. These findings should apply across a range of settings and populations. There is sufficient evidence that client reminders increase colorectal cancer screening by guaiac‐based FOBT.'60
Invitation/reminder method The impact is not uniform across the different types of invitations and reminders. Denhaerynck et al. (2003) undertook a meta‐analysis of direct‐contact invitation studies (i.e. invitations by telephone or face‐to‐face contact) and found that telephone contacts (18 studies) had a statistically significant positive impact on attendance for mammography screening (RR = 1.24, 95% CI 1.11–1.39) but the impact of face‐to‐face contact (7 studies) did not reach statistical significance (RR = 1.14, 95% CI 0.95–1.36).61 In a Cochrane review published in 2009, Bonfill et al. computed the impact of different types of invitations compared with controls. The largest impact was for an invitation letter plus a follow‐
up telephone call, but even a simple mailed letter increased the likelihood of women having a mammogram (Table 13).55 1
This study was published in 1998 but has been included because it was the only study targeting a specific ethnic group with mailed invitations. VOL 3– 22 Knowledge translation: a review of strategies to increase participation in cancer screening Table 13 Impact of different forms of invitations on mammography participation at 12 months follow‐up (odds ratios and 95% CI) Letters of invitation Phone calls Invitation letter plus phone call Home visits Number of studies 5 2 3 2 Odds ratio 1.66 1.94 2.53 1.06 95% CI 1.43–1.92 1.70–2.23 2.02–3.18 0.80–1.40 Source: Bonfill et al. (2009)55 Nature of the screening test The nature of the test may have an important bearing on the impact of invitations and reminders. For example, Jepson et al. (2000) found invitation letters were effective in encouraging women to attend for Pap tests but less effective for mammograms.18 Although Bonfill et al. (2009) found that phone calls had the largest impact in relation to mammography, Forbes et al. (2007) reported that it was unclear whether or not telephone calls were more effective than invitation letters in relation to Pap tests.59 Finally, there were few reviews of the evidence of the impact of reminder letters on FOBT screening. One meta‐analysis by Stone et al. (2002) found that invitations and reminders increased screening for colorectal cancer (OR = 2.75, 95% CI 1.90–3.97).53 Target group characteristics Impacts in relation to invitations and reminders may also depend on the characteristics of the women being targeted. For example, Tseng, Cox and Plane (2001) found that reminder letters for women who were due for a Pap test were not effective for women of low socioeconomic status (Table 13).50 Denhaerynck et al. (2003) found that the impact of direct contact invitations (telephone and personal combined) were different for different groups of women (e.g. under‐
screened, non‐responders to previous invitations, general population), but these differences did not reach statistical significance.61 Table 14 Impact of reminder letters on participation by women due for Pap tests Number of studies Low socioeconomic status Mixed socioeconomic status Odds ratio 95% CI 2 1.16 (0.99–1.35) 1.17 (0.99–1.37) 8 2.02 (1.79–2.28) 2.00 (1.74–2.31) 50
Source: Tseng, Cox and Plan (2001) Fixed vs open appointments In 1999, Sin and St Leger concluded that the inclusion of appointments in invitation letters increased participation in mammography.57 However, Jepson et al. (2000) reached the following conclusions:  Pap tests – two randomised controlled trials (RCTs) found a fixed appointment more effective than an open appointment, and one reported no difference  screening mammography – two RCTs and one controlled trial found a fixed appointment more effective than an open appointment, and that the inclusion of an appointment time appears to be more effective than an open appointment Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 23  FOBT – one RCT found a letter with a fixed appointment more effective than a letter inviting people to make an appointment.18 Anderson (2004) concluded in relation to Pap tests, that including appointments in letters appeared to be a 'promising' strategy, and that 'two studies reviewed suggest that multiple reminders may be a beneficial strategy.'62 Summary The peer‐reviewed literature included studies from a range of countries and for a range of interventions. It is not dominated by studies from the USA or by studies relating to mammography. In total there were 36 intervention arms, and in just under two‐thirds there was a statistically significant positive impact on screening rates. Different methods of issuing the invitations and reminders were included in the studies (e.g. mail, telephone, face‐to‐face, multi‐method). Just under half the mailed invitations and reminders reported a statistically significant positive impact, and nearly all the telephone invitations and reminders had a positive impact (Table 14). Table 15 Overview of the impact of invitations and reminders on screening uptake Type Mail Telephone Face‐to‐face Multi‐method Increased uptake n % 11 48 7 88 3 75 1 100 22 61 Impact unclear n % 5 22 5 15 No impact n % 7 30 1 13 1 25 9 25 Total n 23 8 4 1 36 % 100 100 100 100 100 There is little reporting of the impact of invitations and reminders on specific ethnic groups, and in the one study that was included, the intervention did not have an impact. Seven studies targeted other hard‐to‐reach groups – three non‐urban and four low‐income. Results were mixed: three reported positive impacts on screening; three did not; and in one the results were unclear. The only study testing face‐to‐face invitations and reminders reported an increase in uptake due to the intervention. The reviews and grey literature emphasise that the impact on uptake of invitations and reminders vary by the method of administration, the nature of the test, and the characteristics of the target group. But the overall conclusion appears to be that invitations and reminders do increase uptake, at least among mid to high socioeconomic groups. Telephone calls and letters plus telephone calls appear to have a bigger impact than mailed invitations and reminders. However, the evidence for face‐to‐face is less clear. Fixed appointments in mailed invitations appear to be more effective than open appointments in mammograms and Pap tests. VOL 3– 24 Knowledge translation: a review of strategies to increase participation in cancer screening 4.2 Message framing Peer‐reviewed primary studies There were 27 peer‐reviewed studies relating to strategies involving some form of message framing (Table 16). Over half of the studies (56%) indicated that framing the message had statistically significant positive results compared to the comparison groups (usual care or an active intervention). The nine studies (33%) that did not have statistically significant results included eight Level II studies and one Level III study. The two unclear results included one Level II and one Level IV study. Table 16 Peer‐reviewed primary studies: message framing (n = 27 studies) Country Type of cancer Hard‐to‐reach Level of evidence Impact Australia (2) UK (2) Israel (1) Netherlands (1) Spain (1) USA (20) Breast only (16) Cervical only (3) Colorectal only (6) Breast & cervical (1) Breast, cervical & colorectal (1) Low income (5) Non‐urban (2) Ethno‐specific (3) II (22) III‐1 (1) III‐2 (2) IV (2) Stat sig results (15) No stat sig results (9) Results unclear (2) Varies by type of framing (1) Tailoring These strategies tailored the message to individual characteristics or barriers.  Screening mammograms (11 studies) Three studies reported statistically significant positive impacts on uptake.63‐65 Three studies did not report statistically significant improvements in uptake.23, 66, 67 In one study the results were unclear.68  Pap tests or faecal occult blood testing (2 studies) One study reported a statistically significant increase in the uptake of Pap tests69 and another reported a statistically significant increase in the uptake of FOBT.70  Multi‐test (2 studies) One study targeting both mammograms and Pap tests reported a statistically significant positive impact on uptake.71 Another study targeting mammograms, Pap tests and FOBT did not report any statistically significant increases in uptake.72 Endorsement (9 studies) Five studies reported on the impact of endorsement by a doctor on the uptake of mammograms. Four reported statistically significant positive impacts on uptake21, 44, 73, 74 and one did not.75 Risk (4 studies) All four studies related to the uptake of FOBT. Two studies reported statistically significant impacts on uptake76, 77 and two studies did not.39, 78 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 25 Loss/Gain (3 studies) This type of framing relates to how the reasons for undertaking screening are presented. They can be in a positive frame emphasising the benefits that arise from screening or in a negative frame emphasising what women stand to lose if they are not screened regularly. All three studies were concerned with the uptake of mammograms. One study reported statistically significant positive impacts on uptake79 and two did not.80, 81 Other framing (3 studies) These studies involved framing in relation to cost, fear, empowerment and severity of CRC. The study that investigated framing in terms of the severity of CRC reported a statistically significant increase in uptake of FOBT.76 In one study the framing did not increase the uptake of mammograms81 and in the other, the impact on uptake of mammograms was unclear.82 Hard‐to‐reach groups Included in the above results were thirteen studies targeting hard‐to‐reach groups.  Ethno‐specific (3 studies) Two studies targeted African American women and aimed to increase the uptake of mammograms. One involving tailoring reported a statistically significant positive impact63 and one involving tailoring did not.75 The third study involving tailoring targeted Asian American women and reported a statistically significant positive impact on uptake.64  Non‐urban (2 studies) Both studies aimed to increase the uptake of Pap tests in the UK. One involving tailoring reported a statistically significant positive impact on uptake69 and one involving endorsement by a celebrity did not.83  Low income (8 studies) All 8 studies aimed to increase the uptake of mammograms. Two studies reported a statistically significant positive impact on uptake. One involved ‘gain’ framing79 and the other, endorsement by a doctor.73 Three studies did not report a positive impact on uptake. One involved tailoring,23 and two involved framing in terms of loss/gain and empowerment.80, 81 Literature reviews and grey literature  GP endorsement In the reviews, the information relating to GP endorsement was inconclusive. Sin and St Leger (1999) concluded that endorsement of the letter of invitation by a GP did not boost uptake in women invited for a mammogram. Despite their analysis, the authors concluded that 'it is plausible that GP endorsement of the invitation …improves the acceptance of screening' and screening programs could consider incorporating this strategy.57 Jepson et al. (2000) reviewed the effectiveness of sending letters from different sources and concluded it was not possible to detect which approach was more effective.18 Based on one study with statistically significant results, Black et al. (2002) concluded that personalised invitation letters from family physicians were more effective in promoting participation in Pap testing than a letter of invitation from a clinic involving a female nurse VOL 3– 26 Knowledge translation: a review of strategies to increase participation in cancer screening practitioner (not necessarily known to the client), and controls.52, 54 In relation to mammography screening, Denhaerynck et al. (2003) concluded that: 'Because published studies comparing letters signed by the screening centre coordinator and the physician/general practitioner all argue in favour of a physician/general‐practitioner‐signed letter, this is probably true for any invitation letter designed to increase participation.'61  Risk In a review of RCTs on the effects of communicating risks in screening programs, Edwards et al. (2003) indicated that the majority of trials came from the area of mammography, and that caution should be used in generalising to other topic areas. Overall, they found that individual risk communication (however presented) was associated with an increased uptake of screening (OR 1.5 95% CI 1.11–2.03). However, there was some evidence that more detailed communication of individual risk may lead to smaller increases in screening. For example: o individualised risk presented as numerical risk calculations OR = 1.22 (95% CI 0.56–
2.68) o risk estimates presented as categories (e.g. high, medium, low) OR = 1.43 (95% CI 1.07–1.88) o list of personal risk factors OR = 1.7 (95% CI 1.17–2.48). However, the differences between the categories did not reach statistical significance. Edwards et al. (2003) also pointed out that the effects seemed greater among consumers or patients deemed to be at higher risk than average, and this may mean that risk status may be an important effect modifier for individualised risk communication.84 
Loss/gain (positive/negative or benefits/harms) In a review of RCTs (RCTs), cluster RCTs or quasi RCTs, Forbes et al. (2007) were not able to find any appropriate studies relating to positive/negative message framing in relation to Pap testing.59 The only other review that included conclusions about positive/negative message framing was that undertaken by Anderson (2004) in relation to breast screening. Anderson (2004) concluded: 'From the limited evidence available, the framing of breast screening information in terms of benefits or harms has little effect on screening behaviour …taking the evidence as a whole there appears to be support for the efficacy of loss‐framed reminder letters compared to gain‐framed letters …tailored letters do not appear to be a strategy worth pursuing in the breast or cervical screening area.'62 In an unpublished Australian study, Mullins (2005) found that there were no statistically significant differences in uptake of Pap tests between women who had received a gain framed letter and those who received a loss framed letter at 3 months (Chi sq = 0.465, df = 1, p = 0.5) and 6 months (Chi sq = 0.177, df = 1, p = 0.674) follow‐up.8 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 27 Summary In the peer‐reviewed literature, over half the studies related to mammograms (58%) and nearly three‐quarters (74%) originated in the USA. Just under half of the 33 interventions arms had a statistically significant impact (Table 17). Endorsement by a doctor seems to be particularly effective, with 6 of the 7 studies that used this strategy reporting increased uptake. The evidence for other message framing strategies is mixed. Thirteen studies targeted hard‐to‐reach groups. Five (38%) reported increased uptake in either American ethnic groups, or non‐urban or low‐income groups, eight (62%) did not. Table 17 Type Endorsement GP Other Tailoring Risk Loss/Gain Other Impact of message framing in screening uptake Increased uptake n 6 6 2 1 1 16 % 86 55 50 20 25 48 Impact on uptake unclear n % 1 9 0 0 2 50 3 9 Did not increase uptake n % 1 14 2 100 4 36 2 50 4 80 1 25 14 42 Total n % 7 2 11 4 5 4 33 100 100 100 100 100 100 100 The conclusions from the reviews are based on a narrower range of interventions and a fewer number of studies (mostly confined to Level II studies). The conclusions from this literature are that there is little evidence for GP endorsement. Framing in terms of risk appears to increase uptake but there is little evidence of benefit in terms of loss/gain framing. 4.3 Education Peer‐reviewed primary studies There were 24 peer‐reviewed studies relating to educational strategies (Table 18). One‐third indicated that educational strategies had statistically significant positive results compared to the comparison groups. Three of the studies included strategies targeted at multiple screenings (one was not statistically significant). The studies that did not have statistically significant results included 6 Level II studies, five Level III studies and one Level IV. The unclear results included one Level III‐3 study, two Level III‐2 studies, and one Level IV study. Eleven of the interventions were targeted at hard‐to‐reach groups. Table 18 Country Australia (1) South Africa (1) Germany (1) Hong Kong (1) USA (20) Peer‐reviewed primary studies: education (n = 24 studies) Type of cancer Breast only (8) Cervical only (3) Colorectal only (10) Breast & cervical (1) Breast, cervical & colorectal (2) Hard‐to‐reach Ethno‐specific (8) Non‐urban (1) Low literacy (1) Squatters (1) Level of evidence II (13) III‐1 (1) III‐2 (7) III‐3 (1) IV (2) Impact Stat sig results (8) No stat sig results (12) Results unclear (4) VOL 3– 28 Knowledge translation: a review of strategies to increase participation in cancer screening  Print material (5 studies) One study reported a statistically significant positive impact on the uptake of mammograms.85 Four studies did not report a positive impact – one mammogram,86 one Pap test87 and two FOBT.43, 88  Interactive (group or one‐on‐one) (7 studies) One interactive group study reported statistically significant positive impacts on the uptake of mammograms.89 In two studies the results were unclear (one mammogram,90 one Pap test91) and in two studies targeting multiple tests there were no statistically significant positive impacts reported.92, 93  Videos (6 studies) Three studies (two FOBT94, 95 and one mammogram96) reported statistically significant positive impacts on uptake. Three studies (two FOBT97, 98 and one mammogram99) did not.  Multi‐strategy (7 Studies) Three studies (one mammogram,96 one Pap test100 and one FOBT95) reported statistically significant positive impacts on screening. In three studies there were no positive impacts reported – two FOBT94, 101 and one mammogram.102 In one study the impact on FOBT screening was unclear.103  Other strategies (2 studies) One study which investigated the impact of an interactive computer program on FOBT did not report a positive impact.104 In another study, the education intervention extended over 5 weeks but the strategies used could not be discerned, and there was no positive impact reported.105 Hard‐to‐reach groups 
Ethno‐specific (10 studies) Five educational interventions reported statistically significant positive impacts on the uptake of screening, two on the uptake of mammograms using videos 94, 96 and one using one‐on‐one education.106 One targeted Pap testing using a multi‐strategy intervention,100 and two targeted FOBT using videos94 and a multi‐strategy intervention.94 In two studies the results were unclear. One study involved face‐to‐face group education and aimed to increase the uptake of mammograms90 and the other was a multi‐strategy intervention which aimed to increase the uptake of mammograms, Pap tests and FOBT.105 In three studies no statistically significant impacts were reported. One aimed to increase the uptake of mammograms using a multi‐strategy intervention.96 Another aimed to increase the uptake of FOBT using a video,97 and the other aimed to increase the uptake of mammograms and Pap tests using group education.93 
Other: rural, squatters, low literacy (3 studies) The strategies used in these studies involved a video,98 a multi‐strategy intervention involving a video plus print material101 and print material alone.87 No statistically significant results were reported for any of the studies. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 29 Literature reviews and grey literature Two meta‐analyses presented an assessment of the impact of 'education'. Yabroff et al. (1999) divided the educational interventions into two groups – generic education and theory‐based education. Overall, the generic interventions did not lead to a statistically significant increase in mammography screening (1.1%, 95% CI 2.4%–4.6%). The four theory‐based educational interventions with usual‐care controls were effective in increasing the rate of mammography utilisation by 23.6% (95% CI 16.4%–30.1%).51 In 2002 Stone et al. found that education increased participation in mammography (OR = 1.31, 95% CI 1.12–1.52), cervical cytology (OR = 1.53, 95% CI 1.30–1.81) but did not increase participation in FOBT (OR = 1.38 95% CI 0.84–
2.25).53  Interactive education Yabroff et al. (1999) combined the results from four interactive educational interventions (i.e. by telephone or in person) and found that theory‐based interactive interventions were effective in increasing mammography utilisation (7.9%, 95 % CI 2.3–13.5).51 In 2000 Jepson et al. concluded that, overall, educational materials were found to be of limited effectiveness, apart from home visits for which there was some evidence of effectiveness.18 Baron et al. (2008) divided face‐to‐face education into group education and one‐on‐one education. They concluded that there was insufficient evidence to determine the effectiveness of group education on increasing mammography, Pap tests and FOBT. This was due to 'inconclusive' findings for mammography, few studies with inconsistent findings for Pap tests and inconsistent findings in the only study for FOBT.60 However, they concluded that there was 'strong evidence' that one‐on‐one education increases participation in mammography screening and Pap testing, and that these findings should apply across a broad range of settings and populations when 'interventions are adapted to the target populations and delivery contexts.’ There was insufficient evidence to determine effectiveness for FOBT because there were too few studies with methodological limitations.60  Printed material/ videos/computer programs There were a series of reviews that looked at the effect of these forms of educational interventions on participation in mammography screening. In 1999 a review undertaken by Sin and St Leger found that non‐discriminant leaflet drops did not lead to an increase in subsequent participation in screening mammography when distance and car ownerships were taken into account.57 In the same year a review published by Yabroff et al. found that theoretically based educational interventions delivered by letter or videotape were not more effective in increasing participation in mammography screening compared to active control groups (OR = 0.4 95% CI 5.4–6.2).51 Ten years later in a review published in 2009, Bonfill et al. concluded that a mailed psycho‐educational booklet did increase the likelihood of participation in mammogram screening (OR = 2.81, 95% CI 1.96–4.02).55 These conclusions were based on a small number of studies – two for the Sin and St Leger (1999) review, four for the Yabroff et al. (1999) review and one study published in 1992 for the 2009 review. VOL 3– 30 Knowledge translation: a review of strategies to increase participation in cancer screening There were two reviews looking at the impact on Pap testing. Yabroff et al. (2003) compared theory‐based educational interventions with active controls (3 studies) and found that this form of educational intervention did not increase participation compared to active controls.107 Ellis et al. (2003) also found that the use of printed material (three reviews) did not increase participation (RR = 1.03, 95% CI 0.75–1.43).58 In a review that examined the evidence in relation to mammography (17 studies), Pap testing (12 studies) and FOBT (7 studies), Baron et al. (2008) classified letters, brochures, pamphlets, flyers and newsletters that provide educational or motivational information to promote cancer screening as small media and concluded that: 'There is strong evidence that small media increase breast, cervical and colorectal cancer screening by mammography, Pap test, and guaiac‐based FOBT, respectively. These findings should apply across a range of settings and populations.'60 Summary In the peer‐reviewed literature 83% of the studies originated in the USA. The studies were evenly spread across the types of cancer screening (mammograms, Pap tests and FOBT). Just over one‐third of the interventions resulted in increased uptake. Reaching conclusions based on the peer‐reviewed studies is hampered by the small number of studies in most of the strategies. However, it appears that interventions using print material generally do not report statistically significant positive impacts on uptake (Table 19). Thirteen interventions targeted hard‐to‐reach groups and five (38%) increased uptake in ethnic groups in the USA. For two interventions the results were unclear, and the remaining 6 interventions did not increase uptake in the targeted hard‐to‐reach group. Table 19 Overview of the impact of educational strategies on screening uptake Type Videos Print material Interactive education One‐on‐one Group Multi‐strategy Other Increased uptake n % 3 50 1 20 2 100 1 20 3 43 10 37 Impact unclear n % 0 0 0 2 40 1 14 1 50 4 15 No impact n % 3 50 4 80 0 2 40 3 43 1 50 13 48 Total n 6 5 0 2 5 7 2 27 % 100 100 100 100 100 100 100 The conclusions reached in the literature reviews tend to be based on a small number of high quality studies, and the conclusions are contradictory. The most positive conclusions for the use of printed material included 'letters' in this category. If these letters are, in fact, invitations then this overlaps with the invitations and reminders strategy and may account for the positive conclusion. Overall, the evidence for education appears to be mixed, and no clear trends can be discerned. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 31 4.4 Counselling
Peer‐reviewed primary studies There were 17 peer‐reviewed studies relating to strategies involving counselling (Table 20). The majority (94%) involved breast cancer screening. Twelve (71%) showed clear evidence of statistically significant, positive results in favour of the counselling strategy. The studies that did not have statistically significant results included one Level I, and two Level II studies. One study had a significant result, but the counselling had a negative result on screening. Five of the interventions were targeted at hard‐to‐reach groups. Table 20 Country Israel (1) USA (16) Peer‐reviewed primary studies: counselling (n = 17 studies) Type of cancer Breast only (16) Breast, cervical & colorectal (1) Hard‐to‐reach Ethno‐specific (4) Low income (1) Level of evidence I (1)* II (14) III‐2 (2) Impact Stat sig results (12) No stat sig results (3) Varies (Yes – face‐to‐face, No – telephone) (1) Stat sig (counselling had negative result on screening) (1) Notes: * A study which combined data from individual projects being run under the same funding source. Screening mammograms  Telephone (13 studies) Twelve interventions had a statistically significant positive impact on the uptake of mammograms.40, 42, 63, 74, 75, 85, 86, 108‐112 One study reported a statistically significant decrease in uptake due to telephone counselling.113 Four studies did not report any statistically significant improvements in uptake.74, 85, 108, 109  Face‐to‐face (6 studies) All six studies reported statistically significant positive impacts on the uptake of mammograms.74, 75, 110, 114‐116 Multi‐tests: mammograms, Pap tests and faecal occult blood testing  Telephone (1 study) One study reported statistically significant positive impacts on the uptake of screening.117 Hard‐to‐reach groups  Ethno‐specific (2 studies, 4 intervention arms) These two studies targeted African American women and both reported statistically significant positive impacts for both telephone and face‐to‐face counselling.75, 115  Low income (1 study) This was the study that aimed to increase screening rates for mammograms, Pap tests and FOBT, and reported statistically significant positive impacts.117 VOL 3– 32 Knowledge translation: a review of strategies to increase participation in cancer screening Literature reviews and grey literature Jepson et al. (2000) evaluated 4 telephone counselling studies (3 for mammography and one for Pap testing) and concluded that telephone counselling appeared to be effective; three of the four had statistically significant positive results in favour of telephone counselling. There were only two face‐to‐face counselling interventions (one for mammography and one for Pap testing) and neither of these had statistically significant results. Jepson et al. (2000) concluded that face‐
to‐face counselling was of 'limited' effectiveness.18 Ellis et al. (2005) reviewed strategies to increase participation in mammography and concluded that telephone counselling could not be separated out from other interventions.58 Forbes et al. (2007) defined counselling as either face‐to‐face or on the telephone to encourage participation in Pap testing. The strategy had to involve a discussion of barriers to screening as well as an educational component. Only two studies were located; one involving telephone counselling and the other face‐to‐face by a GP. The authors combined the two studies and computed a relative risk score, which indicated a statistically significant positive effect for counselling (RR = 1.23, 95% CI 1.07–1.41). However, the intervention involving telephone counselling also involved provider prompts and was, therefore, a multi‐target strategy rather than a strategy targeted solely at the screening population. The authors also noted that there were quality issues with both studies.59 Taking these factors into account, it can be concluded that this review does not provide strong evidence for the effectiveness of counselling, despite the statistically significant RR score. Summary The peer‐reviewed literature was dominated by studies originating in the USA and on mammography (i.e. the target group was women). The majority of the interventions involved telephone counselling. All the face‐to‐face interventions, and just over 70% of the telephone interventions, reported increased uptake (Table 21). Six of the interventions targeted hard‐to‐
reach groups and five reported increases in uptake. One reported a decrease. Table 21 Telephone Face‐to‐face Total Overview of the impact of counselling on screening uptake Increased uptake n 10 6 16 % 71 100 80 No increase on uptake n 3 3 % 21 15 Decreased uptake n % 1 7% 1 5% Total n 14 6 20 % 100 100 100 In contrast to the peer‐reviewed literature, the literature reviews provide limited evidence for the effectiveness of counselling. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 33 4.5 Coaching Peer‐reviewed primary studies There were 15 peer‐reviewed studies relating to strategies involving some form of coaching (Table 22). Just under two‐thirds (60%) indicated that the coaching had statistically significant positive results compared to comparison groups. Twelve (80%) of the interventions were in the USA. The studies with no statistically significant results included two Level II studies and one Level III study. The unclear results included two Level III studies and one Level IV study. Table 22 Country Canada (1) Greece (1) Puerto Rico (1) USA (12) Peer‐reviewed primary studies: coaching (n = 15 studies) Type of cancer Breast only (5) Cervical only (6) Colorectal only (1) Breast & cervical (2) Breast, cervical & colorectal (1) Hard‐to‐teach Ethno‐specific (12) Non‐urban (1) Level of evidence II (8) III‐2 (3) III‐3 (2) IV (2) Impact Stat sig results (9) No stat sig results (3) Results unclear (3)  Screening mammograms (5 studies) Three studies reported statistically significant positive impacts on uptake. 118‐120 One Level II study did not report a statistically significant positive impact121 and in one Level III study the results were unclear.122  Pap tests (6 studies) Four Level II studies reported statistically significant positive impacts on uptake.100, 123‐125 In one Level III126 and one Level IV study127 the results were unclear.  Faecal occult blood testing (1 study) This Level II study did not report a statistically significant positive impact on the uptake of FOBT.128  Multi‐tests (3 studies) One Level II study which aimed to increase the uptake of mammograms and Pap tests reported statistically significant positive effects on uptake72 and one Level III study did not.129 One Level II study which targeted mammograms, Pap tests and FOBT reported statistically significant positive effects on uptake.130 Hard‐to‐reach groups  Ethno‐specific (12 studies) Two‐thirds of the studies (eight) reported statistically significant positive impacts on screening. Three related to mammograms,118‐120 four to Pap tests,100, 123‐125 and one to multiple tests (mammograms, Pap tests and FOBT).130 The effective studies targeted Native, African, Vietnamese, Chinese and Cambodian Americans, Latinos and low‐income ethnic minorities in the USA. Three studies did not report statistically significant positive impacts. One related to mammograms,121 one to mammograms and Pap tests combined,129 and one to FOBT.128 VOL 3– 34 Knowledge translation: a review of strategies to increase participation in cancer screening These studies targeted African Americans and Hispanics, and a low‐income ethnically diverse group in the USA. In one study targeting Native Canadians, aimed at increasing uptake in Pap tests, the results were unclear.126  Low income (4 studies) All four studies targeted low‐income ethnic groups. One study that aimed to increase uptake in multiple tests reported a statistically significant positive impact130 while the other three did not.121, 128, 129  Non‐urban (3 studies) Two of the three studies targeted non‐urban ethnic groups. One study, which aimed to increase the uptake of mammograms in non‐urban African Americans, reported statistically significant positive results.119 The other two studies aimed to increase the uptake of Pap tests, and the impact in both studies was unclear.126, 127 Literature reviews and grey literature  Screening mammography Yabroff et al. (1999) categorised strategies to increase mammography participation into three groups: i) behavioural strategies which alter the cues or stimuli associated with screening and include such things as invitations and reminders; ii) cognitive strategies which provide new information and education, increase knowledge and clarify misperceptions; and iii) sociological interventions which use social norms, peers, friends and lay health advisors to increase screening behaviour. It is the sociological interventions that most closely resemble the definition of coaching. Yabroff et al. (1999) concluded that sociological interventions improved mammography utilisation by 12.6% (95% CI 7.4–17.9).51 Ellis et al. (2005) analysed three systematic reviews relating to social networks (including the Yabroff et al. (1999) review) and concluded that networks such as community peers and lay health advisers increase mammography participation.58  Pap tests In a review published in 2002, Black et al. concluded that strategies involving lay health workers or volunteers using individual or group approaches, and coming from the same cultural and/or socioeconomic background as the target population were significantly more effective in promoting Pap test participation compared to controls.52, 54 Yabroff et al. (2003) studied the impact of sociological strategies on participation in Pap testing. Most of the sociological strategies included in this review targeted specific ethnic groups and used lay health workers. The authors concluded that most of the interventions improved participation but not all were statistically significant.107 Scrutiny of the effect sizes calculated for the individual studies indicate that only two of the six studies (33%) had statistically significant positive results in favour of the intervention, and one of the strategies was a culturally sensitive video that, in the categorisation of strategies used for the current review, would be included under education. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 35 Summary The peer‐reviewed literature in relation to coaching is dominated by studies originating in the USA, and is mainly aimed at increasing the uptake of mammograms and Pap tests. Only one of the interventions targeted both men and women, and only 2 did not specifically target hard‐to‐
reach groups. Half of the interventions increased the uptake of screening. However, in terms of the proportion of interventions that increased uptake, this strategy appears to be most effective when targeted at ethnic groups in an urban environment (Table 23). Table 23 Overview of the impact of coaching on screening uptake Increased uptake Impact unclear No impact n % n % n % Non‐urban (a) 1 33 2 67 Low income (b) 1 25 3 75 Ethnic groups (c) 4 100 General population 1 50 1 50 7 54 3 23 3 23 Notes: (a) 2 targeted non‐urban ethnic groups (b) All targeted low‐income ethnic groups (c) 2 targeted Native Americans, one each of Vietnamese, Chinese, Cambodian, Latino. Total n 3 4 4 2 13 % 100 100 100 100 100 The reviews concluded that this form of intervention increases the uptake of mammography but conclusions were less consistent in relation to Pap testing. 4.6 Community interventions Peer‐reviewed primary studies There were 13 peer‐reviewed studies relating to community intervention strategies; one Australian, one Mexican, and eleven from the USA (Table 23). Five of the studies were aimed at improving participation in more than one type of cancer screening, and only two did not target a hard‐to‐reach group. Almost two‐thirds (62%) of the studies were Level III and seven (54%) indicated statistically significant improvements in screening, but for two of these the impact varied by type of cancer. The studies with no statistically significant results included two Level II studies and one Level III study. The unclear results included one Level III study and two Level IV studies. Table 24 Country Australia (1) Mexico (1) USA (11) Peer‐reviewed primary studies: community interventions (n = 13 studies) Type of cancer Breast only (3) Cervical only (4) Colorectal only (1) Breast & cervical (3) Breast, cervical, colorectal (2) Hard‐to‐reach Ethno‐specific (7) Low income (1) Non‐urban (2) Homeless (1) Level of evidence II (3) III‐2 (7) III‐3 (1) IV (2) Impact Stat sig results (5) Varied by test (2) Results unclear (3) No stat sig results (3) VOL 3– 36 Knowledge translation: a review of strategies to increase participation in cancer screening  Screening mammograms (3 studies) One study did not report a statistically significant positive impact on the uptake of mammograms.131 In one study the results were unclear132 and in one there was a statistically significant positive impact on uptake.133  Pap tests (3 studies) Two studies reported statistically significant positive impacts on the uptake of Pap tests134, 135
and one Australian study did not.136  Faecal occult blood testing (1 study) In this Level III study the results were unclear.137  Multi‐tests (6 studies) Mammograms and Pap tests (4 studies): One study reported statistically significant positive impacts on screening rates.138 In two studies the impact varied by screening test. Both reported statistically significant positive impacts on the uptake of Pap tests but not on the uptake of mammograms.139, 140 In the remaining study, the results were unclear.141 Mammograms, Pap tests and FOBT (2 studies): One Level III study reported statistically significant positive impacts on screening uptake across all tests142 and one Level II study did not.143 Hard‐to‐reach groups Eleven of these community intervention studies were targeted at hard‐to‐reach groups.  Ethno‐specific (7 studies) The target groups for these interventions were: i) African American; ii) Indigenous groups in Mexico and Hawaii; and iii) Vietnamese and Latinos in the USA. Five of the 7 studies reported statistically significant positive impacts on the uptake of one or more screening tests,134, 135, 138, 139, 142 one did not143 and in one, the results were unclear.141  Non‐urban/low income/homeless (4 studies) Two Level II studies target non‐urban populations. One reported statistically significant positive impacts on uptake133 and one did not.136 One Level III study targeted low‐income populations, but the results were unclear.137 One Level III study targeted the homeless and the results varied by test.140 Literature reviews and grey literature Only one of the systematic reviews looked at community interventions. Jepson et al. (2000) described community interventions as 'interventions aimed at whole communities which often involve multiple interventions including media campaigns and community participation'. According to the authors, the community intervention studies rated poorly in terms of quality. For example all studies used a cluster design but only two used the cluster as the unit of analysis. They concluded that there was some evidence of effectiveness of multi‐component community interventions.18 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 37 Summary The peer‐reviewed literature is dominated by studies originating in the USA. All but 2 of the interventions were targeted at hard‐to‐reach groups (i.e. specific ethnic groups, non‐urban and low‐income groups). Nearly two‐thirds of the interventions increased uptake in the hard‐to‐
reach groups but none reported increase in uptake in the general population (Table 25). The only review which considered this type of intervention concluded that there was some evidence for the effectiveness of community interventions but the quality of the evidence was poor. Table 25 Overview of the impact of community interventions on screening uptake Hard‐to‐reach General population Total Increased uptake (a) n % 7 64 7 54 Impact unclear n % 2 18 1 50 3 23 No increase n % 2 18 1 50 3 23 Total n 11 2 13 % 100 100 100 Notes: (a) Includes two interventions in which the intervention increased the uptake of Pap tests but not mammograms. 4.7 Mass media campaigns
Peer‐reviewed primary studies The peer‐reviewed literature contained four studies relating to mass media campaigns that were not embedded in larger interventions. Three of these related to Australian campaigns and one to a campaign in the USA targeted at Vietnamese Americans (Table 25). Only one of the Australian studies indicated that the campaign had a statistically significant, positive effect on screening rates.144 The two studies with no statistically significant results included one Level II study and one Level IV study. The unclear result was for a Level III study. Table 26 Country Australia (3) USA (1) Peer‐reviewed primary studies: mass media campaigns (n = 4 studies) Type of cancer Cervical (2) Breast and cervical (1)
Breast (1) Hard‐to‐reach Ethno‐specific (2) Level of evidence II (1) III‐2 (1) III‐3 (1) IV (1) Impact Stat sig findings (1) No stat sig findings (2) Findings unclear (1) Literature reviews and grey literature A 1998 review of mass media campaigns on cervical cancer screening rates cited in Mullins (2008) concluded that: 'the effects of mass media campaigns tend to be of short durations, and these strategies are best used in combinations with other approaches. Conventional wisdom, as well as a great deal of empirical evidence suggests that the strength of mass media campaigns is to increase awareness (i.e. ‘agenda setting’) and to VOL 3– 38 Knowledge translation: a review of strategies to increase participation in cancer screening provide a background context for other intervention strategies, rather than serving as a singular cue to action for behaviour change.'144, 145 Three reviews published between 2000 and 2004 have reached the conclusion that there was insufficient evidence to determine the effectiveness of mass media campaigns alone as a strategy for increasing participation in cancer screening.18, 53, 146 In the Black et al. (2002) review, only one of the 4 studies that related to mass media campaigns alone was effective. That study targeted a specific sub‐population. The authors concluded that paid publicity on ethnic radio may be an effective strategy for increasing Pap testing participation among women from non‐English speaking backgrounds, but concluded in relation to Pap testing that: 'Mass media campaigns should not be implemented on their own, but rather should be complemented with other social marketing and promotion interventions (e.g. group education, free screening, physician education, and/or letters of invitation or reminder) in order to obtain the greatest effect.’52, 54 Ellis et al. (2005) in an overview of systematic reviews found that: i)
for mammography one review showed increased rates but two others reported conflicting results; and ii)
for Pap testing there was inadequate information about the impact.58 In 2008, Baron et al. (2008), concluded that there was insufficient evidence to determine the effectiveness of mass media alone in increasing breast, colorectal, and cervical cancer screening rates.60 (2) In Australia there have been two national cervical screening media campaigns; the first in 1993 and the second in 1999.144 According to Fernbach (2001) the 1999 campaign 'increased the number of Victorian women who attended for a test at the time the campaign was on air by 16%.' 144, 147 In 2004, the Cancer Council Victoria developed a campaign titled Don't Just Sit There which was designed to increase cervical cancer screening rates in under‐screened women. Mullins (2008) evaluated the Victorian campaign using an interrupted time series design (the Level III‐3 study inTable 26 had statistically significant results) using Victorian Cervical Cytology Registry (VCCR) data for the period June 2002–December 2004. The results of the negative binomial regression indicated that 'the campaign had a significant positive effect on screening behaviour, with an 18% increase during the campaign, and this result was statistically significant'. A similar campaign was run in New South Wales in 2007 and was evaluated using a similar design to the Victorian evaluation. Perez et al. (2009) concluded that the 'number of Pap tests across all ages …[increased] by 22 per cent during the campaign, compared to the same period in 2006.'148 These evaluations of Australian media campaigns indicate that they have an effect, at least in the short‐term. As shown in Figure 3, the cervical screening rates did increase in Victoria and New South Wales in 1998–99 when the national campaign was run. However, the long‐term 2
Conclusions available at http://www.thecommunityguide.org/cancer/screening/client‐
oriented/index.html Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 39 linear trend in Victoria is downward. In New South Wales, which started from a lower base than Victoria, the long‐term trend is slightly upwards. Due to lack of available data, it is not possible to draw any firm conclusions about the longer term impact of the Don't Just Sit There campaigns.
Age‐standardised cervical screening rates 1996–97 to 2004–05 Age Standardized Screening Rates
Figure 3 70
65
60
55
50
1996–
1997
19971998
19981999
19992000
20002001
20012002
20022003
20032004
20042005
NSW
55.0
58.1
59.4
58.9
59.1
59.4
58.8
58.4
58.2
VIC
66.7
66.9
67.7
65.2
64.6
64.9
64.2
64.8
65.4
Source: AIHW (2008)7 Excel Spreadsheet csa04‐05‐xdi1.xls available at www.aihw.gov.au/publications/can/csa04‐05/csa04‐05‐xdi1.xls A radio and print media campaign was developed for women who spoke Arabic, Cantonese, Greek, Italian or Mandarin, to increase Pap test screening rates among these groups. Media space was booked in multicultural print and radio for seven weeks. An evaluation of the impact of PapScreen's campaign on culturally and linguistically diverse (CALD) women indicated that there was no evidence that the CALD campaign increased screening rates across Victoria, and there was no overall change in the number of tests done during the campaign for either high or low CALD LGAs.149 Summary Three of the 4 peer‐reviewed studies related to interventions undertaken in Australia and two of the four targeted ethnic groups. From this literature the evidence related to the impact of media campaigns on breast and cervical screening rates does not support the conclusion that media campaigns, by themselves, have and statistically significant positive impact on screening rates. Additionally, literature reviews do not offer clear evidence in relation to the impact of media campaigns. The Australian grey literature provides some evidence for, at least, the short‐term impact on Pap test screening rates, but not for evidence of an increase in relation to hard‐to‐
reach groups. Overall, it appears that media campaigns may have a short‐term impact on screening rates but there is insufficient evidence to conclude that one‐off media campaigns will, by themselves, increase screening rates in the longer term. VOL 3– 40 Knowledge translation: a review of strategies to increase participation in cancer screening 4.8 Worksite interventions Peer‐reviewed primary studies There were four peer‐reviewed studies relating to strategies involving interventions in worksites (Table 27). The two Level II studies indicated that the strategy of intervening in worksites has statistically significant positive results, and for the two Level IV studies the results were unclear. Three of the studies related to colorectal screening and two were undertaken in industrial (manufacturing) worksites, one in a regional area in the UK. Table 27 Country Taiwan (1) UK (1) USA (1) USA (1) Peer‐reviewed primary studies: worksite interventions (n = 4 studies) Type of cancer Colorectal only (3) Breast & cervical (1) Hard‐to‐reach Industrial (1) Industrial/regional (1) Level of evidence II (1) IV (2) Impact Stat sig results (1) Results unclear (2) II (1) Stat sig results (1)  Screening mammograms and Pap tests (1 study) USA: This was a 16‐month intervention undertaken in health, government and tertiary education organisations. The strategy had no impact on the proportion of participants having a mammogram within the last 2 years. However, employees exposed to the intervention were more likely than employees in the control sites to have had a Pap test within the last 3 years.150  Faecal occult blood testing (3 studies) UK: The intervention in the UK was in a regional industrial organisation. The impact of the intervention on screening rates was unclear, however, overall compliance with FOBT testing was 25.4% and there were no differences between the screening rates for men and women. Managers had higher compliance rates than other workers and the difference was statistically significant (28.6% vs 23.5%, p < 0.02).151 USA: The worksite intervention in the USA was targeted at past and present employees who had worked in particular areas. The trial lasted for 2 years and employees could complete one or more of a number of colorectal screening tests. There was a small but statistically significant improvement in screening rates compared to the control sites, however, the results for FOBT screening was unclear.152 TAIWAN: FOBT kits were distributed to employees in 10 different worksites in Taiwan. The completion rate was 74%.153 Literature reviews and grey literature There was very limited literature relating to interventions targeting the workplace. A study included in the Jepson et al. (2000) review was of a multi‐component worksite intervention (print media, onsite mammography workshops and incentives). Uptake increased in both the intervention and control groups but there was no comparison of the differences between the groups.18 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 41 In a review published in 2004, Pasick et al. concluded that 'as settings for cancer screening promotion and services, their effectiveness in reaching under‐screened subgroups has yet to be demonstrated because of the lack of high quality research.'146 Ellis et al. (2005) in a review of systematic reviews concluded that making use of 'opinion leaders' in the workplace is a promising strategy but no evidence of the impact on cancer screening rates was included to support the conclusion.58 Summary There were four primary studies relating to worksite interventions; two in the USA, one in the UK and one in Taiwan. In two studies there were statistically significant increases in screening and in two the results were unclear. The worksite intervention in Taiwan targeted colorectal screening and had impressive completion rates compared with completion rates for FOBT in the UK intervention. However, given the cultural differences between workplaces in Taiwan and non‐Asian countries, rates for workplace interventions in Australia are not likely to be as high as those achieved in Taiwan. The literature reviews offer very little evidence for the effectiveness of worksite interventions. Overall the conclusion appears to be that these interventions are promising but the evidence is limited as to their effectiveness. 4.9 Financial incentives/disincentives Peer‐reviewed primary studies There were three peer‐reviewed studies relating to strategies involving interventions with economic incentives/disincentives (Table 28). Two studies were targeted at low income women and both reported statistically significant, positive results for the intervention. The Finnish study (Level IV) revealed that the introduction of a fee reduced the level of screening. Table 28 Country USA (2) Finland (1) Peer‐reviewed primary studies: financial incentives (n = 3 studies) Type of cancer Breast only (3) Hard‐to‐reach Low income (2) Level of evidence II (1) III‐2 (1) IV (1) Impact Stat sig results (2) Stat sig result (fee caused reduction in screening) (1) All three interventions reported statistically significant impacts on the uptake of mammograms. One involving the provision of incentives for low‐income women reported an increase in uptake.20 Another which was also aimed at low income women, and involved providing insurance coverage for screening reported increased uptake in Hispanic women but not black or white women.154 The third study reported on the impact of charging a fee for having a mammogram and reported a decrease in the uptake of screening.155 Literature reviews and grey literature Financial incentives have been divided into two groups in the literature reviews: i) the reduction in financial barriers; and ii) the provision of financial incentives. Jepson et al. (2000) concluded that reducing financial barriers was effective in increasing uptake across a range of screening VOL 3– 42 Knowledge translation: a review of strategies to increase participation in cancer screening programs but that rewards and incentives do not seem to do so.18 This latter finding is contradicted by a meta‐analysis undertaken by Stone et al. (2002) which indicated the provision of financial incentives had statistically significant positive effects on the uptake of: i)
screening mammography (OR = 2.74, 95% CI 1.78–4.24) ii)
Pap tests (OR = 2.82, 95% CI 2.35–3.38) iii)
FOBT (OR = 1.82, 95% CI 1.35–2.46).53 A review of systematic reviews by Ellis et al. (2005) concluded that the removal of financial barriers increased mammography and Pap test participation rates.58 In 2008 Baron et al. concluded that there was sufficient evidence that reducing out‐of‐pocket costs increases breast cancer screening by mammography, but there was insufficient evidence to determine effectiveness in relation to Pap tests or FOBT.60 In another review looking at interventions to increase demand for cancer screening, Baron et al. (2008) concluded that there was insufficient evidence to determine the effectiveness of client incentives (i.e. small non‐coercive rewards in the form of cash or coupons) alone in increasing screening for breast, cervical or colorectal cancer.60 Summary Based on limited evidence it appears that: i) making screening available free of charge does not necessarily increase screening rates across all ethnic groups; and ii) offering financial incentives to low‐income women increases participation. However, this latter result is based on one Level II study only. 4.10 Procedures
Peer‐reviewed primary studies There were 15 peer‐reviewed studies related to strategies involving changes to the way the tests are conducted. Unlike many of the other interventions, the studies related to interventions across a number of different countries (Table 29). Six related to FOBT testing, 6 to Pap tests and 3 to mammograms. Just under two‐thirds (60%; six Level II, two Level III, one Level IV) had statistically significant results, indicating that changes in procedures increased screening rates. One Level II study was significant after a post hoc test was performed. The 3 studies with no statistically significant results included one Level II study, one Level III, and one Level IV study. The unclear results were a Level III study and a Level IV study. Table 29 Peer‐reviewed primary studies: procedures (n = 15 studies) Country Australia (2) Italy (2) USA (3) One each from: France, India, Israel, Finland, Netherlands, New Zealand, Sweden, Taiwan Type of cancer Breast only (3) Cervical only (6) Colorectal only (6) Hard‐to‐reach Ethno‐specific (1) Non‐urban (2) Low income (1) Level of evidence II (8) III‐1 (1) III‐2 (2) III‐3 (1) IV (3) Impact Stat sig results (9) Stat sig result (post hoc test only) (1) No stat sig results (3) Results unclear (2) Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 43  Screening mammography (3 studies) Two Level II studies reported statistically significant positive impacts on the uptake of mammograms. One involved the offer of opportunistic screening and free transport when women visited their doctors for other reasons156 and the other involved using health checks to recommend screening.157 The Level IV study on the use of a mobile screening van in New Zealand did not report a positive impact.158  Pap tests (6 studies) Four studies reported on the use of self‐sampling devices (Netherlands, Sweden, Finland India). Two reported a statistically significant positive impact on the uptake of screening.159, 160 In one study the results were unclear161 and in one there was no statistically significant impact on uptake.162 One study investigated the use of a mobile clinic in Taiwan, and the other investigated training Community Health Workers to undertake Pap tests. In the first study the results were unclear163 and in the second there was no impact on screening rates.164  Faecal occult blood testing (6 studies) Four studies investigated whether the manner in which the FOBT kit was distributed impacted upon uptake, and all four reported statistically significant uptake rates.78, 165‐167 One study reported statistically significant positive impacts on uptake when the dietary restrictions were lifted168 and another found that the type of kit had statistically significant impacts on uptake.169 Hard‐to‐reach groups  Ethno‐specific (1 study) The study reporting on the training of Community Health Workers to undertake Pap testing was targeted at native Alaskans and did not have a statistically significant impact.164  Non‐urban (2 studies) The New Zealand study investigating the use of a mobile van was targeted at a non‐urban population, and did not have an impact on the uptake of mammograms.158 The other study involved self‐sampling for Pap testing by rural women in India, and it reported a statistically significant positive impact on uptake.160  Low income (1 study) The study which investigated the use of opportunistic screening and free transport was targeted at low‐income women and it reported statistically significant positive increases in screening.156 Literature reviews and grey literature  Dietary restrictions, test period (faecal occult blood testing) Jepson et al. (2000) included 2 studies relating to diet in their review. In one the decrease in uptake when participants were asked to restrict their diet was not statistically VOL 3– 44 Knowledge translation: a review of strategies to increase participation in cancer screening significant, and in the other it was statistically significant. They also included two studies that compared uptake when participants were required to perform FOBTs over a 3‐ or 6‐
day period. In one there was a small, statistically significant decrease in uptake among participants offered the 6‐day testing, but in the other there was no difference.18  Opportunistic screening Jepson et al. (2000) concluded that there was some evidence to suggest that opportunistic screening may be effective in increasing participation. The review included four studies relating to Pap tests (3 studies) and FOBT (1 study). The FOBT study found that uptake was higher when GPs offered an FOBT kit opportunistically compared to sending the FOBT with a letter (RR = 1.48; 95% CI 1.43–1.55). Among the three Pap test studies: o one found that opportunistic screening was less effective than a combined invitation for cervical smear testing and breast screening (RR = 0.60, 95%CI 0.47–0.77) o another targeted 'poor, elderly black women', and indicated that among this group opportunistic screening increased uptake o the last indicated that offering women Pap tests while they were in hospital increased uptake.18 In a 2004 review of the literature relating to outpatient cervical screening, Anderson concluded that there was 'mixed evidence' for the introduction of an 'opportunistic' outpatient screening program in Victoria. But, the offer of same‐day screening (e.g. having a near‐by Pap test clinic), compared to patients making appointments for a test at a later date, is likely to increase the uptake rate of opportunistic screening.62  Periodic Health Evaluations In a review of Periodic Health Evaluations (PHE), Boulware et al. (2007) found greater rates of delivery of Pap tests and FOBT in persons undergoing the PHE than in those not undergoing the PHE, but had mixed effects on mammography. The authors concluded that it was possible that the PHE is more effective in improving the delivery of preventive services that are performed at the time of the visit than on services that require patients to schedule appointments outside the visit for the PHE.170  Mobile vans, transport In a review of systematic reviews, Ellis et al. (2005) concluded (on the basis of two systematic reviews) that participation increases with the use of mobile vans and the provision of transport.58 Baron et al. (2008) concluded that: 'There is strong evidence that reducing structural barriers is effective in increasing mammography and FOBT. Questions remain about whether additional interventions are needed when focusing on specific populations, such as people who have never been screened or who may be hard to reach for screenings. There is insufficient evidence to determine the effectiveness of reducing structural barriers in increasing Pap testing because there were too few qualifying studies with adequate quality of design and execution.'60 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 45 Summary The peer‐reviewed literature originated from studies undertaken in various countries and was mainly focused on increasing uptake of Pap tests and FOBT. Overall, two‐thirds of the interventions increased uptake, and all the interventions aimed at increasing the uptake of FOBT did so (Table 30). Four of the interventions targeted hard‐to‐reach groups; two reported increases in screening uptake and two did not. Table 30 Overview of the impact of procedures on screening uptake Pap test (self sampling) FOBT Mobile clinics/vans Other Increased uptake n % 2 50 6 100 2 67 10 67 Impact on uptake unclear n % 1 25 1 50 2 13 Did not increase uptake n % 1 25 1 50 1 33 3 20 Total n 4 6 2 3 15 % 100 100 100 100 100 The literature reviews and grey literature were generally positive about the potential for opportunistic screening, PHEs, and the use of mobile vans. The evidence for changing dietary restrictions and methods of screening for FOBT were mixed. 4.11 Multi‐component interventions Peer‐reviewed primary studies There were 21 peer‐reviewed studies relating to multi‐component interventions (Table 31), with the majority originating from the USA. Ten studies (48%) were aimed at hard‐to‐reach groups, and just over half (53%) of the studies indicated statistically significant improvements in screening (8 Level II and 3 Level III). The seven studies that did not have statistically significant results were all Level II. Table 31 Country Australia (1) UK (1) Sweden (1) USA (18) Peer‐reviewed primary studies: multi‐component (n = 21 studies) Type of cancer Breast only (11) Cervical only (2) Colorectal only (4) Breast & cervical (3) Breast, cervical & colorectal (1) Hard‐to‐reach Ethno‐specific (7) Low income (1) Non‐urban (2) Level of evidence II (16) III‐1 (1) III‐2 (2) III‐3 (1) IV (1) Impact Stat sig results (11)
Results unclear (3) No stat sig results (7)  Screening mammograms (11 studies) Six studies reported statistically significant positive impacts on the uptake of mammograms.63, 67, 74, 75, 171, 172 Two Level II studies did not have an impact on the uptake of mammograms.102, 133 In three studies (one each of Level II, Level III, Level IV) the results were unclear.173‐175 VOL 3– 46 Knowledge translation: a review of strategies to increase participation in cancer screening  Pap tests (2 studies) One Level III study reported a statistically significant positive impact on the uptake of Pap tests176 and one Level II study did not.29  Faecal occult blood testing (4 studies) Two Level III studies reported statistically significant positive impacts on the uptake of FOBT177, 178 and two Level II studies did not.179, 180  Multi‐tests (4 studies): Mammograms and Pap tests (3 studies): Two Level II studies reported statistically significant impacts on uptake,72, 181 and one Australian study did not.182 Mammograms, Pap tests and FOBT (1 study): This Level II study did not report a statistically significant impact on uptake.71 Hard‐to‐reach groups Included in the above results were ten studies targeting hard‐to‐reach groups.  Ethno‐specific (7 studies) Three studies reported statistically significant positive impacts on the uptake of mammograms by African American women63, 75, 172 and one that targeted ethnic minorities in Wales did not.173 One study reported a positive impact on the uptake of Pap tests by low‐income Korean women in the USA176 and one study reported that an intervention designed to increase the uptake of FOBT did not have an impact among Native Hawaiians.179  Low income (2 studies) One study reported a statistically significant positive impact on the uptake of FOBT among a low income, ethnically diverse target group178 and the other reported a statistically significant positive impact upon the uptake of mammograms among low income women.171  Non‐urban (2 studies) Neither study reported a statistically significant impact on the uptake of mammograms or Pap tests.133, 182 Literature reviews and grey literature  Invitations and reminders + o Screening mammography Yabroff et al. (1999) found that a combination of behavioural (e.g. invitations and reminders) plus cognitive (educational) strategies to increase mammography had variable effectiveness, ranging from little effect (n = 3) to a maximum effect of 33% (n = 1 study).51 Two studies included in the Jepson et al. (2000) review reported statistically significant positive impacts on uptake for: i) invitation plus follow‐up telephone counselling plus a $15 grocery incentive; and ii) invitation letter plus education plus mammography van.18 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 47 Bonfill et al. (2009) found that the combination of letters of invitation to multiple examinations plus educational material did not increase uptake, and concluded that this form of intervention should be avoided.55 o Pap tests Two of the three interventions used mailed generic educational information and reminders, and were not associated with significant improvements in participation rates. The third which incorporated a telephone call from a health educator with the reminder, reported a 13.5% increase in participation rates for the intervention group (95% CI 7.3–19.4). Anderson (2004) reported on a study in which a brochure containing detailed information did not have a statistically significant effect on participation (+1.3% difference; 95% CI –0.03–2.9).183 Multi‐test (mammograms and Pap tests) One study in the Jepson et al. (2000) review indicated that invitation letters plus telephone counselling increased uptake of cervical and breast cancer screening by low‐
income women.18 o Faecal occult blood testing The Jepson et al. (2000) review contained one RCT which found that a personal letter from a GP with an appointment and an educational booklet was more effective than a letter by itself (RR = 1.25; 95% CI 1.17–1.33). In an Australian study, Millard (2006) concluded that despite targeted liaison and education programs, the response to postal invitations was reduced among people with historically poor uptake of health services (i.e. the culturally and linguistically diverse, and Aboriginal and Torres Strait Islander communities).37  Sociological + In the Yabroff et al. (2003) review, one of the two studies that used sociological and cognitive strategies improved participation. Two of the three sociological, behavioural and cognitive interventions led to large increases in participations rates. The first used lay health workers, educational pamphlets and financial incentives and was culturally specific to Vietnamese American women (13%). However, the authors concluded that the apparent effectiveness of the second multi‐component intervention may have been due to the decline in participation from 39% to 17% in the concurrent control group.107 Summary The peer‐reviewed literature was dominated by studies originating in the USA, and most related to the uptake of mammograms and/or Pap tests. Overall, 11 of the 21 interventions improved uptake. Ten of the interventions targeted hard‐to‐reach groups and 6 increased uptake. All three strategies aimed at low income groups increased uptake, but neither of the two targeted at non‐urban groups increased uptake (Table 32). The literature reviews contained some evidence that some multi‐component interventions improved uptake of screening rates, but it was difficult to discern patterns or trends. VOL 3– 48 Knowledge translation: a review of strategies to increase participation in cancer screening Table 32 Non‐urban Low income Ethno‐specific Gen. population Overview of the impact of multi‐component interventions Increased uptake n % 3 100 3 60 5 45 11 52 Impact unclear n % 1 20 2 18 3 14 No impact n % 2 100 1 20 4 36 7 33 Total n 2 3 5 11 21 % 100 100 100 100 100 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 49 5 STRATEGIES TARGETING PROVIDERS Of the community interventions reported in the previous section, five specifically mentioned having a strategy targeting providers.131, 134, 135, 137, 140 These five studies have not been included in this section because it is impossible to determine the impacts of the different components. Eleven studies have been included, and all the interventions took place in the USA. Nine of the 11 studies had results relating to provider‐only interventions (Table 33) and four contained results for strategies targeting both the provider and the screening population (Table 34). 5.1. Provider only strategies Table 33 Country USA (9) Peer‐reviewed primary studies: provider only strategies (n = 9 studies) Type of cancer Breast only (3) Colorectal only (3) Breast, cervical, colorectal (3) Hard‐to‐reach Low literacy (1) Level of evidence II (6) III‐2 (3) Impact Stat sig results (1) No stat sig results (6) Varied by type of cancer (2)  Provider audit and feedback (1 study) One study with statistically significant results indicated that patients whose doctors had attended some feedback sessions had higher FOBT completion rates than those patients of doctors who had not attended any of the sessions (30.2% vs 17.9%, p = 0.01).101  Provider education (4 studies) None of the four provider education strategies (3 mammography, 1 FOBT) had statistically significant results.85, 108, 112, 184  Provider prompts (3 studies) All 3 provider prompt strategies were multi‐test interventions, aimed at increasing the uptake of mammograms, Pap tests and FOBT. One did not report statistically significant impacts71 and two did.185, 186 However, both the citations with statistically significant results reported on the same intervention – one at 12 months follow‐up and the other at 24 months. At 12 months, the intervention had succeeded in increasing the uptake of mammograms and FOBT,182 and at 24 months only the uptake of FOBT had improved.185, 186
 Multi‐component (1 study) One study involved provider education on the use of prompts and the subsequent provision of provider prompts for FOBT.33 This study did not report statistically significant results. VOL 3– 50 Knowledge translation: a review of strategies to increase participation in cancer screening 5.2 Table 34 Country USA (4) Provider and screening population strategies combined Peer‐reviewed primary studies: provider/screening‐population strategies (n = 4 studies) Type of cancer Breast only (4) Hard‐to‐reach Low income (1) Level of evidence II (3) III‐2 (1) Impact Stat sig results (1) No stat sig results (2) Unclear (1)  Provider education + patient counselling (1 study) No statistically significant results.112  Provider prompts + patient invitations and reminders (1 study) No statistically significant results.25  Provider education + patient invitations and reminders (1 study) The results are unclear, however a subgroup analysis indicated that the combined strategies resulted in a statistically significant increase in mammograms in the 8 weeks following the intervention in the 'never‐screened' but not in the 'ever‐screened'.175  Provider education + patient education + patient counselling (1 study) This intervention aimed at increasing the uptake of mammograms in women 65 years and older, and the combination of strategies resulted in a statistically significant positive impact (14.4% vs 10.2%, p = 0.020). Subgroup analysis indicated that the statistically significant positive effect was confined to the group of women who had not had a mammogram in the previous 15 months, rather than those who had a mammogram in the 12 months preceding the baseline data collection (i.e. those up to date with screening).85 5.3 Hard‐to‐reach groups  Ethno‐specific None of these interventions targeted specific ethnic groups.  Low‐income/low literacy The provider prompts + patient invitations and reminders strategy was targeted at low income women but did not have a statistically significant impact on the uptake of mammograms.25 The provider audit and feedback strategy was targeted at patients with low literacy, and this study did report statistically significant increases in the uptake of FOBT.101 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 51 5.4 Literature reviews and grey literature In a review of provider strategies Jepson et al. (2000) concluded that there was:  evidence of effectiveness of prompts;  some evidence to suggest that audit and feedback increase uptake;  not enough evidence that physician education increases uptake;  evidence that provider prompts and invitations to individuals increase uptake.18 However, close reading of the details of the studies on which these conclusions were reached indicates that these conclusions may be optimistic in relation to mammography and Pap tests. In 2001, Kupets and Covens published a review of provider strategies, and reached the conclusion that 'interventions targeting physicians alone resulted in absolute increases in cervical screening rates ranging from 9% to 40%. Breast cancer screening was increased by 6–
35%.'187 As shown in Table 34, the interventions with the highest success rates included physician reminder systems, both computerised and manual. Table 35 Impact of provider strategies on the uptake of screening mammograms and Pap tests All studies Computer‐generated prompts Flowsheet/Information sheet on chart Audit and feedback Computer prompts + audit and feedback 6 (SSR = 4) Pap tests Increases in screening rates in studies with SSR 9%–30% Screening mammograms All studies Increases in screening rates in studies with SSR 6 (SSR = 4) 6%–30% 3 (SSR = 1) 40% 2 (SSR = 1) 35% 2 (SSR = 0) 1 (SSR = 0) 1 (SSR = 1) 1 (SSR = 0) 14% Source: Kupets and Covens (2001)187 Notes: SSR: Statistically significant results Yabroff et al. (2000) reviewed provider targeted interventions to increase the uptake of Pap tests. Of the 19 studies relating to prompts included in the review, five had statistically significant positive results. The impact on uptake ranged from a statistically significant 18% decrease to a 44% increase. One of the three studies which examined the impact of seminars or audit and feedback had statistically significant results, with an 8% increase in uptake.107 In a review of systematic reviews, Ellis et al. (2005) concluded that:  prompts or reminders increase the uptake of mammography (10 reviews) and Pap tests (eight reviews).  training and education increase mammography uptake (4 reviews) but the results for Pap tests were inconsistent (3 reviews). VOL 3– 52 Knowledge translation: a review of strategies to increase participation in cancer screening  audit and feedback in combination with other interventions may lead to increases in the uptake of mammography (4 reviews).58 Summary There is scant evidence in the peer‐reviewed primary studies for the effectiveness of the strategies targeting providers. The literature reviews appear to indicate that prompts may have some effect in terms of increasing uptake for mammograms and Pap tests. However, the evidence for the impact of prompts on FOBT is not clear. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 53 6. ENGAGING THE HARD‐TO‐REACH A number of issues dominated in the search for parallel evidence. The first was a resourcing issue. Given the breadth of the literature relating to cancer screening, it was not possible to undertake a similar analysis of even one area that could have been expected to yield a similar amount of literature (e.g. cardiovascular disease) within the given time and resource constraints. The second was the difficulties in determining similar screening or testing behaviour that would engender some confidence that the evidence would translate to mammography, Pap testing and FOBT. The third consideration related to the inclusion criteria for the studies in relation to knowledge translation, i.e. studies which included interventions designed to have an impact on screening uptake. This excluded some of the newer activities where trials have been reported (e.g. screening for aneurysms) but for which there are not studies reporting manipulation of interventions designed to increase uptake. To overcome these difficulties, this section on engaging hard‐to‐reach groups has been divided into three sub‐sections. 1. The first focuses on engaging men in health promotion activities. The first part presents appropriate subgroup analyses from the FOBT interventions included in the primary studies in Section 4. The aim was to identify if any of those strategies appear to have been most effective in engaging men. The second part includes a more general discussion of the evidence relating to engaging men in health related behaviours. 2. The second focuses on engaging hard‐to‐reach women in cancer screening behaviour. It includes a re‐analysis of the primary studies focusing on studies targeting specific ethnic groups, low‐income groups and non‐urban populations. It also includes the results of two literature reviews (one from the peer‐reviewed literature, and one from the grey literature) that focus on the effectiveness of interventions designed to increase the uptake of mammography and Pap tests by women who are either under‐served or under‐
screened. 3. The final sub‐section presents information gleaned from three literature reviews that focused not only on the uptake of cancer screening, but also on other health related activities. 6.1 Engaging men
Colorectal cancer screening (primary studies) This section of the literature review appraises the intervention studies contained in section 4 to examine the effect of gender on the rate of participation in colorectal cancer screening (CRCS). More specifically, it aims to highlight intervention strategies that may potentially be implemented in Australia to promote the participation of men in CRCS. A total of thirty‐nine studies pertaining to CRCS were identified from the original list of studies. Twenty‐six of these studies reported screening programs implemented in the USA and another 6 were from Australia. The remaining studies were programs conducted in several European countries, Israel and Taiwan. Two studies that targeted those aged 40 years and above; in all VOL 3– 54 Knowledge translation: a review of strategies to increase participation in cancer screening other interventions the target population was aged 50 years and above. Only twenty‐one of the thirty‐nine presented subgroup analyses by gender. Two types of subgroup analyses were evident in the studies: i)
analyses of gender as a predictor to participation in order to examine differential participation between men and women; and ii)
analyses of the effectiveness of the screening program by gender with no comparison of differences between men and women. It needs to be noted that not reporting a differential uptake by gender may be acceptable if it indicates that the intervention is equally acceptable to men and women. Such a conclusion relies on the requirement that screening rates reported in the study for both men and women are comparable to those currently being attained in Australia, unless the intervention specifically targets those who did not participate in screening prior to the intervention (i.e. hard‐to‐reach population).  Hard‐to‐reach groups There were three studies specifically targeted to populations that under utilised screening services, two of which aimed at improving screening rates in socially disadvantaged populations in the USA, and a third that was directed towards non‐responders to the routine recruitment efforts for CRCS in France.97, 167, 188 The study by Basch et al. (2006) reported that tailored telephone counselling increased FOBT rates among New York urban minority women (RR = 3.9; 95% CI 2.1–7.3) and in men (RR = 6.3; 95% CI 2.0–20.1). However, overall participation was significantly lower in men (n = 132) than in women (n = 324).188 The French study found that sending FOBT test kits to all non‐responders resulted in higher participation rates than sending the test kit only to those who requested it after receiving a recall letter (14.7% vs 8.3%, p < 0.00001). There was no difference between the groups in terms of gender.167 The study by Friedman et al. (2001) tested the efficacy of a videotaped intervention using peer educators as well as a health professional amongst low income attendants to a medical outpatient clinic.97 They found significant improvement in screening rates (43.6% vs 36.0%), and that gender was not a predictor of screening. However, no information was presented regarding the absolute screening uptake in men versus women. Although this intervention targeted low income individuals, the participants were in fact already accessing healthcare services. This probably explains the much higher screening rate observed in this study than in the other two studies.  General population In the studies aimed at the general population, the study participants were usually recruited from electoral rolls, healthcare databases, workplaces, or specific geographical locations. Four studies were conducted in South Australia.39, 168, 169, 189 The 2001 study by Coles et al. found that not having dietary restrictions when undertaking FOBT achieved a significantly higher participation rate for the mailed‐out FOBT than having dietary restriction (65.9% vs 53.3%). This study reported no differential uptake by gender (OR = 0.98; 95% CI: 0.77, 1.24).168 Subsequent studies conducted by Cole et al. examined the effectiveness of different intervention strategies such as having endorsement by primary care practitioners,189 various haemoglobin tests with improved testing procedures,169 or Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 55 receiving an advance notification letter prior to mailing out the testing kit.39 Although statistically significant gains in screening rates were observed compared with the control group in these studies, the high rate of participation reported in the 2001 study was not observed. The reported participation rates in these studies ranged from 23.4% to 48.3%. A differential uptake between women and men was only reported in the 2007 study involving the advance notification letter (RR = 1.13; 95% CI: 1.03, 1.26).39 The other studies did not observe differential uptakes. The study by Courtier et al. (2002) found that having a face‐to‐face invitation by a trained non‐health professional who supplied a signed invitation letter and FOBT kit resulted in a higher participation rate than not having a face‐to‐face invitation, both in women (45.2%) and men (48.7%).47 The study by Crotta et al. (2008) found that mailed invitation letters containing an appointment at a local municipal office where volunteers distributed testing kits resulted in high participation rates both in men (57.8%) and women (64.4%).35 Some other studies also reported promising final screening rates between 40% and 50%, and no differential uptakes between men and women were observed. The range of interventions examined in these studies included mailed invitation with mail‐out FOBT kits,37 information booklets,33, 34 telephone reminders,43, 178 and face‐to‐face education.190 There were a number of studies reporting substantial increases in the uptake of FOBT due to the intervention, but these studies may be of less relevance to the Australian decision context because the baseline participation rates were much lower than those observed in Phase I of the National Bowel Cancer Screening Project (36.8% for those aged 55 and 44.3% for those aged 65). The study by Church et al. (2004) reported substantial improvements in participation rates both in men and women but the baseline participation rate was 21.4%; (95% CI 19.2%–23.8%).38 The study by Markus and colleagues also had low baseline rates of screening in both women (20%) and men (22%), although the rates following interventions improved substantially to 39%–53% according to different intervention modalities.70 The workplace based study by Hart and colleagues was also not likely to be relevant because it resulted in substantially higher screening participation from women (49%) than men (34%) (Chi sq = 12.2, p < 0.001).191 Engaging men in healthcare – lessons from other areas At a conceptual level, two intervention approaches can be used to promote engagement in men: i)
Non sex‐specific interventions with a strategic focus on improving uptake in men; and ii)
A male‐specific intervention. Robertson et al. (2008) conducted a systematic review to examine the evidence relating to health promoting interventions targeting men.192 Seventeen of the 27 primary intervention studies included in this review were related to health conditions specific to men, such as prostate cancer and testicular cancer. The remaining studies were related to interventions that addressed a range of prevention areas, such as smoking cessation, diet and physical activities, cardiovascular disease, skin cancer and alcohol consumption. VOL 3– 56 Knowledge translation: a review of strategies to increase participation in cancer screening Overall, the authors found limited published evidence on how to improve male uptake of services.192 Consequently, it was not possible to reach a recommendation about whether to offer a generic intervention that is gender sensitive, or to design a male‐specific intervention. Nevertheless, this review did identify several novel intervention approaches that may be of interest. For example, in the study by Holland et al. (2005), a very high participation rate for preventive healthcare screenings (91.9%, n = 457) was achieved amongst men who received a personalised reminder followed by a postcard reminder addressed 'To Someone Who Loves [member’s name]' and whose provider received a reminder sticker for their medical chart.193 However, Robertson et al. noted that such indirect targeting of men through their doctors and spouses may be controversial.192 The systematic review also identified several interventions that were implemented in the workplace. Such interventions may have some merits from a theoretical viewpoint as ‘many men define themselves via their work, often feeling more comfortable in the workplace than in health oriented settings.’194 However, only men participating in the workforce can benefit from such intervention approaches.192 Summary Although there were individual studies that demonstrated the effectiveness of a particular strategy in improving participation rates among men, there is currently no clear evidence about which is the most effective strategy, especially within the Australian context. It is important to note, as Galadas et al. (2005) state: 'not all men are the same, nor does it make sense to assume that individual men behave similarly in all help‐seeking contexts.'195 For this reason, it is unlikely that all men can be engaged through one particular technique or strategy.194, 195 6.2 Engaging hard‐to‐reach women in cancer screening
Indigenous groups (6 studies) There were six studies among the primary studies which targeted indigenous groups.  Native Americans (2 coaching studies) One of the interventions to recruit Native American women to undergo breast screening targeted women in the Denver metropolitan area.197 The program ran from November 1995 to January 1997. Local American Indian women were recruited and trained as lay health advisers (Native Sisters). The Native Sisters undertook a series of activities which reflect the nature of what has been termed 'coaching' in this review. These activities included: 1. contacting American Indian women to increase their awareness of the need for breast cancer screening 2. participating in American Indian community meetings and gatherings to speak about the project and the importance of screening Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 57 3. leading traditional social support circles of American Indian women and discussing breast cancer prevention and early detection 4. mailing NAWA project newsletter and educational materials to women to reinforce the need for screening 5. arranging for transportation to the mammography facility if needed 6. accompanying women to mammography appointments and using follow‐up procedures to provide support.197 The study used an interrupted time series to present the results of the intervention (Level III‐
3 study) and the authors indicated that the intervention had a statistically significant positive impact on the uptake of mammograms.118 However, the data used to determine the impact of the program was for the period February 1995 to August 1996 and does not indicate what happened to participation rates in the final few months of the project. The second USA study aimed to increase the uptake of Pap tests in Native Americans in two tribes in North Carolina. The intervention also involved the use of lay health advisors but focussed more on individual face‐to‐face education, identification of barriers and ways to overcome the barriers (through extended family or church groups), a health risk appraisal for participating women, and watching a videotape with the women in relation to cervical cancer and Pap tests. There was no direct help offered to the women in terms of arranging transport or accompanying them on their visits. The study employed a Solomon Four‐Group design that compared the uptake of Pap tests among those who had participated in pre‐test activities to those who had not. As shown in Table 38, the differences in uptake between the control and intervention groups were statistically significant for both of the Cherokee groups. For the Lumbee groups, only the differences in uptake for the pre‐test groups were statistically significant.124 Table 36 Cherokee Lumbee Percentage of women who reported having a Pap test in the previous 12 months Control 65.1% 69.1% Pre‐test Intervention 71.0%* 74.6%* Control 62.5% 66.8% No pre‐test Intervention 76.0%* 74.0% Notes: * Indicates a statistically significant difference between the intervention and control groups.  Native Hawaiians (1 community intervention study) This community intervention was a prospective cohort study (Level III‐2) aimed to increase the uptake of mammograms and Pap tests. At the heart of this intervention was the Kokua Group led by Native Hawaiian lay health educators. As well as educational material, women in these groups were given vouchers for free mammograms and pap tests; one voucher for themselves and one for a friend. During the project two other activities included a support group for cancer patients and a designated women's clinic at the health centre which was open only to Kokua Group participants. At this clinic women could receive screening from a female health provider with one of the lay health educators in attendance. Compliance with Pap tests in the intervention group communities increased from 59% to 67% and the change was statistically significant. VOL 3– 58 Knowledge translation: a review of strategies to increase participation in cancer screening There were no statistically significant changes in mammography compliance (Table 37). The authors also state that Kokua Group participation was 'associated with increased likelihood of screening participation after controlling for the effects of other variables in the models'. However, close inspection of the results indicate that this conclusion appears to relate to clinical breast examination and mammograms but not to Pap tests.139 Table 37 Changes in screening rates pre‐ and post‐intervention for Native Hawaiian women Intervention group n % Pap tests Baseline 188 59 Follow‐up 213 67 p value* < 0.05 Screening mammograms Baseline 104 60 Follow‐up 105 61 p value* > 0.05 Source: Gotay et al. (2000)139 Control group n % 227 232 63 64 > 00.5 124 126 59 60 > 0.05 Notes: * Based on Chi sq tests  Native Canadians (1 coaching study) This intervention was an historical control study design (Level III‐2) which aimed to increase the uptake of Pap tests in First Nations women in a rural reserve on an island off the north west coast of British Columbia. Community health representatives (CHRs) form part of the staff in the at a community health centre in the reserve. In addition to the mailed invitations and reminder letters to two Pap clinics, CHRs made home visits to the eligible women to schedule appointments at the clinics. The women were telephoned prior to their appointment to remind them of their appointment. The Pap clinics were conducted at the health centre by two female doctors known to the community with the assistance of the CHRs and volunteers. At the clinic educational materials were also provided. The authors concluded that the screening rate rose from 62% to 77%.139 No tests of statistical significance were included in the study.  Native Alaskans (1 procedures study) This was a prospective cohort study (Level III‐2) which aimed to increase the uptake of Pap tests in Native Alaskan women living in remote villages. Community health aides (CHAs) were indigenous village residents who had received training in basic emergency care, acute health problem assessment and well‐child healthcare. The intervention provided CHA training and on‐going supervision for CHAs in the women's health services (history taking, patient education Pap tests, collecting samples for sexually transmitted disease testing, and clinical breast examination).139 The way in which the results were presented for this study make it difficult to determine whether or not the intervention had a statistically significant impact on the uptake of Pap tests. The authors indicate that: Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 59 o the mean proportion of women in the participating villages who had a Pap test increased from 0.44 at baseline to 0.48 at follow‐up and in the comparison villages the proportion decreased from 0.42 to 0.39 o the mean proportion of women aged 45 or older in the participating villages who had a Pap test increased from 0.35 at baseline to 0.39 at follow‐up and in the comparison villages the proportion decreased from 0.39 to 0.28.164  Native Mexicans (1 community intervention study) This was a prospective cohort study (Level III‐2) which aimed to increase the uptake of pap tests in ten villages in with low annual screen rates for cervical cancer. In each village, rural health promoters were recruited to implement the intervention. The local health promoters invited women to attend workshops using door‐to‐door visits, messages through loudspeakers and getting in touch with local groups. There were women only workshops, men only workshops and workshops for men and their partners. There was also an intervention directed at professional healthcare providers as well as a 'dissemination' promotion campaign.134 The authors indicate that a multi‐level analysis (with women nested in villages) indicated that the probability of women having a Pap test increased in the intervention villages compared with the controls and this increase was statistically significant (t(11) = 2.66, p = 0.02).134 Ethno‐specific (primary studies) As shown in Table 38, there were twenty‐seven studies (involving thirty‐two intervention arms) in the data set which specifically targeted ethnic group women. The majority (89%) targeted ethnic groups in America, two tarted ethnic groups in Australia and one targeted ethnic minorities in Wales. Table 38 Target groups in ethno‐specific studies (27 studies, 32 intervention arms) Target groups American: African American American: Asian (1), Cambodian (1), Chinese (2), Korean (1), Vietnamese (3) American: Latino (2), Hispanic (1) American: ethnic minority Australian: Italians (1) Vietnamese (1) UK (Wales): ethnic minorities Total Studies 13 7 3 1 2 1 27 Intervention arms 17 8 3 1 2 1 32  African American women (17 intervention arms) Eleven interventions reported statistically significant positive impacts on the uptake of screening: three counselling interventions,63, 75, 115 three multi‐component interventions,63, 75, 172
two education interventions,94, 106 and one each of coaching,119 community,138 and message‐framing interventions.63 In one study, counselling was associated with a statistically significant decrease in screening uptake.113 In one education90 and one community intervention,141 the results VOL 3– 60 Knowledge translation: a review of strategies to increase participation in cancer screening were unclear. In one coaching intervention,121 one education intervention96 and one message framing intervention75 study there were no statistically significant improvements in uptake.  Other American ethnic groups (12 intervention arms) Eight interventions reported statistically significant positive impacts on the uptake of screening: five coaching interventions,100, 120, 123, 125, 130 one community intervention,135 one education intervention,100 one message framing intervention,64 and one multi‐component intervention.176  Other groups (3 studies) One study reported on a media campaign to target Italians in Australia and did not report statistically significant increases in uptake.196 The other Australian study used invitations and reminders to target Vietnamese women and did not report statistically significant increases in uptake.30 The other study targeted ethnic minorities in Wales using a multi‐
component strategy, and the results were unclear.173 Other hard‐to‐reach groups (primary studies) There was some overlap between studies targeting ethnic groups and those targeting low‐
income and non‐urban women. For example, five of the studies targeting specific ethnic groups targeted low‐income ethnic groups. As it is far from clear whether ethnicity or socioeconomic status is the most important factor in determining participation in screening activities, the studies targeting low‐income ethnic women and non‐urban ethnic women have also been included in this section. As shown in Table 39, there were twenty‐one interventions targeting low income women and seventeen targeting non‐urban women. Twelve studies targeting low‐income women (57%) and 8 targeting non‐urban women (43%) reported increased uptake of screening. In the other interventions the impact was either unclear or there was no statistically significant increase in uptake reported. Table 39 Impact of interventions targeting low‐income and non‐urban women Coaching Community intervention Counselling Financial incentives/disincentives Education Invitations and reminders Message framing Multi‐component Procedures Low income (a) Increased Other uptake 1 130 2 121, 129 1 138 1 117 20, 154
2 20, 32
2 2 73, 79 2 171, 176 1 157 12 (57%) 1 87 23, 25, 197
3 3 23, 80, 81 9 (43%) Non‐urban Increased Other uptake 1 119 2 126, 127 2 133, 134 1 136 1 133 2 94, 106 1 45 1 69 160
1 8 (47%) 1 83 133, 182
2 158, 164
2 9 (53%) Notes: (a) Includes one study targeting squatters in South Africa Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 61 Literature reviews and grey literature Two reviews looked at interventions designed to increase either 'under‐served' populations or women who have historically been defined as under‐users. In a literature review undertaken for the Alberta Cancer Board, Oelke (2002) concluded that most types of strategies have been found to be effective in some situations. Oelke (2002) indicated which strategies can be considered best practice, which are promising and which strategies have been found to be ineffective and, presumably, not recommended (Table 40). The problem with this breakdown is that strategies appear in more than one recommendation category. For example, under breast cancer screening 'Lay health worker' is classified as effective, potentially effective and ineffective. Table 40 Under‐served populations Recommendation Breast cancer screening Best practice (effective) Physician/clinic‐based Lay health worker Promising (potentially effective) Physician/clinic‐based Media Education Lay health worker Multi‐strategy Physician/clinic‐based Education Lay health worker Multi‐strategy Not recommended (ineffective) Cervical cancer screening Education (video) Multi‐strategy Physician/clinic‐based Education Lay health worker Multi‐strategy Physician/Clinic‐based Media Multi‐strategy Both Physician/clinic‐based Lay health worker Multi‐strategy Physician/clinic‐based Lay health worker Multi‐strategy Physician/clinic based Lay health worker Multi‐strategy Source: Oelke (2002)198 Legler et al. published a review in 2002 focusing on which strategies are most effective for women with historically lower use of mammography; namely older women, women with low incomes and non‐Caucasian women.199 The interventions were grouped under the following headings:  individual directed (e.g. one‐on‐one counselling, tailored and untailored letters and reminders, and telephone counselling)  access enhancing (e.g. transportation to appointments, facilitated scheduling, mobile vans, vouchers and reduced cost mammograms)  social network e.g. peer leaders and lay health advisors  community education  mass media  multi‐strategy (combinations of the other interventions). When the odds ratios (OR) for the types of interventions were studied, the most effective appeared to be individual directed strategies generated from within a healthcare setting (e.g. a Health Maintenance Organisation). The OR for the same types of strategies undertaken in a community setting did increase uptake but the result was not statistically significant. Access‐
enhancing strategies and community education also resulted in statistically significant increases VOL 3– 62 Knowledge translation: a review of strategies to increase participation in cancer screening in uptake. Social network strategies and media campaigns did not (Table 40). Legler et al. (2004) state that the most impressive effects were not for single strategy approaches but for combinations of strategies. However, they did not include overall effect sizes or ORs for these strategies. They did state that of 'special note' is the result for nine studies relating to the combination of access enhancing/individual directed strategies (effect size (ES) = 26.9; 95% CI 9.9–43.9) and the result for five studies using the combination access enhancing/system directed strategies (ES = 19.6; 95% CI 8.2–30.6) Table 41 Impact of single strategies on the uptake of mammograms by under‐users Strategy Individual directed – healthcare setting Access enhancing Community education Social network Individual directed – community setting Media campaigns Number of studies 15 14 14 7 13 6 Odds ratio (OR) 2.5 2.3 1.5 1.4 1.3 1.3 95% CI 1.9–3.4 1.7–3.1 1.2–1.9 1.0–2.0 1.0–1.6 1.0–1.8 Source: Legler et al. (2004)199 Summary There were 6 primary studies targeting indigenous women in American, Canada, Alaska, Hawaii, Mexico. All but one of the strategies involved either coaching or community interventions, five targeted Pap tests only and one targeted both mammograms and Pap tests (Table 42). Because of the small number of studies in each group it was difficult to draw conclusions about the results. Table 42 Overview of studies targeting indigenous women Cervical cancer only Coaching Community intervention Procedures Breast & cervical cancer Community Intervention Number of studies Stat sig results 3 1 1 1 1 0 1 For Pap tests only The primary studies focusing on ethnic women are dominated by studies originating in the USA. The most effective strategies (i.e. those that provided evidence of a statistically significant positive impact) appear to be coaching, community interventions, multi‐component interventions and counselling (Figure 4). There is less evidence for the effectiveness of education and message framing, and a small number of studies indicate that media campaigns, invitations and reminders, and procedures did not report positive impacts. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 63 Figure 4 Overview of effectiveness of strategies targeting specific ethnic groups Procedures
1
Invitations and Reminders
1
Media Campaign
2
Message Framing
2
1
Community Intervention
2
1
Counselling
4
Multi-component
4
Education
3
3
Coaching
2
6
0 Positive
1 Impact
2
1
3
4
No/Unclear
5
6 Impact
7
8
9
The primary studies focusing on low‐income women are dominated by studies originating in the USA. It is difficult to discern trends because of the low number of studies in each strategy. However, the overall effectiveness in increasing uptake (as determined by the number of studies reporting statistically significant positive impacts) appears to be quite low. The primary studies focusing on non‐urban women originate in a range of countries including Australia. Once again it is difficult to discern trends in terms of effectiveness because of the small number of studies in each of the strategies, but the proportion of studies reporting positive impacts on screening is less than half. There was only one literature review with usable results which focused on women who were historically low‐users of mammography. For these women the most effective strategies appear to be what have been described as individual directed, access enhancing and community education. However, these groupings included a wide range of strategies which makes it difficult to draw conclusions. For example, individual directed strategies include invitations and reminders, telephone counselling and face‐to‐face counselling. VOL 3– 64 Knowledge translation: a review of strategies to increase participation in cancer screening 6.2 Parallel evidence from other areas Stone et al. (2002) provided a summary of effective strategies for increasing the uptake of adult immunisation.53 Organisational change was the most effective strategy (Table 43), and was defined as changes in the work processes in organisations including redesign of jobs, changes in clinical procedures, or changes in facilities or infrastructure. The organisational change studies included the establishment of a separate clinic, use of planned care visits, use of techniques similar to continuous quality improvement, and designation of specific responsibilities to non medical doctor staff. Organisational change was also included in models of the effectiveness of components to improve the use of screening services for mammograms (OR = 2.47, 95% CI 1.97–3.10), Pap tests (OR = 3.03, 95% CI 2.56–3.58) and colon cancer screening (OR = 17.6, 95% CI 12.3–25.2).53 However, care needs to be exercised in extrapolating the colon cancer results to FOBT as it is not clear whether or not flexible sigmoidoscopy and colonoscopy were included in the review. Among the provider strategies, reminders/prompts were the most effective. Provider financial incentives and feedback were not effective. Among the strategies targeting patients, financial incentives, reminders and education were all effective but the financial incentives were the most effective (Table 43). Table 43 Effectiveness of intervention strategies to improve the uptake of immunisation (29 studies) Strategy Organisational change Provider‐targeted strategies Reminders/Prompts Education Financial incentives Feedback Patient‐targeted strategies Financial incentive Reminders Education Adjusted odds ratio 16.00 95% confidence interval (11.2–22.8) 3.80 3.21 1.26 1.23 (3.31–4.37) (2.24–4.61) (0.83–1.90) (0.96–1.58) 3.42 2.52 1.29 (2.89–4.06) (2.24–2.82) (1.14–1.45) Source: Stone et al. (2002)53 The review by Jepson et al. (2000) contained a number of studies relating to the impact of interventions on screening practices in other areas. However, the authors do point out that the evidence base is dominated by cancer screening and mammography in particular.200  Invitations and reminders (fixed appointment) Osteoporosis screening: One RCT showed that a fixed appointment with the option to change the time was no more effective than a fixed appointment requiring telephone confirmation (RR = 0.93; 95% CI 0.85–1.01).201  Message framing (risk) Cardiovascular disease: An RCT evaluating a questionnaire appraising risk of coronary heart disease found that it was effective in increasing uptake of cholesterol screening for all participants (regardless of risk status) (RR = 2.87; 95% CI 1.86–4.44). It also increased uptake among those who met pre‐defined screening criteria (high risk of coronary heart disease, RR = 5.99; 95% CI 2.96–12.10).202 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 65  Education/coaching/multi‐component Prostate cancer screening: For prostate cancer screening, one controlled trial undertaken in the USA evaluated different types of group teaching: peer educated method (using men of the same age and race as teachers and demonstrators); client navigator method (using a social worker to assist the men in navigating the healthcare system, making their appointment, arranging transportation, and remembering to attend); combination method (peer educator and client navigator methods combined); and traditional method (control). Men who received the client navigator method or the combination method were more likely to attend the free prostate cancer screening than men who received the control or traditional intervention (p < 0.05).203  Community intervention Cholesterol testing: One controlled trial assessed the impact of cholesterol education, targeted media‐intensive screening campaigns with organised screening events at worksites, public areas, churches and special events. The net intervention effect compared to the control was 8.6% (p < 0.001).204  Media campaigns Snyder et al. (2004) propose that it should be easier to convince people to commence a new behaviour than to cease an old behaviour. New screening behaviours may fall in between, since the undesirable behaviour is not yet habitual and may therefore be more susceptible to influence. Table 44 illustrates the effect size of campaigns for different topics and different types of behaviour.205 However, as shown in Table 44, the impact of media campaigns on mammography has been quite small in comparison with, for example, oral health and heart health media campaigns. Table 44 Average campaign effect size by campaign topic Campaign topic Oral health Heart Mammography Number of campaigns 2 4 5 All studies Average effect size 0.13 0.05 0.04 SD 0.00 0.03 0.00 Studies with % change data Number of Average % SD campaigns change 1 0.14 0.00 3 0.04 0.04 5 0.03 0.02 Source: Snyder et al. (2004)205  Financial incentive/disincentive Cholesterol screening: Adding the incentive of being entered into a competition to win a microwave did not increase the uptake of cholesterol screening (RR = 0.98; 95% CI, 0.93–
1.03).206 A study in Denmark reported on the results of a study in which 49‐year‐old men were invited to attend for a health examination of ischaemic heart disease. The participation rate was 66% in the 'free' area and 37% in the fee‐paying area.207  Procedures Diabetes screening: one Level II study assessed the uptake of self‐testing for glycosuria using foil‐wrapped dip‐sticks. Pre‐prandial and post‐prandial tests were compared with a single post‐prandial test. The difference between the two groups in uptake was not statistically significant (78% vs 80%; RR = 1.02; 95% CI, 0.99–1.06).208 VOL 3– 66 Knowledge translation: a review of strategies to increase participation in cancer screening Blood pressure screening: One Level II study evaluated the effectiveness of the use of a health promotion nurse in general practice using a computerised follow‐up system. When the study started in 1986 there were no statistically significance differences in the recording of preventive activities in the control and intervention groups. However, at two years follow‐up there were statistically significant differences between the control and the intervention groups in the recording of screening activities (Table 45). The authors concluded that 'with organisation and resources, over 90% recording and follow‐up of risk factors in general practice can be achieved even in the most adverse inner city conditions'.209 Table 45 Impact of health prevention nurse on recording of screening activities Test recorded Blood pressure recorded within preceding 5 years Hypertensive patients with blood pressure record within preceding year Patients with positive family history of heart attack who had serum cholesterol recorded at any time Women with Pap test recorded within preceding three years (excl. women with hysterectomy) Intervention nurse + doctor 93% Control doctor 73% 1.28 1.23–1.32 97% 69% 1.41 1.24–1.60 40% 28% 2.32 1.67–3.22 76% 49% 1.56 1.44–1.69 RR 95% CI Source: Based on Robson et al. (1989)209 RR: Relative risk Summary The literature relating to parallel evidence was not exhaustive and comes mainly from reviews that also included evidence relating to cancer screening. It appears that reviews to increase screening across a number of areas are dominated by interventions to increase cancer screening and, in particular, the uptake of mammograms. The strategies adopted in other areas are similar to those used to increase participation in cancer screening and the results are quite similar in terms of the effectiveness of the intervention. Of particular interest are:  the effectiveness of coaching (patient navigation) for increasing participation in prostate cancer screening  the negative impact of charging a fee on men's participate in a health examination of ischaemic heart disease  the positive impact of a health prevention nurse in general practice on the recording of (and presumably participation in) screening activities including Pap tests. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 67 7. DISCUSSION AND CONCLUSIONS Preparation of this review presented a number of difficulties. Interventions vary along a number of dimensions and this makes them extremely difficult to classify. In order to present a coherent document, the classification developed by Jepson et al. (2000) was adapted to the needs of the review.200 Data from the systematic reviews are presented in line with the classification, although there is not always a neat fit between the groupings of strategies in the systematic reviews and the classification used here. Where it appears that differences in classification schemes may have affected the conclusions drawn in the reviews, this has been pointed out. Another difficulty was deciding on what constituted 'effectiveness'. For the purposes of this review, strategies were deemed effective if they demonstrated statistically significant impacts on the uptake of screening. This means that some studies that may have been just outside significance have been classified as not being effective. The annotated bibliography included in Appendix B includes the results for each of the 154 primary studies, and can be consulted for a more detailed description of the results of each study. There are a number of studies where the impact on screening has been classified as 'unclear'. This occurred for a number of reasons: i) no significance tests were included even when there was a comparison group; ii) the study did not include a comparison group; or iii) the results were presented in a way that made it difficult to be sure of the impact on screening rates. One of the strengths of the review is that is has included studies from all levels of evidence. In contrast, many of the literature reviews had a much narrower inclusion criteria and this means that the results in the literature reviews are based on fewer studies. Sometimes the authors appear to have based their conclusions on the results of one or two studies. Where there was doubt about the conclusions in the reviews because of the small number of studies, this has been pointed out. 7.1 Overview of primary studies There are a number of issues related to the primary studies focusing on cancer screening which need to be kept in mind when considering the implications of the findings.  Number of studies on which the conclusions are based There were 154 primary intervention studies included in this analysis, and these studies included 188 interventions arms. Over two‐thirds of the screening‐population strategies (69%) involved invitations and reminders, message framing, education, multi‐component interventions and counselling. There were few studies (< 5) related to mass media campaigns, worksite interventions and financial (dis)incentives published in the ten‐year period that included some indication of the impact on screening rates. There were also few studies focusing on strategies aimed at the screening population and/or providers (Table 46).  Healthcare system to which the studies relate Most of the studies relating to message framing, education, multi‐component interventions, and community interventions originated in the USA (Figure 5). VOL 3– 68 Knowledge translation: a review of strategies to increase participation in cancer screening Table 46 Number of studies investigating each of the strategies (154 studies, 188 intervention arms) Target group/Strategy Screening‐population strategies Invitations and reminders Message framing Education Multi‐component interventions Counselling Coaching Procedures Community interventions Mass media campaigns Worksite interventions Financial incentives/disincentives Studies Number Provider strategies Provider + screening‐population strategies % 32 27 24 21 17 15 15 13 4 4 3 175 9 4 13 18 15 14 12 10 9 9 7 2 2 2 100 69 31 100  Type of screening test targeted Over half of the message framing and multi‐component intervention studies, all but one of the counselling studies and all of the financial (dis)incentive studies related to the uptake of mammograms. FOBT were well‐represented in invitations and reminders, education, procedures and provider strategies. Pap tests were well‐represented in the coaching, procedures and provider strategies (Figure 6). Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 69 Figure 5 Overview of country from which primary intervention studies originated (154 studies, 183 intervention arms) Provider + Screening Pop Strategies
4
Provider Strategies
9
Screening Population Strategies
Financial Incentives/ Disincentives
2 1
Worksite Interventions
Mass Media Campaigns
2
2
3
1
Community Interventions
11
11
Procedures
2
10
3
Coaching
3
12
Counselling
1
16
Multi-Component Interventions
2 1
18
Education
20
Message Framing
20
Invitations and Reminders
5
USA
2
5
5
10
17
0
1
3
10
15
OTHER
20
25
30
35
AUSTRALIA
VOL 3– 70 Knowledge translation: a review of strategies to increase participation in cancer screening Figure 6 Overview of type of cancer screening test targeted in primary intervention studies Screening Population Strategies
Provider + Screening Pop Strategies
4
Provider Strategies
3
Financial Incentives/ Disincentives
3
Worksite Interventions
3
Mass Media Campaigns
3
3
1
1 2 1
Community Interventions
3
Procedures
3
Coaching
1
4
5
6
6
1
6
5
Counselling
3
16
Multi-Component Interventions
1
2
11
Education
3
8
Message Framing
4
4
10
5
2
10
5
15
0
6
3
16
Invitations and Reminders
3
10
Mammograms
15
Pap Tests
20
25
FOBT
2
30
35
Multi-Test
7.2 Summary of results for cancer screening strategies Bearing in mind the caveats around the peer‐reviewed literature, a summary of the findings for each of the strategies targeting screening populations is shown in Table 47. There is scant evidence in the peer‐reviewed primary studies for the effectiveness of the strategies targeting providers. The literature reviews appear to indicate that prompts may have some effect in terms of increasing uptake for mammograms and Pap tests. However, the evidence for the impact of prompts on FOBT is not clear. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 71 Table 47 Summary of the findings for the strategies Strategy Invitations and reminders Message framing Education Counselling Coaching Community interventions Findings In the primary studies, different methods of issuing the invitations and reminders were included in the studies (e.g. mail, telephone, face‐to‐face, multi‐method). Just under half of the mailed invitations and reminders, and nearly all of the telephone invitations and reminders reported a statistically significant positive impact. The reviews/grey literature emphasise that the impact of invitations and reminders on uptake of screening will vary by method of administration, nature of the test, and the characteristics of the target group. The overall conclusion appears to be that invitations and reminders do increase uptake, at least among mid‐high socioeconomic groups. Telephone calls, and letters plus telephone calls appear to have a bigger impact than mailed invitations and reminders. The evidence for face‐to‐face is less clear. Fixed appointments in mailed invitations appear to be more effective than open appointments, mammograms and Pap tests. In the primary studies, endorsement by a doctor seems to be particularly effective, with six of the seven studies that used this strategy reporting increased uptake. The evidence for other message framing strategies is mixed. The conclusions from the reviews/grey literature are based on a narrow range of interventions and a fewer number of studies. The conclusions from this literature are: i) there is little evidence for GP endorsement; ii) framing in terms of risk appears to increase uptake; and iii) there is little evidence of benefit in terms of loss/gain framing. Reaching conclusions based on the primary studies is hampered by the small number of studies for most of the different types of education strategies. However, it appears that interventions using print material generally do not report statistically significant positive impacts on uptake. The conclusions reached in the literature reviews/grey literature are also based on a small number of studies, and the conclusions are contradictory. The review with the most positive conclusions for the use of print material included 'letters' in this category. If these letters are, in fact, invitations then this overlaps with the invitations and reminders strategy and may account for the positive conclusion. In the primary studies, the majority of the counselling interventions involved telephone counselling. All the face‐to‐face interventions and just over seventy per cent of the telephone interventions reported an increased uptake. In contrast, the literature reviews/grey literature provided limited evidence for the effectiveness of counselling. Half of the interventions in the primary studies increased the uptake of screening. The conclusions from the literature reviews/grey literature were that this form of intervention increases the uptake of mammography but conclusions were less consistent in relation to Pap testing. All but two of the interventions were targeted at hard‐to‐reach groups (i.e. specific ethnic groups, non‐urban and low‐income groups). Nearly two‐thirds of the interventions increased uptake in the hard‐to‐reach groups but none reported an increase in uptake in the general population. The only review which considered this type of intervention concluded that there was some evidence for the effectiveness of community interventions but the quality of the evidence was poor. VOL 3– 72 Knowledge translation: a review of strategies to increase participation in cancer screening Strategy Mass media campaigns Findings Overall, it appears that media campaigns may have a short‐term impact on screening rates but there is insufficient evidence to conclude that one‐off campaigns will, by themselves, increase screening rates in the longer term. Worksite Overall it appears that these interventions are promising, but the evidence is interventions limited as to their effectiveness. Although encouraging, results for FOBT uptake in Taiwanese workplaces may not easily translate to western countries because of cultural differences. Financial incentives/ Based on limited evidence it appears that: i) making screening available free of disincentives charge does not necessarily increase screening rates across all ethnic groups; and ii) offering financial incentives to low‐income women increases participation. Procedures Overall two‐thirds of the interventions in the peer‐reviewed literature increased uptake, and all the interventions aimed at increasing the uptake of FOBT did so. The literature reviews/grey literature were generally positive about the potential for opportunistic screening, periodic health examinations, and the use of mobile vans. The evidence for changing dietary restrictions and methods of screening for FOBT were mixed. Multi‐component Overall, eleven of the twenty‐one interventions in the peer‐reviewed literature interventions improved uptake. Ten of the interventions were targeted hard‐to‐reach groups and six increased uptake. The literature reviews/grey literature contained some evidence that some multi‐component interventions improved uptake of screening rates, but it was difficult to discern patterns or trends. Provider strategies There were few primary studies relating to provider strategies, but the overall conclusion based on these and the literature reviews was that there is some evidence that provider prompts (computer generated and on patients' files or charts) are effective in increasing uptake of mammograms and Pap tests. However the impact on FOBT is unclear. Provider + screening There were few primary studies reporting on these strategies and only one population strategies produced a statistically significant positive impact on uptake. 7.3 Summary of results for hard‐to‐reach groups For the purposes of this report the definition 'hard‐to‐reach' as has generally been in terms of socio‐demographic factors: ethnicity, place of residence, income, gender. However, these categories are not discrete and there is considerable overlap. The analysis for hard‐to‐reach groups looked at the evidence in relation to: i) engaging men in FOBT screening; ii) engaging women from specific ethnic groups, low‐income women and women living in non‐urban areas; (iii) evidence for success in engaging these groups in other screening activities. A summary of the findings in relation to hard‐to‐reach groups are shown in Table 48. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 73 Table 48 Overview of findings for strategies targeting hard‐to‐reach groups Focus Engaging men Findings Although there were individual studies that demonstrated the effectiveness of a particular strategy in improving FOBT participation rates amongst men, there is currently no clear evidence about which is the most effective strategy, especially within the Australian context. It is important to note, as Galadas et al. (2005) state: 'not all men are the same, nor does it make sense to assume that individual men behave similarly in all help‐seeking contexts'.195 Engaging hard‐to‐reach women in cancer screening Parallel evidence from other health behaviours For this reason, it is unlikely that all men can be engaged through one particular technique or strategy.194, 195 Primary studies: (i) Indigenous groups: The small number of studies make it difficult to discern trends in these data. However, the strategies that were tried included coaching, community interventions and procedures. All were aimed at Pap tests and/or mammograms. (i) Ethnic groups: The most effective strategies appear to be coaching, community interventions, multi‐component interventions and counselling. There is less evidence for the effectiveness of education and message framing, and a small number of studies indicate that media campaigns, invitations and reminders and procedures did not report positive impacts. (ii) Low income: It was difficult to discern trends because of the low number of studies in each strategy. However, the overall effectiveness (as judged by the number of studies reporting a statistically significant increase in uptake) appears to be quite low (57%). Strategies with studies only reporting increased uptake included: counseling, financial (dis)incentives, procedures and multi‐component interventions. The only strategy with no studies reporting increased uptake was education. (iii) Non‐urban: It was difficult to discern trends in terms of effectiveness because of the small number of studies in each of the strategies, but the proportion of studies reporting positive impacts on screening was less than half (47%). Strategies with studies only reporting increased uptake (i.e. strategies with no studies reporting unclear, or no increase results) included: education and invitations and reminders. The strategies with no studies reporting increased uptake were counselling and multi‐component. In the literature reviews/grey literature. There was only one literature review with usable results which focused on women who were historically low‐users of mammography. For these women the most effective strategies appeared to be 'individual directed', 'access enhancing' and 'community education'. However, these groupings included a wide range of strategies which limited the usefulness of the conclusions. The literature relating to parallel evidence was not exhaustive and came mainly from reviews that also included evidence relating to cancer screening. It appears that reviews to increase screening across a number of areas were dominated by interventions to increase cancer screening and, in particular, the uptake of mammograms. The strategies adopted in other areas were similar to those used to increase VOL 3– 74 Knowledge translation: a review of strategies to increase participation in cancer screening participation in cancer screening and the results were quite similar in terms of the effectiveness of the intervention. Of particular interest are: i) the effectiveness of coaching (patient navigation) for increasing participation in prostate cancer screening; ii) the negative impact of charging a fee for men's participation in a health examination of ischaemic heart disease; iii) the positive impact of a health prevention nurse in general practice who followed‐up patients on the recording of (and presumably participation in) screening activities including Pap tests; and iv) the positive impact of organisational change on immunisation. 7.4 Conclusions Many countries appear to be facing declining screening rates, and the reasons for this are not clear. So the search is on for interventions or ways to increase participation, particularly among hard‐to‐reach groups. Many of the interventions presented in the review are short‐term. If short‐term interventions do succeed in increasing the uptake of cancer screening, then methods for sustaining the successful interventions, or maintaining the increased uptake, need to considered.210, 211 Bearing this and the other caveats in relation to the data in mind, Table 50 provides an assessment of the strategies for increasing cancer screening rates for the general population and for hard‐to‐reach groups. The assessments in 49 and Table 50 have taken into account the findings from the other areas and the section of hard‐to‐reach groups. In reading these tables it needs to be kept in mind that the number of studies targeting hard‐to‐
reach groups is much lower than the overall number of studies. Therefore, the assessments in Table 50 are somewhat more tentative. Also in this Table, 'X' generally means there is insufficient evidence to make a judgment rather than the evidence indicating that the strategy is ineffective. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 75 Table 49 General population: assessment of the strategies Invitations and reminders Mailed Telephone Face‐to‐face Message framing Tailoring GP endorsement Risk Loss/gain Other framing Education Print material Interactive (group/one‐on‐one) Videos Multi‐method Counselling Telephone Face‐to‐face Coaching Community intervention Mass media campaigns Worksite interventions Financial incentives/disincentives Procedures Multi‐strategy interventions Assessment ? Issues  X Inclusion of appointment. Decreasing returns on multiple reminders, could be increased by multi‐method strategy with telephone reminders for those who do not respond. X  X X X X ? ? ? Not as a stand‐alone strategy The combination of methods needs to be clarified   ? X ? ?  ? ? One‐off campaigns have short‐term impact The strategies differ by type of cancer and there is need for clarity around which strategies will work for which screening test. More clarity needed around which are the most effective combinations of strategies. Provider strategies  Prompts (computer or chart) Education X Audit and feedback X Not as a stand‐alone strategy Provider + screening population X Notes:  Sufficient evidence to suggest that this strategy has a direct impact on screening rates and is worth pursuing ? Evidence for the impact on screening rates is inconclusive and needs further investigation ? Sufficient evidence to suggest that this strategy has a direct, positive impact on screening rates and is worth pursuing but there are some issues that need further investigation X Insufficient evidence to suggest that this strategy has a direct impact on screening rates VOL 3– 76 Knowledge translation: a review of strategies to increase participation in cancer screening Table 50 Hard‐to‐reach groups (indigenous, ethnic, low income, non‐urban): assessment of the strategies Assessment Indigenous Ethnic Low income Non‐urban Invitations and reminders  Mailed X ? ? Telephone X X X X Face‐to‐face X X X X Message framing Tailoring X ? X X GP endorsement X X ? X Risk X X X X Loss/Gain X X ? X Other framing X X X Education Print material X X X X  Interactive (group/one‐on‐one) X X X Videos X ? X X  Multi‐strategy X X X Counselling  Telephone X ? X  Face‐to‐face X ? X  Coaching ? ? ? Community intervention ? X ? X Mass media campaigns X X X X Worksite interventions X X X X  Financial incentives/disincentives X ? X  Procedures X X ?   Multi‐component interventions X X Provider strategies Prompts (computer or chart) X X X X Education X X X X Audit and feedback X X X X Provider + screening population Notes:  Sufficient evidence to suggest that this strategy has a direct, positive impact on screening rates and is worth pursuing in the Australian context ? Evidence for the impact on screening rates is inconclusive and needs further investigation in the Australian context ? Sufficient evidence to suggest that this strategy has a direct, positive impact on screening rates and is worth pursuing in the Australian context but there are some issues that need further investigation. X Insufficient evidence to suggest that this strategy has a direct impact on screening rates Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 77 Table 51 Opportunities to combine interventions across cancer screening programs Recommendation Reminder letters Invitation letters Mass media campaigns Targeted strategies Training of aboriginal health workers Pilot self‐sampling HPV test Financial incentives for ATSI clients Workplace strategies Relevance to individual screening program Of special relevance to cervical screening, and BowelScreen as BreastScreen already send up to 3 reminders. Targeted approach e.g. geographic, with locally framed messages, could be used across programs. Already sent by BreastScreen and BowelScreen. Most relevant to cervical screening but potential to combine with BreastScreen. Potential to target campaigns for cancer screening in general. Need to balance with the need for different messages and different service providers. Given requirement for intensive resource allocation, a combined approach should be considered, particularly in areas shown to have low screening rates across the board. Cervical screening but additional education about mammography and FOBT could be reinforced. Cervical screening Applicable across programs Applicable across programs HPV: Human papilloma virus; ATSI: Aboriginal and Torres Strait Islander; FOBT: Faecal occult blood testing VOL 3– 78 Knowledge translation: a review of strategies to increase participation in cancer screening REFERENCES 1. Australian Institute of Health and Welfare. 2008. BreastScreen Australia monitoring report 2004‐
2005. Rep. Cancer series no. 42, Cat no. 37, AIHW, Canberra 2. Department of Health and Ageing. 2009. BreastScreen Australia Evaluation MBS Mammography Analysis Project. Final report, Australian Government Department of Health and Ageing, Canberra 3. Schopper D, de Wolf C. 2007. Breast cancer screening by mammography: International evidence and the situation in Switzerland, Bern, Switzerland 4. Mitchell H, Medley G. 1987. Age trends in Pap smear usage, 1971‐1986. Community Health Studies 11: 183‐5 5. Straton JA, Holman CD, Edwards BM. 1993. Cervical cancer screening in Western Australia in 1992: progress since 1983. Medical Journal Australia 159: 657‐61 6. Condron J, Fernbach M. 2001. Young women's cervical screening: A literature review, Centre for Behavioural Research in Cancer, Melbourne, Victoria 7. Australian Institute of Health and Welfare. 2008. Cervical screening in Australia 2005‐2006. Rep. Cancer series no. 41, Cat no. CAN 36, AIHW, Canberra 8. Mullins R. 2005. Evaluation of a strategy to increase cervical screening through the use of personalised letters, The Cancer Council Victoria, Melbourne, Victoria 9. Australian Institute of Health and Welfare. 2009. Cervical screening in Australia 2006‐2007. Rep. Cancer series no. 47, Cat no. CAN 43, Australian Institute of Health and Welfare, Canberra 10. Linos A, Riza E. 2000. Comparisons of cervical cancer screening programmes in the European Union. European Journal of Cancer 36: 2260‐5 11. Ahti A, von Karsa L, Aasmaa A, Fender M, Patnick J, et al. 2009. Cervical cancer screening policies and coverage in Europe. European journal of cancer (Oxford, England : 1990) 45: 2649‐58 12. Ahti A, Guglielmo R. 2009. Description of the national situation of cervical cancer screening in the member states of the European Union. European Journal of Cancer (Oxford, England : 1990) 45: 2685‐708 13. Australian Institute of Health and Welfare, Australian Government Department of Health and Ageing. 2008. National Bowel Cancer Screening Program monitoring report 2008. Rep. Cancer series 44, Cat no. 40, AIHW, Canberra 14. Bastos J, Peleteiro B, Gouveia J, Coleman M, Lunet N. 2009. The state of the art of cancer control in 30 European countries in 2008. Int J Cancer 126: 2700‐15 15. UK Colorectal Cancer Screening Pilot Project Group. 2004. Results of the first round of a demonstration pilot of screening for colorectal cancer in the United Kingdom. Bmj 329: 133 16. Department of Health. 1998. First Report of the National Screening Committee, Department of Health, London 17. National Health and Medical Research Council. 2008. NHMRC additional levels of evidence and grades for recommendations for developers of guidelines, NHMRC, Canberra 18. Jepson R, Clegg A, Forbes C, Lewis R, Sowden A, Kleijnen J. 2000. The determinants of screening uptake and interventions for increasing uptake: A systematic review. Health Technology Assessment 4 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 79 19. Jean S, Major D, Rochette L, Brisson J. 2005. Screening mammography participation and invitational strategy: the Quebec Breast Cancer Screening Program, 1998‐2000. Chronic Dis Can 26: 52‐8 20. Slater JS, Henly GA, Ha CN, Malone ME, Nyman JA, et al. 2005. Effect of direct mail as a population‐based strategy to increase mammography use among low‐income underinsured women ages 40 to 64 years. Cancer Epidemiology Biomarkers and Prevention 14: 2346‐52 21. Mayer JA, Lewis EC, Slymen DJ, Dullum J, Kurata H, et al. 2000. Patient reminder letters to promote annual mammograms: a randomized controlled trial. Prev Med 31: 315‐22 22. Barr JK, Franks AL, Lee NC, Antonucci DM, Rifkind S, Schachter M. 2001. A randomized intervention to improve ongoing participation in mammography. American Journal of Managed Care 7: 887‐94 23. McCaul KD, Wold KS. 2002. The effects of mailed reminders and tailored messages on mammography screening. Journal of Community Health 27: 181‐90 24. Rutter DR, Steadman L, Quine L. 2006. An implementation intentions intervention to increase uptake of mammography. Annals of Behavioral Medicine 32: 127‐34 25. Simon MS, Gimotty PA, Moncrease A, Dews P, Burack RC. 2001. The effect of patient reminders on the use of screening mammography in an urban health department primary care setting. Breast Cancer Res Treat 65: 63‐70 26. Hayes C, O'Herlihy B, Hynes M, Johnson Z. 1999. The impact of reminder letters on attendance for breast cancer screening. Irish Journal of Medical Science 168: 29‐32 27. de Jonge E, Cloes E, de Beeck LO, Adriaens B, Lousbergh D, et al. 2008. A quasi‐randomized trial on the effectiveness of an invitation letter to improve participation in a setting of opportunistic screening for cervical cancer. European Journal of Cancer Prevention 17: 238‐42 28. Morrell S, Taylor R, Zeckendorf S, Niciak A, Wain G, Ross J. 2005. How much does a reminder letter increase cervical screening among under‐screened women in NSW? Australian and New Zealand Journal of Public Health 29: 78‐84 29. Eaker S, Adami HO, Granath F, Wilander E, Sparen P. 2004. A large population‐based randomized controlled trial to increase attendance at screening for cervical cancer. Cancer Epidemiology Biomarkers & Prevention 13: 346‐54 30. Del Mar C, Glasziou P, Adkins P, Hua T, Brown M. 1998. Do personalised letters in Vietnamese increase cervical cancer screening among Vietnamese women? Australian and New Zealand Journal of Public Health 22: 824‐5 31. Reeves MJ, Remington PL. 1999. Use of patient reminder letters to promote cancer screening services in women: A population‐based study in Wisconsin. Preventive Medicine 29: 202‐9 32. Burack RC, Gimotty PA, Simon M, Moncrease A, Dews P. 2003. The effect of adding Pap smear information to a mammography reminder system in an HMO: Results of randomized controlled trial. Preventive Medicine 36: 547‐54 33. Sequist TD, Zaslavsky AM, Marshall R, Fletcher RH, Ayanian JZ. 2009. Patient and physician reminders to promote colorectal cancer screening: A randomized controlled trial. Archives of Internal Medicine 169: 364‐71 34. Myers RE, Sifri R, Hyslop T, Rosenthal M, Vernon SW, et al. 2007. A randomized controlled trial of the impact of targeted and tailored interventions on colorectal cancer screening. Cancer 110: 2083‐91 35. Crotta S, Dagnes B, Grazzini G, Senore C. 2008. Strategies to enhance participation in a screening program for colorectal cancer in a mountain community. Gastroenterology 134: A483‐A VOL 3– 80 Knowledge translation: a review of strategies to increase participation in cancer screening 36. Grazzini G, Ciatto S, Cislaghi C, Castiglione G, Falcone M, et al. 2008. Cost evaluation in a colorectal cancer screening programme by faecal occult blood test in the District of Florence. Journal of Medical Screening 15: 175‐81 37. Millard FB. 2006. Bowel cancer screening in Australia. Aust Fam Physician 35: 188‐91 38. Church TR, Yeazel MW, Jones RM, Kochevar LK, Watt GD, et al. 2004. A randomized trial of direct mailing of fecal occult blood tests to increase colorectal cancer screening. J Natl Cancer Inst 96: 770‐80 39. Cole SR, Smith A, Wilson C, Turnbull D, Esterman A, Young GP. 2007. An advance notification letter increases participation in colorectal cancer screening. J Med Screen 14: 73‐5 40. Fishman P, Taplin S, Meyer D, Barlow W. 2000. Cost‐effectiveness of strategies to enhance mammography use. Eff Clin Pract 3: 213‐20 41. Page A, Morrell S, Chiu C, Taylor R, Tewson R. 2006. Recruitment to mammography screening: A randomised trial and meta‐analysis of invitation letters and telephone calls. Australian and New Zealand Journal of Public Health 30: 111‐8 42. Taplin SH, Barlow WE, Ludman E, MacLehos R, Meyer DM, et al. 2000. Testing reminder and motivational telephone calls to increase screening mammography: a randomized study. Journal of the National Cancer Institute 92: 233‐42 43. Potter MB, Namvargolian Y, Hwang J, Walsh JME. 2009. Improving colorectal cancer screening: A partnership between primary care practices and the American Cancer Society. Journal of Cancer Education 24: 22‐7 44. Segura JM, Castells X, Casamitjana M, Macia F, Porta M, Katz SJ. 2001. A randomized controlled trial comparing three invitation strategies in a breast cancer screening program. Preventive Medicine 33: 325‐32 45. Shenson D, Cassarino L, DiMartino D, Marantz P, Bolen J, et al. 2001. Improving access to mammograms through community‐based influenza clinics: A quasi‐experimental study. American Journal of Preventive Medicine 20: 97‐102 46. Chalapati W, Chumworathayi B. 2007. Can a home‐visit invitation increase Pap smear screening in Samliem, Khon Kaen, Thailand? Asian Pac J Cancer Prev 8: 119‐23 47. Courtier R, Casamitjana M, Macia F, Panades A, Castells X, et al. 2002. Participation in a colorectal cancer screening programme: influence of the method of contacting the target population. European Journal of Cancer Prevention 11: 209‐13 48. Feldstein AC, Perrin N, Rosales AG, Schneider J, Rix MM, et al. 2009. Effect of a Multimodal Reminder Program on Repeat Mammogram Screening. American Journal of Preventive Medicine 37: 94‐101 49. Hermens RP, Tacken MA, Hulscher ME, Braspenning JC, Grol RP. 2000. Attendance to cervical cancer screening in family practices in The Netherlands. Preventive Medicine 30: 35‐42 50. Tseng DS, Cox E, Plane MB, Hla KM. 2001. Efficacy of patient letter reminders on cervical cancer screening: a meta‐analysis. J Gen Intern Med 16: 563‐8 51. Yabroff KR, Mandelblatt JS. 1999. Interventions targeted toward patients to increase mammography use. Cancer Epidemiology Biomarkers and Prevention 8: 749‐57 52. Black ME, Yamada J, Mann V. 2002. A systematic literature review of the effectiveness of community‐based strategies to increase cervical cancer screening. Can J Public Health 93: 386‐93 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 81 53. Stone EG, Morton SC, Hulscher ME, Maglione MA, Roth EA, et al. 2002. Interventions that increase use of adult immunization and cancer screening services: A meta‐analysis. Annals of Internal Medicine 136: 641‐51 54. Lee E, Robeson P. 2007. 12th September 2009 http://health‐evidence.ca/articles/show/16134 55. Bonfill CS, Marzo CM, Pladevall VM, Marti J, Emparanza JI. 2009. Strategies for increasing the participation of women in community breast cancer screening (Review). Cochrane database of systematic reviews 56. Bonfill CS, Marzo CM, Pladevall VM, Marti J, Emparanza JI. 2001. Strategies for increasing women participation in community breast cancer screening. Cochrane database of systematic reviews 57. Sin JP, St Leger AS. 1999. Interventions to increase breast screening uptake: Do they make any difference? Journal of Medical Screening 6: 170‐81 58. Ellis P, Ciliska D, Sussman J, Robinson P, Armour T, et al. 2005. A systematic review of studies evaluating diffusion and dissemination of selected cancer control interventions. Health Psychology 24: 488‐500 59. Forbes C, Jepson R, Martin‐Hirsch P. 2002. Interventions targeted at women to encourage the uptake of cervical screening. Cochrane Database of Systematic Reviews Issue 3: Art No CD002834. DOI: 10.1002/14651858.CD002834 60. Baron RC, Rimer BK, Coates RJ, Kerner J, Kalra GP, et al. 2008. Client‐directed interventions to increase community access to breast, cervical, and colorectal cancer screening: a systematic review. Am J Prev Med 35: S56‐66 61. Denhaerynck K, Lesaffre E, Baele J, Cortebeeck K, Van Overstraete E, Buntinx F. 2003. Mammography screening attendance ‐ Meta‐analysis of the effect of direct‐contact invitation. American Journal of Preventive Medicine 25: 195‐203 62. Anderson J. 2004. Literature Review of Outpatient Cervical Screening, The Cancer Council, Melbourne, Victoria 63. Champion VL, Skinner CS, Menon U, Seshadri R, Anzalone DC, Rawl SM. 2002. Comparisons of tailored mammography interventions at two months postintervention. Ann Behav Med 24: 211‐8 64. Tu SP, Taplin SH, Barlow WE, Boyko EJ. 1999. Breast cancer screening by Asian‐American women in a managed care environment. American Journal of Preventive Medicine 17: 55‐61 65. Williams‐Piehota P, Pizarro J, Schneider TR, Mowad L, Salovey P. 2005. Matching health messages to monitor‐blunter coping styles to motivate screening mammography. Health Psychology 24: 58‐67 66. Rakowski W, Lipkus IM, Clark MA, Rimer BK, Ehrich B, et al. 2003. Reminder letter, tailored stepped‐care, and self‐choice comparison for repeat mammography. American Journal of Preventive Medicine 25: 308‐14 67. Rimer BK, Halabi S, Skinner CS, Kaplan EB, Crawford Y, et al. 2001. The short‐term impact of tailored mammography decision‐making interventions. Patient Education and Counseling 43: 269‐85 68. Williams‐Piehota P, Schneider TR, Pizarro J, Mowad L, Salovey P. 2003. Matching Health Messages to Information‐Processing Styles: Need for Cognition and Mammography Utilization. Health Communication 15: 375‐92 69. Sheeran P, Orbell S. 2000. Using implementation intentions to increase attendance for cervical cancer screening. Health Psychology 19: 283‐9 VOL 3– 82 Knowledge translation: a review of strategies to increase participation in cancer screening 70. Marcus AC, Mason M, Wolfe P, Rimer BK, Lipkus I, et al. 2005. The efficacy of tailored print materials in promoting colorectal cancer screening: Results from a randomized trial involving callers to the National Cancer Institute's Cancer Information Service. Journal of Health Communication 10: 83‐104 71. Rimer BK, Conaway M, Lyna P, Glassman B, Yarnall KSH, et al. 1999. The impact of tailored interventions on a community health center population. Patient Education and Counseling 37: 125‐40 72. Valanis BG, Glasgow RE, Mullooly J, Vogt TM, Whitlock EP, et al. 2002. Screening HMO women overdue for both mammograms and Pap tests. Preventive Medicine 34: 40‐50 73. Van Harrison R, Janz NK, Wolfe RA, Tedeschi PJ, Chernew M, et al. 2003. Personalized targeted mailing increases mammography among long‐term noncompliant medicare beneficiaries: a randomized trial. Med Care 41: 375‐85 74. Saywell Jr RM, Champion VL, Zollinger TW, Maraj M, Skinner CS, et al. 2003. The cost effectiveness of 5 interventions to increase mammography adherence in a managed care population. American Journal of Managed Care 9: 33‐44 75. Champion V, Maraj M, Hui S, Perkins AJ, Tierney W, et al. 2003. Comparison of tailored interventions to increase mammography screening in nonadherent older women. Preventive Medicine 36: 150‐8 76. Lipkus IM, Green LG, Marcus A. 2003. Manipulating Perceptions of Colorectal Cancer Threat: Implications for Screening Intentions and Behaviors. Journal of Health Communication 8: 213‐28 77. Lipkus IM, Skinner CS, Dement J, Pompeii L, Moser B, et al. 2005. Increasing colorectal cancer screening among individuals in the carpentry trade: Test of risk communication interventions. Preventive Medicine 40: 489‐501 78. Ore L, Hagoel L, Lavi I, Rennert G. 2001. Screening with faecal occult blood test (FOBT) for colorectal cancer: assessment of two methods that attempt to improve compliance. European Journal of Cancer Prevention 10: 251‐6 79. Abood DA, Black DR, Coster DC. 2005. Loss‐framed minimal intervention increases mammography use. Women's Health Issues 15: 258‐64 80. Schneider TR, Salovey P, Apanovitch AM, Pizarro J, McCarthy D, et al. 2001. The effects of message framing and ethnic targeting on mammography use among low‐income women. Health Psychol 20: 256‐66 81. Consedine NS, Horton D, Magai C, Kukafka R. 2007. Breast screening in response to gain, loss, and empowerment framed messages among diverse, low‐income women. Journal of Health Care for the Poor and Underserved 18: 550‐66 82. Boling W, Laufman L, Lynch GR, Weinberg AD. 2005. Increasing mammography screening through inpatient education. Journal of Cancer Education 20: 247‐50 83. Stein K, Lewendon G, Jenkins R, Davis C. 2005. Improving uptake of cervical cancer screening in women with prolonged history of non‐attendance for screening: a randomized trial of enhanced invitation methods. Journal of Medical Screening 12: 185‐9 84. Edwards A, Unigwe S, Elwyn G, Hood K. 2003. Effects of communicating individual risks in screening programmes: Cochrane systematic review. British Medical Journal 327: 703‐7 85. Michielutte R, Sharp PC, Foley KL, Cunningham LE, Spangler JG, et al. 2005. Intervention to increase screening mammography among women 65 and older. Health Education Research 20: 149‐62 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 83 86. Lipkus IM, Rimer BK, Halabi S, Strigo TS. 2000. Can tailored interventions increase mammography use among HMO women? American Journal of Preventive Medicine 18: 1‐10 87. Risi L, Bindman JP, Campbell OMR, Imrie J, Everett K, et al. 2004. Media interventions to increase cervical screening uptake in South Africa: An evaluation study of effectiveness. Health Education Research 19: 457‐68 88. Trevena LJ, Irwig L, Barratt A. 2008. Randomized trial of a self‐administered decision aid for colorectal cancer screening. J Med Screen 15: 76‐82 89. Hurdle DE. 2007. Breast cancer prevention with older women: A gender‐focused intervention study. Health Care for Women International 28: 872‐87 90. Kidder B. 2008. P.O.W. (Protect Our Women): Results of a breast cancer prevention project targeted to older African‐American women. Social Work in Health Care 47: 60‐72 91. Twinn S. 2001. The evaluation of the effectiveness of health education interventions in clinical practice: A continuing methodological challenge. Journal of Advanced Nursing 34: 230‐7 92. Resnick B. 2003. Health promotion practices of older adults: Testing an individualized approach. Journal of Clinical Nursing 12: 46‐55 93. Navarro AM, Raman R, McNicholas LJ, Loza O. 2007. Diffusion of cancer education information through a Latino community health advisor program. Preventive Medicine 45: 135‐8 94. Powe BD. 2002. Promoting fecal occult blood testing in rural African American women. Cancer Practice 10: 139‐46 95. Powe BD, Ntekop E, Barron M. 2004. An intervention study to increase colorectal cancer knowledge and screening among community elders. Public Health Nursing 21: 435‐42 96. Husaini BA, Sherkat DE, Levine R, Bragg R, Cain VA, et al. 2002. The effect of a church‐based breast cancer screening education program on mammography rates among African‐American women. Journal of the National Medical Association 94: 100‐6 97. Friedman LC, Everett TE, Peterson L, Ogbonnaya KI, Mendizabal V. 2001. Compliance with fecal occult blood test screening among low‐income medical outpatients: A randomized controlled trial using a videotaped intervention. Journal of Cancer Education 16: 85‐8 98. Powe BD, Weinrich S. 1999. An intervention to decrease cancer fatalism among rural elders. Oncol Nurs Forum 26: 583‐8 99. Wood RY, Duffy ME. 2004. Video breast health kits: Testing a cancer education innovation in older high‐risk populations. Journal of Cancer Education 19: 98‐104 100. Taylor VM, Hislop TG, Jackson JC, Tu SP, Yasui Y, et al. 2002. A randomized controlled trial of interventions to promote cervical cancer screening among Chinese women in North America. J Natl Cancer Inst 94: 670‐7 101. Fitzgibbon ML, Ferreira MR, Dolan NC, Davis TC, Rademaker AW, et al. 2007. Process evaluation in an intervention designed to improve rates of colorectal cancer screening in a VA medical center. Health Promot Pract 8: 273‐81 102. Rimer BK, Halabi S, Sugg Skinner C, Lipkus IM, Strigo TS, et al. 2002. Effects of a mammography decision‐making intervention at 12 and 24 months. American Journal of Preventive Medicine 22: 247‐57 103. Loss J, Eichhorn C, Nagel E. 2006. The effects of promoting colorectal cancer screening on screening utilisation: Evaluation of the German campaign "aktiv gegen darmkrebs" (action against colorectal cancer). Zeitschrift fur Gastroenterologie 44: 1127‐34 VOL 3– 84 Knowledge translation: a review of strategies to increase participation in cancer screening 104. Miller Jr DP, Kimberly Jr JR, Case LD, Wofford JL. 2005. Using a computer to teach patients about fecal occult blood screening: A randomized trial. Journal of General Internal Medicine 20: 984‐8 105. Mann BD, Sherman L, Clayton C, Johnson RF, Keates J, et al. 2000. Screening to the converted: an educational intervention in African American churches. J Cancer Educ 15: 46‐50 106. Powell ME, Carter V, Bonsi E, Johnson G, Williams L, et al. 2005. Increasing mammography screening among African American women in rural areas. J Health Care Poor and Underserved 16: 11‐21 107. Yabroff KR, Mangan P, Mandelblatt J. 2003. Effectiveness of interventions to increase Papanicolaou smear use. Journal of the American Board of Family Practice 16: 188‐203 108. Costanza ME, Stoddard AM, Luckmann R, White MJ, Spitz Avrunin J, Clemow L. 2000. Promoting mammography. Results of a randomized trial of telephone counseling and a medical practice intervention. American Journal of Preventive Medicine 19: 39‐46 109. Stoddard AM, Fox SA, Costanza ME, Lane DS, Andersen MR, et al. 2002. Effectiveness of telephone counseling for mammography: results from five randomized trials. Prev Med 34: 90‐9 110. Champion VL, Skinner CS, Foster JL. 2000. The effects of standard care counseling or telephone/in‐person counseling on beliefs, knowledge, and behavior related to mammography screening. Oncol Nurs Forum 27: 1565‐71 111. Luckmann R, Savageau JA, Clemow L, Stoddard AM, Costanza ME. 2003. A randomized trial of telephone counseling to promote screening mammography in two HMOs. Cancer Detect Prev 27: 442‐50 112. Messina CR, Lane DS, Grimson R. 2002. Effectiveness of women's telephone counseling and physician education to improve mammography screening among women who underuse mammography. Annals of Behavioral Medicine 24: 279‐89 113. Duan N, Fox SA, Derose KP, Carson S. 2000. Maintaining mammography adherence through telephone counseling in a church‐based trial. Am J Public Health 90: 1468‐71 114. Bowen DJ, Powers D, Greenlee H. 2006. Effects of breast cancer risk counseling for sexual minority women. Health Care Women Int 27: 59‐74 115. Champion VL, Ray DW, Heilman DK, Springston JK. 2000. A tailored intervention for mammography among low‐income African‐American women. Journal of Psychosocial Oncology 18: 1‐13 116. Giveon S. 2000. Patient adherence to family practitioners' recommendations for breast cancer screening: A historical cohort study. Family Practice 17: 42‐5 117. Dietrich AJ, Tobin JN, Cassells A, Robinson CM, Reh M, et al. 2007. Translation of an efficacious cancer‐screening intervention to women enrolled in a Medicaid managed care organization. Ann Fam Med 5: 320‐7 118. Burhansstipanov L, Dignan MB, Wound D, Tenney M, Vigil G. 2000. Native American recruitment into breast cancer screening: The NAWWA project. Journal of Cancer Education 15: 28‐32 119. Earp JA, Eng E, O'Malley MS, Altpeter M, Rauscher G, et al. 2002. Increasing use of mammography among older, rural African American women: Results from a community trial. American Journal of Public Health 92: 646‐54 120. Sauaia A, Min S‐J, Lack D, Apodaca C, Osuna D, et al. 2007. Church‐based breast cancer screening education: impact of two approaches on Latinas enrolled in public and private health insurance plans. Prev Chronic Dis 4: A99 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 85 121. Zhu K, Hunter S, Bernard LJ, Payne‐Wilks K, Roland CL, et al. 2002. An intervention study on screening for breast cancer among single African‐American women aged 65 and older. Preventive Medicine 34: 536‐45 122. Oliver‐Vazquez M, Sanchez‐Ayendez M, Suarez‐Perez E, Velez‐Almodovar H, Arroyo‐Calderon Y. 2002. Breast cancer health promotion model for older Puerto Rican women: Results of a pilot programme. Health Promotion International 17: 3‐11 123. Lam TK, McPhee SJ, Mock J, Wong C, Doan HT, et al. 2003. Encouraging Vietnamese‐American women to obtain Pap tests through lay health worker outreach and media education. J Gen Intern Med 18: 516‐24 124. Messer L, Steckler A, Dignan M. 1999. Early detection of cervical cancer among native American women: A qualitative supplement to a quantitative study. Health Education and Behavior 26: 547‐62 125. Taylor VM, Jackson JC, Yasui Y, Kuniyuki A, Acorda E, et al. 2002. Evaluation of an outreach intervention to promote cervical cancer screening among Cambodian American women. Cancer Detection and Prevention 26: 320‐7 126. Calam B, Norgrove L, Brown D, Wilson MA. 1999. Pap screening clinics with native women in Skidegate, Haida Gwaii ‐ Need for innovation. Canadian Family Physician 45: 355‐60 127. Vivilaki V, Romanidou A, Theodorakis P, Lionis C. 2005. Are health education meetings effective in recruiting women in cervical screening programmes? An innovative and inexpensive intervention from the island of Crete. Rural Remote Health 5: 376 128. Jandorf L, Gutierrez Y, Lopez J, Christie J, Itzkowitz SH. 2005. Use of a patient navigator to increase colorectal cancer screening in an urban neighborhood health clinic. Journal of Urban Health‐Bulletin of the New York Academy of Medicine 82: 216‐24 129. Lopez VA, Castro FG. 2006. Participation and program outcomes in a church‐based cancer prevention program for Hispanic women. Journal of Community Health 31: 343‐62 130. Dietrich AJ, Tobin JN, Cassells A, Robinson CM, Greene MA, et al. 2006. Telephone care management to improve cancer screening among low‐income women ‐ A randomized, controlled trial. Annals of Internal Medicine 144: 563‐71 131. Worden JK, Mickey RM, Vacek PM, Flynn BS, Solomon LJ, et al. 2002. Evaluation of a community breast screening promotion program. Preventive Medicine 35: 349‐61 132. English KC, Fairbanks J, Finster CE, Rafelito A, Luna J, Kennedy M. 2008. A socioecological approach to improving mammography rates in a tribal community. Health Educ Behav 35: 396‐
409 133. Andersen MR, Yasui Y, Meischke H, Kuniyuki A, Etzioni R, Urban N. 2000. The effectiveness of mammography promotion by volunteers in rural communities. Am J Prev Med 18: 199‐207 134. Givaudan M, Leenen I, Pick S, Angulo A, Poortinga YH. 2008. Enhancement of underused cervical cancer prevention services in rural Oaxaca, Mexico. Revista Panamericana de Salud Publica/Pan American Journal of Public Health 23: 135‐43 135. Nguyen TT, McPhee SJ, Bui‐Tong N, Luong T‐N, Ha‐Iaconis T, et al. 2006. Community‐based participatory research increases cervical cancer screening among Vietnamese‐Americans. J Health Care Poor and Underserved 17: 31‐54 136. Hancock L, Sanson‐Fisher R, Perkins J, Corkrey R, Burton R, Reid S. 2001. Effect of a community action intervention on cervical cancer screening rates in rural Australian towns: The CART project. Preventive Medicine 32: 109‐17 VOL 3– 86 Knowledge translation: a review of strategies to increase participation in cancer screening 137. Sarfaty M, Feng S. 2006. Choice of screening modality in a colorectal cancer education and screening program for the uninsured. Journal of Cancer Education 21: 43‐9 138. Paskett ED, Tatum CM, D'Agostino Jr R, Rushing J, Velez R, et al. 1999. Community‐based interventions to improve breast and cervical cancer screening: Results of the Forsyth County Cancer Screening (FoCaS) project. Cancer Epidemiology Biomarkers and Prevention 8: 453‐9 139. Gotay CC, Banner RO, Matsunaga DS, Hedlund N, Enos R, et al. 2000. Impact of a culturally appropriate intervention on breast and cervical screening among native Hawaiian women. Preventive Medicine 31: 529‐37 140. Ciaranello AL, Molitor F, Leamon M, Kuenneth C, Tancredi D, et al. 2006. Providing health care services to the formerly homeless: A quasi‐experimental evaluation. Journal of Health Care for the Poor and Underserved 17: 441‐61 141. Fouad MN, Partridge E, Dignan M, Holt C, Johnson R, et al. 2006. A community‐driven action plan to eliminate breast and cervical cancer disparity: successes and limitations. J Cancer Educ 21: S91‐100 142. Blumenthal DS, Fort JG, Ahmed NU, Semenya KA, Schreiber GB, et al. 2005. Impact of a two‐city community cancer prevention intervention on African Americans. J Natl Med Assoc 97: 1479‐88 143. Thompson B, Coronado G, Chen L, Islas I. 2006. Celebremos la salud! a community randomized trial of cancer prevention (United States). Cancer Causes Control 17: 733‐46 144. Mullins R, Wakefield M, Broun K. 2008. Encouraging the right women to attend for cervical cancer screening: results from a targeted television campaign in Victoria, Australia. Health Educ. Res. 23: 477‐86 145. Marcus AC, Crane LA. 1998. A review of cervical cancer screening intervention research: implications for public health programs and future research. Prev Med 27: 13‐31 146. Pasick RJ, Hiatt RA, Paskett ED. 2004. Lessons learned from community‐based cancer screening intervention research. Cancer 101: 1146‐64 147. Fernbach M. 2001. In PapScreen Victoria Evaluation Report ed. M Fernbach, pp. 73‐85. Melbourne: Anti‐Cancer Council of Victoria 148. Perez D, Hardy M, Walsberger S, Morrell S, Hung A, et al. 2009. Don't Just Sit There: NSW Cervical Screening Campaign Evaluation: February 2009, Cancer Institute NSW, Alexandria, NSW 149. Mullins R. 2006. Evaluation of the impact of PapScreen's campaign on Culturally and Linguistically Diverse (CALD) women, The Cancer Council Victoria, Melbourne, Victoria 150. Allen JD, Stoddard AM, Sorensen G, Mays J. 2001. Promoting Breast and Cervical Cancer Screening in the Workplace: Results From the Woman to Woman Study. American Journal of Public Health 91: 584‐90 151. Hart AR, Glover N, Howick‐Baker J, Mayberry JF. 2003. An industry based approach to colorectal cancer screening in an asymptomatic population. Postgraduate Medical Journal 79: 646‐9 152. Tilley BC, Vernon SW, Myers R, Glanz K, Lu M, et al. 1999. The Next Step Trial: Impact of a worksite colorectal cancer screening promotion program. Preventive Medicine 28: 276‐83 153. Hou SI, Chen PH. 2004. Home‐administered fecal occult blood test for colorectal cancer screening among worksites in Taiwan. Preventive Medicine 38: 78‐84 154. Fox SA, Stein JA, Sockloskie RJ, Ory MG. 2001. Targeted mailed materials and the Medicare beneficiary: increasing mammogram screening among the elderly. Am J Public Health 91: 55‐61 155. Immonen‐Raiha P, Kauhava L, Parvinen I, Helenius H, Klemi P. 2001. Customer fee and participation in breast‐cancer screening. Lancet 358: 1425 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 87 156. Dolan NC, McDermott MM, Morrow M, Venta L, Martin GJ. 1999. Impact of same‐day screening mammography availability: Results of a controlled clinical trial. Archives of Internal Medicine 159: 393‐8 157. Gregory‐Mercado KY, Will J, True S, Royalty J, Starcher II ET, et al. 2007. A combined approach to women's health is associated with a greater likelihood of repeat mammography in a population of financially disadvantaged women. Prev Chronic Dis 4: A89 158. Thomson RM, Crengle S, Lawrenson R. 2009. Improving participation in breast screening in a rural general practice with a predominately Maori population. New Zealand Medical Journal 122: 39‐47 159. Bais AG, van Kemenade FJ, Berkhof J, Verheijen RHM, Snijders PJF, et al. 2007. Human papillomavirus testing on self‐sampled cervicovaginal brushes: an effective alternative to protect nonresponders in cervical screening programs. Int J Cancer 120: 1505‐10 160. Sowjanya AP, Paul P, Vedantham H, Ramakrishna G, Vidyadhari D, et al. 2009. Suitability of self‐
collected vaginal samples for cervical cancer screening in periurban villages in Andhra Pradesh, India. Cancer Epidemiol Biomarkers Prev 18: 1373‐8 161. Wikstrom I, Stenvall H, Wilander E. 2007. Attitudes to self‐sampling of vaginal smear for human papilloma virus analysis among women not attending organized cytological screening. Acta Obstet Gynecol Scand 86: 720‐5 162. Tuominen A, Anttila A, Malila N, Leinonen M, Luostarinen T, Nieminen P. 2009. Self‐sample HPV test ‐ increasing screening coverage among non‐responders in Finland, Malmo, Sweden 163. Chang HC, Hsiung HY, Chen SI, Yen AMF, Chen THH. 2007. Comparison of a community outreach service with opportunity screening for cervical cancer using Pap smears. Journal of Public Health 29: 165‐72 164. Sox CH, Dietrich AJ, Goldman DC, Provost EM. 1999. Improved access to women's health services for Alaska natives through Community Health Aide training. Journal of Community Health 24: 313‐23 165. Federici A, Rossi PG, Borgia P, Bartolozzi F, Farchi S, Gausticchi G. 2005. The immunochemical faecal occult blood test leads to higher compliance than the guaiac for colorectal cancer screening programmes: a cluster randomized controlled trial. Journal of Medical Screening 12: 83‐8 166. Federici A, Rossi PG, Bartolozzi F, Farchi S, Borgia P, Guastcchi G. 2006. The role of GPs in increasing compliance to colorectal cancer screening: a randomised controlled trial (Italy). Cancer Causes & Control 17: 45‐52 167. Tifratene K, Eisinger F, Rinaldi Y, Didelot R, Seitz J‐F. 2007. Colorectal cancer screening program: cost effectiveness of systematic recall letters. Gastroenterol Clin Biol 31: 929‐33 168. Cole SR, Young GP. 2001. Effect of dietary restriction on participation in faecal occult blood test screening for colorectal cancer. Med J Aust 175: 195‐8 169. Cole SR, Young GP, Esterman A, Cadd B, Morcom J. 2003. A randomised trial of the impact of new faecal haemoglobin test technologies on population participation in screening for colorectal cancer. J Med Screen 10: 117‐22 170. Boulware LE, Marinopoulos S, Phillips KA, Hwang CW, Maynor K, et al. 2007. Systematic review: The value of the periodic health evaluation. Annals of Internal Medicine 146: 289‐300 171. Bruce Jr A, Bazargan‐Hejazi S. 2005. Evaluating a tailored intervention to increase screening mammography in an urban area. Journal of the National Medical Association 97: 1350‐60 VOL 3– 88 Knowledge translation: a review of strategies to increase participation in cancer screening 172. Young RF, Waller Jr JB, Smitherman H. 2002. A breast cancer education and on‐site screening intervention for unscreened African American women. Journal of Cancer Education 17: 231‐6 173. Bell TS, Branston LK, Newcombe RG, Barton GR. 1999. Interventions to improve uptake of breast screening in inner city Cardiff general practices with ethnic minority lists. Ethn Health 4: 277‐84 174. Fowler BA, Rodney M, Roberts S, Broadus L. 2005. Collaborative breast health intervention for African American women of lower socioeconomic status. Oncology Nursing Forum 32: 1207‐16 175. Taylor V, Thompson B, Lessler D, Yasui Y, Montano D, et al. 1999. A clinic‐based mammography intervention targeting inner‐city women. Journal of General Internal Medicine 14: 104‐11 176. Fang CY, Ma GX, Tan Y, Chi N. 2007. A multifaceted intervention to increase cervical cancer screening among underserved Korean women. Cancer Epidemiology Biomarkers and Prevention 16: 1298‐302 177. Lewis CL, Brenner AT, Griffith JM, Pignone MP. 2008. The uptake and effect of a mailed multi‐
modal colon cancer screening intervention: A pilot controlled trial. Implementation Science 3 178. Potter MB, Phengrasamy L, Hudes ES, McPhee SJ, Walsh JME. 2009. Offering annual fecal occult blood tests at annual flu shot clinics increases colorectal cancer screening rates. Annals of Family Medicine 7: 17‐23 179. Braun KL, Fong M, Kaanoi ME, Kamaka ML, Gotay CC. 2005. Testing a culturally appropriate, theory‐based intervention to improve colorectal cancer screening among native Hawaiians. Preventive Medicine 40: 619‐27 180. Pignone M, Harris R, Kinsinger L. 2000. Videotape‐based decision aid for colon cancer screening. A randomized, controlled trial. Annals of Internal Medicine 133: 761‐9 181. Vogt TM, Glass A, Glasgow RE, La Chance PA, Lichtenstein E. 2003. The safety net: A cost‐
effective approach to improving breast and cervical cancer screening. Journal of Womens Health 12: 789‐98 182. Newell SA, Sanson‐Fisher RW, Girgis A, Davey HM. 2002. Can personal health record booklets improve cancer screening behaviors? Am J Prev Med 22: 15‐22 183. Anderson J. 2004. Literature review of effective reminder letter strategies used in screening programs, The Cancer Council Victoria, Melbourne, Victoria 184. Schenck AP, Peacock S, Pignone M, Jackson E, Gunter N, Klabunde CN. 2006. Increasing colorectal cancer testing: Translating physician interventions into population‐based practice. Health Care Financing Review 27: 25‐35 185. Roetzheim RG, Christman LK, Jacobsen PB, Cantor AB, Schroeder J, et al. 2004. A randomized controlled trial to increase cancer screening among attendees of community health centers. Ann Fam Med 2: 294‐300 186. Roetzheim RG, Christman LK, Jacobsen PB, Schroeder J, Abdulla R, Hunter S. 2005. Long‐term results from a randomized controlled trial to increase cancer screening among attendees of community health centers. Ann Fam Med 3: 109‐14 187. Kupets R, Covens A. 2001. Strategies for the implementation of cervical and breast cancer screening of women by primary care physicians. Gynecol Oncol 83: 186‐97 188. Basch CE, Wolf RL, Brouse CH, Shmukler C, Neugut A, et al. 2006. Telephone outreach to increase colorectal cancer screening in an urban minority population. American Journal of Public Health 96: 2246‐53 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 89 189. Cole SR, Young GP, Byrne D, Guy JR, Morcom J. 2002. Participation in screening for colorectal cancer based on a faecal occult blood test is improved by endorsement by the primary care practitioner. Journal of Medical Screening 9: 147‐52 190. Stokamer CL, Tenner CT, Chaudhuri J, Vazquez E, Bini EJ. 2005. Randomized controlled trial of the impact of intensive patient education on compliance with fecal occult blood testing. Journal of General Internal Medicine 20: 278‐82 191. Hart AR, Eaden J, Barnett S, de Bono AM, Mayberry JF. 1998. Colorectal cancer prevention. An approach to increasing compliance in a faecal occult blood test screening programme. J Epidemiol Community Health 52: 818‐20 192. Robertson LM, Douglas F, Ludbrook A, Reid G, van Teijlingen E. 2008. What works with men? A systematic review of health promoting interventions targeting men. BMC Health Serv Res 8: 141 193. Holland D, Bradley D, Khoury J. 2005. Sending men the message about preventive care: an evaluation of communication strategies. International Journal of Men's Health 4: 97‐114 194. Malcher G. 2009. Engaging men in health care. Aust Fam Physician 38: 92‐5 195. Galdas PM, Cheater F, Marshall P. 2005. Men and health help‐seeking behaviour: literature review. J Adv Nurs 49: 616‐23 196. Page A, Morrell S, Tewson R, Taylor R, Brassil A. 2005. Mammography screening participation: Effects of a media campaign targeting Italian‐speaking women. Australian and New Zealand Journal of Public Health 29: 365‐71 197. Dinshaw K, Mishra G, Shastri S, Badwe R, Kerkar R, et al. 2007. Determinants of compliance in a cluster randomised controlled trial on screening of breast and cervix cancer in Mumbai, India ‐ 1. Compliance to screening. Oncology 73: 145‐53 198. Oelke ND. 2002. Cervical and Breast Cancer Screening Literature Review, Alberta Cancer Board, Alberta 199. Legler J, Meissner HI, Coyne C, Breen N, Chollette V, Rimer BK. 2002. The effectiveness of interventions to promote mammography among women with historically lower rates of screening. Cancer Epidemiology Biomarkers and Prevention 11: 59‐71 200. Jepson RG, Forbes CA, Sowden AJ, Lewis RA. 2001. Increasing informed uptake and non‐uptake of screening: Evidence from a systematic review. Health Expectations 4: 116‐26 201. Garton MJ, Torgerson DJ, Donaldson C, Russell IT, Reid DM. 1992. Recruitment methods for screening programmes: trial of a new method within a regional osteoporosis study. Bmj 305: 82‐
4 202. Hutchison B, Birch S, Evans CE, Goldsmith LJ, Markham BA, et al. 1998. Screening for hypercholesterolaemia in primary care: randomised controlled trial of postal questionnaire appraising risk of coronary heart disease. Bmj 316: 1208‐13 203. Weinrich SP, Weinrich MC, Boyd MD, Atkinson C. 1998. The impact of prostate cancer knowledge on cancer screening. Oncol Nurs Forum 25: 527‐34 204. Heath GW, Fuchs R, Croft JB, Temple SP, Wheeler FC. 1995. Changes in blood cholesterol awareness: final results from the South Carolina Cardiovascular Disease Prevention Project. Am J Prev Med 11: 190‐6 205. Snyder LB, Hamilton MA, Mitchell EW, Kiwanuka‐Tondo J, Fleming‐Milici F, Proctor D. 2004. A meta‐analysis of the effect of mediated health communication campaigns on behavior change in the United States. Journal of Health Communication 9: 71‐96 VOL 3– 90 Knowledge translation: a review of strategies to increase participation in cancer screening 206. Owen N, James R, Henrikson D, Van Beurden E. 1990. Community cholesterol screenings: the impact of follow‐up letters and incentives on retest rates and biometric changes in follow‐up screenings. Am J Health Promotion 5: 58‐61 207. Christensen B. 1995. Characteristics of attenders and non‐attenders at health examinations for ischaemic heart disease in general practice. Scand J Prim Health Care 13: 26‐31 208. Davies M, Alban‐Davies H, Cook C, Day J. 1991. Self testing for diabetes mellitus. Bmj 303: 696‐8 209. Robson J, Boomla K, Fitzpatrick S, Jewell AJ, Taylor J, et al. 1989. Using nurses for preventive activities. Bmj 298: 433‐6 210. National Advisory Committee on Health and Disability. 1998. Population Screening for Colorectal Cancer, Ministry of Health, Wellington, New Zealand 211. Curbow B, Bowie J, Garza MA, McDonnell K, Scott LB, et al. 2004. Community‐based cancer screening programs in older populations: making progress but can we do better? Preventive Medicine 38: 676‐93 212. Fernbach M. 2002. The Impact of a Media Campaign on Cervical Screening Knowledge and Self‐
efficacy. J Health Psychol 7: 85‐97 213. Jenkins CN, McPhee SJ, Bird JA, Pham GQ, Nguyen BH, et al. 1999. Effect of a media‐led education campaign on breast and cervical cancer screening among Vietnamese‐American women. Prev Med 28: 395‐406 214. Reuben DB, Bassett LW, Hirsch SH, Jackson CA, Bastani R. 2002. A randomized clinical trial to assess the benefit of offering on‐site mobile mammography in addition to health education for older women. AJR Am J Roentgenol 179: 1509‐14 Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 91 VOL 3– 92 Knowledge translation: a review of strategies to increase participation in cancer screening
Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) Abood DA et al. Level III‐1 Int1: Message framing (loss) (2005)79 Breast Received telephonically "loss" framed USA Women script 50–64 Loss‐framed minimal Uninsured and underinsured women who Int2: Usual care intervention increases telephoned to enquire about mammogram Usual procedure (telephoned re: eligibility mammography use at one of two urban clinics randomly requirements and offered appointment) selected – Florida Allen JD et al. Level II Int1: Worksite intervention (2001)150 Breast Cervical Baseline questionnaire + 16 month USA Women intervention involving peer health advisors ≥ 40 overseen by volunteer advisory boards + Promoting breast and Employees employed for more than 15 follow‐up questionnaire cervical cancer hours per week on a permanent basis at 26 screening in the worksites. Worksite eligibility included: i) Int2: Usual care workplace: results minimum of 60 target group members; ii) Baseline questionnaire + usual care + from the woman to union representation among some follow‐up questionnaire woman study segment of the workforce; and iii) geographic location within 1.5 hours of the study centre. Participating sites included public community hospitals and chronic care facilities, private community hospitals, state agencies, state universities and private health organisations. APPENDIX: ANNOTATED BIBLIOGRAPHY (154 primary studies) Screening rates adjusted for sample stratum (i.e. age group) and worksite cluster. (i) Mammogram within past 2 years Int1: baseline = 3.2%, follow‐up = 90.4% change = +7.2% Int2: baseline = 84.3%, follow‐up = 89.9% change = +5.6% Int1 vs Int2 (change): OR 1.14 (95% CI 0.90–1.44); Not stat sig (ii) Pap test within past 3 years Int1: baseline = 85.2%, follow‐up = 89.9% change = +4.7% Int2: baseline = 85.8%, follow‐up = 87.7% change = +1.9% Int1 vs Int2 (change): OR 1.28 (95% CI 1.01–1.62); Stat sig Mammogram during 6‐month study period: Int1: 31/112 (27.7%) Int2: 157/992 (15.8%) Int1 vs Int2: Chi sq = 7.48; p = 0.0063; Adjusted OR = 1.914 (95% CI 1.20–3.05); Stat sig Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 93
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level II Breast Women ≥ 40 Random sample of women living in inner city area of Los Angeles Level II Breast Women 50–80 Residents of 40 communities located predominantly in rural areas of Washington state. Author (Publication year) Country Title Allen B et al. (2005)171 USA Evaluating a tailored intervention to increase screening mammography in an urban area Andersen MR et al. (2000)133 USA The effectiveness of mammography promotion by volunteers in rural communities Screening rates at 6‐month follow‐up Int1 vs Int 2: Adjusted OR = 1.76 (95% CI 1.06–2.92); Stat sig Int1: Multi‐component mailing + counselling (telephone) Enrolment and baseline assessment + culturally and ethnically tailored telephone counselling by trained interviewers + mailed intervention materials + 6 months follow‐up assessment Int2: Usual care Enrolment and baseline assessment + usual care + 6‐month follow‐up assessment. Int1: Usual care Baseline interview + 3‐year follow‐up interview Int2: Counselling (telephone) Mailed brochures + barrier specific telephone counselling of women 'not known' to be regular users of mammography by volunteers Int3: Community intervention Community activities undertaken by trained volunteers Int4: Counselling & community intervention Individual counselling + community activities (i) Regular mammography users at baseline Int1: 50.9% Int2: 50.1% Int3: 50.6% Int4: 48.5%; No stat sig diffs compared to usual care (ii) Conversion of under‐users at baseline to regular users Int1: 57.8% Int2: 60.6% Int3: 59.9% Int4: 60.4%; No stat sig diffs compared to usual care (iii) Regular users at baseline maintained at follow‐up Int1: 92.2% Int2: 91.8% Int3: 95.1% Int4: 93.6% Findings Nature of the intervention VOL 3– 92 Knowledge translation: a review of strategies to increase participation in cancer screening
Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) Abood DA et al. Level III‐1 Int1: Message framing (loss) (2005)79 Breast Received telephonically "loss" framed USA Women script 50–64 Loss‐framed minimal Uninsured and underinsured women who Int2: Usual care intervention increases telephoned to enquire about mammogram Usual procedure (telephoned re: eligibility mammography use at one of two urban clinics randomly requirements and offered appointment) selected – Florida Allen JD et al. Level II Int1: Worksite intervention (2001)150 Breast Cervical Baseline questionnaire + 16 month USA Women intervention involving peer health advisors ≥ 40 overseen by volunteer advisory boards + Promoting breast and Employees employed for more than 15 follow‐up questionnaire cervical cancer hours per week on a permanent basis at 26 screening in the worksites. Worksite eligibility included: i) Int2: Usual care workplace: results minimum of 60 target group members; ii) Baseline questionnaire + usual care + from the woman to union representation among some follow‐up questionnaire woman study segment of the workforce; and iii) geographic location within 1.5 hours of the study centre. Participating sites included public community hospitals and chronic care facilities, private community hospitals, state agencies, state universities and private health organisations. APPENDIX: ANNOTATED BIBLIOGRAPHY (154 primary studies) Screening rates adjusted for sample stratum (i.e. age group) and worksite cluster. (i) Mammogram within past 2 years Int1: baseline = 3.2%, follow‐up = 90.4% change = +7.2% Int2: baseline = 84.3%, follow‐up = 89.9% change = +5.6% Int1 vs Int2 (change): OR 1.14 (95% CI 0.90–1.44); Not stat sig (ii) Pap test within past 3 years Int1: baseline = 85.2%, follow‐up = 89.9% change = +4.7% Int2: baseline = 85.8%, follow‐up = 87.7% change = +1.9% Int1 vs Int2 (change): OR 1.28 (95% CI 1.01–1.62); Stat sig Mammogram during 6‐month study period: Int1: 31/112 (27.7%) Int2: 157/992 (15.8%) Int1 vs Int2: Chi sq = 7.48; p = 0.0063; Adjusted OR = 1.914 (95% CI 1.20–3.05); Stat sig Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 93
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level II Breast Women ≥ 40 Random sample of women living in inner city area of Los Angeles Level II Breast Women 50–80 Residents of 40 communities located predominantly in rural areas of Washington state. Author (Publication year) Country Title Allen B et al. (2005)171 USA Evaluating a tailored intervention to increase screening mammography in an urban area Andersen MR et al. (2000)133 USA The effectiveness of mammography promotion by volunteers in rural communities Screening rates at 6‐month follow‐up Int1 vs Int 2: Adjusted OR = 1.76 (95% CI 1.06–2.92); Stat sig Int1: Multi‐component mailing + counselling (telephone) Enrolment and baseline assessment + culturally and ethnically tailored telephone counselling by trained interviewers + mailed intervention materials + 6 months follow‐up assessment Int2: Usual care Enrolment and baseline assessment + usual care + 6‐month follow‐up assessment. Int1: Usual care Baseline interview + 3‐year follow‐up interview Int2: Counselling (telephone) Mailed brochures + barrier specific telephone counselling of women 'not known' to be regular users of mammography by volunteers Int3: Community intervention Community activities undertaken by trained volunteers Int4: Counselling & community intervention Individual counselling + community activities (i) Regular mammography users at baseline Int1: 50.9% Int2: 50.1% Int3: 50.6% Int4: 48.5%; No stat sig diffs compared to usual care (ii) Conversion of under‐users at baseline to regular users Int1: 57.8% Int2: 60.6% Int3: 59.9% Int4: 60.4%; No stat sig diffs compared to usual care (iii) Regular users at baseline maintained at follow‐up Int1: 92.2% Int2: 91.8% Int3: 95.1% Int4: 93.6% Findings Nature of the intervention VOL 3– 94 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings Int3 vs Int1; p = 0.010; i.e. Community intervention (Int3) sig different from usual care (Int1) Bais AG et al. Level II Int1: Invitations and reminders (mailed) Int1: 48/272 (18%) attended conventional (2007)159 Cervical Extra recall reminder for regular cytology screening; NETHERLANDS Women with an explanatory letter 224/272 (82%) did not respond 30–50 Human Neither responded to the regular invitation Int2: Procedures (self‐sampling) Int2: 736/2352 (30%) submitted SSVS kit; papillomavirus for screening nor to the first 6 months Received a self‐sample cervicovaginal 70/2352 (3%) attended convention screening; testing on self‐
reminder. specimens (SSVS) kit with instructions, an 1546/2352 (66%) did not respond sampled explanatory letter and a return envelope. cervicovaginal No statistical significance tests included. brushes: an effective alternative to protect Post hoc tests by CHPPE (screened/non‐response) nonresponders in Int 2 vs Int1: Chi sq = 110.449, df = 1, p = 0.000 cervical screening programs Barr JK et al. Level II Int1: Invitations and reminders (mailed) Int1 vs Int3: (2001)22 Breast Mail reminder Invitations and reminders (mailed) vs usual care: USA Women RR = 1.08 (95% CI 0.94–1.21); Not stat sig 50–75 Int2: Invitations and reminders (telephone) A randomized Continuously enrolled in a large group‐
Telephone reminder with an option for Int2 vs Int3: intervention to model HMO during the study who appointment scheduling Invitations and reminders (telephone) vs usual care: improve ongoing underwent a bilateral mammogram during RR = 1.39 (95% CI 1.25–1.54); Stat sig participation in the first quarter of 1994 and no Int3: Usual care mammography subsequent mammogram during the next Routine publicity campaign on 18 to 21 months. mammography for all women Basch CE et al. Level II Int1: Counselling (telephone) Verification of screening within 6 months using (2006)188 Colorectal Tailored telephone education and support medical records, billing claims USA Men (median number of calls – 5) > 52 Int1: 6/226 (27.0%) Telephone outreach Mainly Black population in New York metro Int2: 14/230 (6.1%) Author (Publication year) Country Title VOL 3– 94 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings Int3 vs Int1; p = 0.010; i.e. Community intervention (Int3) sig different from usual care (Int1) Bais AG et al. Level II Int1: Invitations and reminders (mailed) Int1: 48/272 (18%) attended conventional (2007)159 Cervical Extra recall reminder for regular cytology screening; NETHERLANDS Women with an explanatory letter 224/272 (82%) did not respond 30–50 Human Neither responded to the regular invitation Int2: Procedures (self‐sampling) Int2: 736/2352 (30%) submitted SSVS kit; papillomavirus for screening nor to the first 6 months Received a self‐sample cervicovaginal 70/2352 (3%) attended convention screening; testing on self‐
reminder. specimens (SSVS) kit with instructions, an 1546/2352 (66%) did not respond sampled explanatory letter and a return envelope. cervicovaginal No statistical significance tests included. brushes: an effective alternative to protect Post hoc tests by CHPPE (screened/non‐response) nonresponders in Int 2 vs Int1: Chi sq = 110.449, df = 1, p = 0.000 cervical screening programs Barr JK et al. Level II Int1: Invitations and reminders (mailed) Int1 vs Int3: (2001)22 Breast Mail reminder Invitations and reminders (mailed) vs usual care: USA Women RR = 1.08 (95% CI 0.94–1.21); Not stat sig 50–75 Int2: Invitations and reminders (telephone) A randomized Continuously enrolled in a large group‐
Telephone reminder with an option for Int2 vs Int3: intervention to model HMO during the study who appointment scheduling Invitations and reminders (telephone) vs usual care: improve ongoing underwent a bilateral mammogram during RR = 1.39 (95% CI 1.25–1.54); Stat sig participation in the first quarter of 1994 and no Int3: Usual care mammography subsequent mammogram during the next Routine publicity campaign on 18 to 21 months. mammography for all women Basch CE et al. Level II Int1: Counselling (telephone) Verification of screening within 6 months using (2006)188 Colorectal Tailored telephone education and support medical records, billing claims USA Men (median number of calls – 5) > 52 Int1: 6/226 (27.0%) Telephone outreach Mainly Black population in New York metro Int2: 14/230 (6.1%) Author (Publication year) Country Title VOL 3– 92 Knowledge translation: a review of strategies to increase participation in cancer screening
Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) Abood DA et al. Level III‐1 Int1: Message framing (loss) (2005)79 Breast Received telephonically "loss" framed USA Women script 50–64 Loss‐framed minimal Uninsured and underinsured women who Int2: Usual care intervention increases telephoned to enquire about mammogram Usual procedure (telephoned re: eligibility mammography use at one of two urban clinics randomly requirements and offered appointment) selected – Florida Allen JD et al. Level II Int1: Worksite intervention (2001)150 Breast Cervical Baseline questionnaire + 16 month USA Women intervention involving peer health advisors ≥ 40 overseen by volunteer advisory boards + Promoting breast and Employees employed for more than 15 follow‐up questionnaire cervical cancer hours per week on a permanent basis at 26 screening in the worksites. Worksite eligibility included: i) Int2: Usual care workplace: results minimum of 60 target group members; ii) Baseline questionnaire + usual care + from the woman to union representation among some follow‐up questionnaire woman study segment of the workforce; and iii) geographic location within 1.5 hours of the study centre. Participating sites included public community hospitals and chronic care facilities, private community hospitals, state agencies, state universities and private health organisations. APPENDIX: ANNOTATED BIBLIOGRAPHY (154 primary studies) Screening rates adjusted for sample stratum (i.e. age group) and worksite cluster. (i) Mammogram within past 2 years Int1: baseline = 3.2%, follow‐up = 90.4% change = +7.2% Int2: baseline = 84.3%, follow‐up = 89.9% change = +5.6% Int1 vs Int2 (change): OR 1.14 (95% CI 0.90–1.44); Not stat sig (ii) Pap test within past 3 years Int1: baseline = 85.2%, follow‐up = 89.9% change = +4.7% Int2: baseline = 85.8%, follow‐up = 87.7% change = +1.9% Int1 vs Int2 (change): OR 1.28 (95% CI 1.01–1.62); Stat sig Mammogram during 6‐month study period: Int1: 31/112 (27.7%) Int2: 157/992 (15.8%) Int1 vs Int2: Chi sq = 7.48; p = 0.0063; Adjusted OR = 1.914 (95% CI 1.20–3.05); Stat sig Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 93
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level II Breast Women ≥ 40 Random sample of women living in inner city area of Los Angeles Level II Breast Women 50–80 Residents of 40 communities located predominantly in rural areas of Washington state. Author (Publication year) Country Title Allen B et al. (2005)171 USA Evaluating a tailored intervention to increase screening mammography in an urban area Andersen MR et al. (2000)133 USA The effectiveness of mammography promotion by volunteers in rural communities Screening rates at 6‐month follow‐up Int1 vs Int 2: Adjusted OR = 1.76 (95% CI 1.06–2.92); Stat sig Int1: Multi‐component mailing + counselling (telephone) Enrolment and baseline assessment + culturally and ethnically tailored telephone counselling by trained interviewers + mailed intervention materials + 6 months follow‐up assessment Int2: Usual care Enrolment and baseline assessment + usual care + 6‐month follow‐up assessment. Int1: Usual care Baseline interview + 3‐year follow‐up interview Int2: Counselling (telephone) Mailed brochures + barrier specific telephone counselling of women 'not known' to be regular users of mammography by volunteers Int3: Community intervention Community activities undertaken by trained volunteers Int4: Counselling & community intervention Individual counselling + community activities (i) Regular mammography users at baseline Int1: 50.9% Int2: 50.1% Int3: 50.6% Int4: 48.5%; No stat sig diffs compared to usual care (ii) Conversion of under‐users at baseline to regular users Int1: 57.8% Int2: 60.6% Int3: 59.9% Int4: 60.4%; No stat sig diffs compared to usual care (iii) Regular users at baseline maintained at follow‐up Int1: 92.2% Int2: 91.8% Int3: 95.1% Int4: 93.6% Findings Nature of the intervention VOL 3– 94 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings Int3 vs Int1; p = 0.010; i.e. Community intervention (Int3) sig different from usual care (Int1) Bais AG et al. Level II Int1: Invitations and reminders (mailed) Int1: 48/272 (18%) attended conventional (2007)159 Cervical Extra recall reminder for regular cytology screening; NETHERLANDS Women with an explanatory letter 224/272 (82%) did not respond 30–50 Human Neither responded to the regular invitation Int2: Procedures (self‐sampling) Int2: 736/2352 (30%) submitted SSVS kit; papillomavirus for screening nor to the first 6 months Received a self‐sample cervicovaginal 70/2352 (3%) attended convention screening; testing on self‐
reminder. specimens (SSVS) kit with instructions, an 1546/2352 (66%) did not respond sampled explanatory letter and a return envelope. cervicovaginal No statistical significance tests included. brushes: an effective alternative to protect Post hoc tests by CHPPE (screened/non‐response) nonresponders in Int 2 vs Int1: Chi sq = 110.449, df = 1, p = 0.000 cervical screening programs Barr JK et al. Level II Int1: Invitations and reminders (mailed) Int1 vs Int3: (2001)22 Breast Mail reminder Invitations and reminders (mailed) vs usual care: USA Women RR = 1.08 (95% CI 0.94–1.21); Not stat sig 50–75 Int2: Invitations and reminders (telephone) A randomized Continuously enrolled in a large group‐
Telephone reminder with an option for Int2 vs Int3: intervention to model HMO during the study who appointment scheduling Invitations and reminders (telephone) vs usual care: improve ongoing underwent a bilateral mammogram during RR = 1.39 (95% CI 1.25–1.54); Stat sig participation in the first quarter of 1994 and no Int3: Usual care mammography subsequent mammogram during the next Routine publicity campaign on 18 to 21 months. mammography for all women Basch CE et al. Level II Int1: Counselling (telephone) Verification of screening within 6 months using (2006)188 Colorectal Tailored telephone education and support medical records, billing claims USA Men (median number of calls – 5) > 52 Int1: 6/226 (27.0%) Telephone outreach Mainly Black population in New York metro Int2: 14/230 (6.1%) Author (Publication year) Country Title VOL 3– 94 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings Int3 vs Int1; p = 0.010; i.e. Community intervention (Int3) sig different from usual care (Int1) Bais AG et al. Level II Int1: Invitations and reminders (mailed) Int1: 48/272 (18%) attended conventional (2007)159 Cervical Extra recall reminder for regular cytology screening; NETHERLANDS Women with an explanatory letter 224/272 (82%) did not respond 30–50 Human Neither responded to the regular invitation Int2: Procedures (self‐sampling) Int2: 736/2352 (30%) submitted SSVS kit; papillomavirus for screening nor to the first 6 months Received a self‐sample cervicovaginal 70/2352 (3%) attended convention screening; testing on self‐
reminder. specimens (SSVS) kit with instructions, an 1546/2352 (66%) did not respond sampled explanatory letter and a return envelope. cervicovaginal No statistical significance tests included. brushes: an effective alternative to protect Post hoc tests by CHPPE (screened/non‐response) nonresponders in Int 2 vs Int1: Chi sq = 110.449, df = 1, p = 0.000 cervical screening programs Barr JK et al. Level II Int1: Invitations and reminders (mailed) Int1 vs Int3: (2001)22 Breast Mail reminder Invitations and reminders (mailed) vs usual care: USA Women RR = 1.08 (95% CI 0.94–1.21); Not stat sig 50–75 Int2: Invitations and reminders (telephone) A randomized Continuously enrolled in a large group‐
Telephone reminder with an option for Int2 vs Int3: intervention to model HMO during the study who appointment scheduling Invitations and reminders (telephone) vs usual care: improve ongoing underwent a bilateral mammogram during RR = 1.39 (95% CI 1.25–1.54); Stat sig participation in the first quarter of 1994 and no Int3: Usual care mammography subsequent mammogram during the next Routine publicity campaign on 18 to 21 months. mammography for all women Basch CE et al. Level II Int1: Counselling (telephone) Verification of screening within 6 months using (2006)188 Colorectal Tailored telephone education and support medical records, billing claims USA Men (median number of calls – 5) > 52 Int1: 6/226 (27.0%) Telephone outreach Mainly Black population in New York metro Int2: 14/230 (6.1%) Author (Publication year) Country Title Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 95
Boling W et al. (2005)82 USA Increasing Screening rates: Int1: 50.7%; Historical control 1994 screening round 32%; Difference = +15.5% (95% CI +8.2%, +22.5%); Stat sig Int1 vs Int2: RR = 4.4 (95% CI 2.6–7.7); Stat sig Findings (i) Atlanta (intervention site) vs Decatur (matched control) Pap tests: effect difference = +1.1%; Not stat sig Mammograms: effect difference = +8.6%; p < 0.05 FOBT: effect difference = +12.3%; p < 0.01 (ii) Nashville (intervention site) vs Decatur (matched control) Pap tests: effect difference = +8.0%; p < 0.01 Mammograms: effect difference = –3.5%; Not stat sig FOBT: effect difference = –5.0%; Not stat sig Level IV Int1: Message Framing (cost) At 3 months: Breast Recruitment + baseline survey + cost 42/147 (28.6%) reported obtaining a mammogram Women brochure + reminder on patient's chart + 3‐ after receiving Int1 or Int2 or Int3 ≥ 40 month telephone survey Hospital patients admitted for non‐cancer No between‐group analysis presented Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) to increase colorectal who had not had recent screening Int2: Education (mailed) cancer screening in Letter, brochure re: colorectal cancer an urban minority prevention and screening options – discuss population with physician Bell TS et al. Level III‐3 Int1: Multi‐component (1999)173 Breast Message framing (translated literature UK (WALES) Women including a GP endorsement literature) + Not stated economic (free transport to the screening Interventions to Patients of three inner city general centre) + coaching (link workers acting as improve uptake of practices in Cardiff with a high proportion interpreters at the screening centre) breast screening in of ethnic minority women on their lists. inner city Cardiff general practices with ethnic minority lists Blumenthal DS et al. Level III‐2 Int1: Community intervention (2005)142 Breast, Colorectal, Cervical 35‐minute telephone pre‐intervention USA Women, Men survey + large scale, community ≥ 18 intervention focused on cancer prevention Impact of a two‐city African Americans 18 years of age and over and lasting 1.5 years in Atlanta and community cancer living in Atlanta (comparison Decatur) and Nashville + 35‐minute post‐intervention prevention Nashville (comparison Chattanooga) telephone survey intervention on African Americans Int2: Usual care 35‐minute telephone surveys, for both time periods, at Decatur and Chattanooga Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 95
Boling W et al. (2005)82 USA Increasing Screening rates: Int1: 50.7%; Historical control 1994 screening round 32%; Difference = +15.5% (95% CI +8.2%, +22.5%); Stat sig Int1 vs Int2: RR = 4.4 (95% CI 2.6–7.7); Stat sig Findings (i) Atlanta (intervention site) vs Decatur (matched control) Pap tests: effect difference = +1.1%; Not stat sig Mammograms: effect difference = +8.6%; p < 0.05 FOBT: effect difference = +12.3%; p < 0.01 (ii) Nashville (intervention site) vs Decatur (matched control) Pap tests: effect difference = +8.0%; p < 0.01 Mammograms: effect difference = –3.5%; Not stat sig FOBT: effect difference = –5.0%; Not stat sig Level IV Int1: Message Framing (cost) At 3 months: Breast Recruitment + baseline survey + cost 42/147 (28.6%) reported obtaining a mammogram Women brochure + reminder on patient's chart + 3‐ after receiving Int1 or Int2 or Int3 ≥ 40 month telephone survey Hospital patients admitted for non‐cancer No between‐group analysis presented Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) to increase colorectal who had not had recent screening Int2: Education (mailed) cancer screening in Letter, brochure re: colorectal cancer an urban minority prevention and screening options – discuss population with physician Bell TS et al. Level III‐3 Int1: Multi‐component (1999)173 Breast Message framing (translated literature UK (WALES) Women including a GP endorsement literature) + Not stated economic (free transport to the screening Interventions to Patients of three inner city general centre) + coaching (link workers acting as improve uptake of practices in Cardiff with a high proportion interpreters at the screening centre) breast screening in of ethnic minority women on their lists. inner city Cardiff general practices with ethnic minority lists Blumenthal DS et al. Level III‐2 Int1: Community intervention (2005)142 Breast, Colorectal, Cervical 35‐minute telephone pre‐intervention USA Women, Men survey + large scale, community ≥ 18 intervention focused on cancer prevention Impact of a two‐city African Americans 18 years of age and over and lasting 1.5 years in Atlanta and community cancer living in Atlanta (comparison Decatur) and Nashville + 35‐minute post‐intervention prevention Nashville (comparison Chattanooga) telephone survey intervention on African Americans Int2: Usual care 35‐minute telephone surveys, for both time periods, at Decatur and Chattanooga VOL 3– 96 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) diagnoses who had no mammogram in previous 12 months, and no select co morbid conditions Nature of the intervention Findings Int2: Message framing (fear) Recruitment + baseline survey + fear brochure + reminder on patient's chart + 3‐
month telephone survey Int3: Message framing (information) Recruitment + baseline survey + lack of information brochure + reminder on patient's chart + 3 month telephone survey
Bowen D J et al. Level II Baseline questionnaires to assess eligibility, Mammogram at baseline and at 24‐month follow‐up (2006)114 Breast then randomised to: for participants aged > 40 years: USA Women Int1: Counselling Int1: 75% to 87% 18–74 Four weekly 2‐hour psychosocial Int2: 75% to 75% Effects of breast Sexual minority women (lesbian and counselling sessions that include Int1 vs Int2: p < 0.05 cancer risk counseling bisexual) information dissemination, group for sexual minority discussion, and skills training, in groups of women 5–8 women led by trained counsellor Int2: Control Delayed counselling, counselling offered after 24 months Braun KL et al. Level II Int1: Education
FOBT screening rates: (2005)179 Colorectal Non‐Hawaiian nurse delivered educational Int1: 85% USA Women, Men presentation including information on CRC Int2: 67% ≥ 50 impact on Native Hawaiians, addressed Int2 vs Int1: Testing a culturally Native Hawaiian members from Civic Clubs cultural barriers, and featured Native OR = 0.364 (95% CI 0.14–0.97) appropriate, theory‐
Hawaiian artwork, words, and faces + based intervention to brochure on CRC (Native Hawaiian faces), improve colorectal phone numbers for local providers, and cancer screening FOBT kit with instructions + reminder Author (Publication year) Country Title mammography screening through inpatient education Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 97
Burhansstipanov L et al. (2000)118 USA Native American Burack RC et al. (2003)32 USA The effect of adding pap smear information to a mammography reminder system in an HMO: Results of randomized controlled trial Author (Publication year) Country Title among native Hawaiians Nature of the intervention Level III‐3 Breast Women Not stated American Indians in Denver Findings (i) Mammogram screening rates at one‐year follow‐
up Int1: 40% Int2: 39% Int2 vs Int1: Adjusted OR = 0.94 (95% CI 0.78–1.14); Not stat sig (ii) Pap test screening rates at one‐year follow‐up Int1: 23% Int2: 30% Int2 vs Int1: Adjusted OR = 1.39 (95% CI 1.07–1.89); Stat sig Int1: Coaching Int 1 vs historical ‘usual care’ controls Native American Women's Wellness through Awareness (NAWWA) Outreach Stat sig differences in the numbers of women having program: mammograms before and after the Native Sister (i) Local American Indian women trained as program was initiated (binomial test p < 0.05) lay health advisers (Native Sisters). phone call after one month for non‐
responders Int2: Education (theory) + counselling Native Hawaiian physician delivered educational presentation, Native Hawaiian CRC survivor told story, free FOBT kits with instructions from physician + multiple phone calls between 4 and 16 weeks post presentation to address and help solve screening‐related barriers Level II Int1: Invitations and reminders (single test) Breast, Cervical Mammogram only reminder letter + Women medical record prompts in patients' files ≥ 40 Women enrolled at one of 3 practice sites Int2: Invitations and reminders (multi‐test)
of a health maintenance organisation Combined mammogram and Pap test (HMO) – serves mainly Medicaid‐eligible reminder letter + medical record prompts population in Detroit, Michigan. in patients' files Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) VOL 3– 98 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention (ii) Located participants and encouraged them to participate in screening. They arranged transport and accompanied the women to mammography appointments if needed. Historical control "usual efforts" Calam B et al. Level III‐3 Int1: Coaching (1999)126 Cervical (i) Mailed invitation, letter of information, CANADA Women consent form – sent 6 weeks prior to 2 ≥ 40 half‐day Pap clinics Pap screening clinics First Nations people off the coast of British (ii) Community Health Reps (CHRs) then with native women in Columbia – Skidegate, Haida Gwaii, Queen made home visits to eligible women to Skidegate, Haida Charlotte islands discuss intent of program, get consent, Gwaii ‐ Need for book appointment times innovation Historical control Chalapati W Level III‐1 All participants were surveyed at home, (2007)46 Cervical then: THAILAND Women Int1: Invitation 35–60 Home visit to deliver invitation and fact Can a home‐visit Women from a rural area in Thailand who sheet + Home visit 4 months after 1st invitation increase did not have a pap test within the past 5 home visit for survey Pap smear screening years Int2: Control in Samliem, Khon Home visit 4 months after 1st home visit Kaen, Thailand? for survey Champion V et al. Level II Int1: Usual care (2003)75 Breast Recruitment + baseline interview + usual USA Women care (general postcard reminder to 50–85 schedule a mammogram) Author (Publication year) Country Title recruitment into breast cancer screening: The NAWWA project (i) Mammogram at six months: Int1: 25.6% Int2: 40.7% Int3: 50.8% Pap test uptake at 4‐month follow‐up: Int1: 11/100 Int2: 5/100 Int1 vs Int2: p = 0.193 Intention‐to‐treat analysis: rate of adequate Pap tests (in accordance with guidelines) Pre = 46/74 (62%) Post = 57/74 (77%) Difference = +15% No sig test presented Post hoc test by CHPPE Chi sq = 3.864, df = 1, p = 0.049 Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 99
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Members of an HMO and attendees at a general medicine clinic who had not had a mammogram within the past 15 months at time of recruitment and who did not have a history of breast cancer. Level II Breast Women 50–85 Women who had not had a mammogram in the last 15 months and had not had breast cancer, identified from medical records of a large HMO and a general medicine clinic in Indianapolis. Author (Publication year) Country Title Comparison of tailored interventions to increase mammography screening in nonadherent older women Champion VL et al. (2000)110 USA The effects of standard care counseling or telephone/in‐person counseling on beliefs, knowledge, and behavior related to mammography screening Findings Int4: 39.8% Int2: Counselling (telephone) Int5: 49.2% Recruitment + baseline interview + Int6: 55.0%; p < 0.001 tailored telephone counselling (ii) Logistic regression adjusting for previous Int3: Counselling (face‐to‐face) mammography, recruitment site, age, race, Recruitment + baseline interview + education, income and contemplation status tailored in‐person counselling compared to Int1: Int2 OR = 2.09 (95% CI 1.15–3.79), p = 0.016 Int4: Message framing (GP) Int3 OR = 2.45 (95% CI 1.37–4.38), p = 0.003 Recruitment + baseline interview + non‐
Int4 OR = 1.60 (95% CI 0.90–2.87), p = 0.112 tailored recommendation letter signed by Int5 OR = 2.22 (95% CI 1.25–3.93) p = 0.006 their primary care physician Int6 OR = 3.53 (95% CI 1.98–6.32), p < 0.001 Int5: Combining Int2 with Int4 Int6: Combining Int3 with Int4 Int1: Usual care Mammogram by 4‐week post intervention follow‐up Recruitment + baseline data collection + no data collection (compliant) counselling + follow‐up data collection (4 Int1: 17% weeks post intervention) Int2: 30% Int3: 33% Int2: Counselling (telephone) Recruitment + baseline data collection + Logistic regression: telephone counselling + follow‐up data Int2 vs Int1: OR = 2.18 (95% CI 1.34, 3.54) collection (4 weeks post intervention) Int3 vs Int1: OR = 2.80 (95% CI 1.75, 4.48) Int3: Counselling (face‐to‐face) Recruitment + baseline data collection + in‐
person counselling + follow‐up data collection (4 weeks post intervention) Nature of the intervention VOL 3– 100 Knowledge translation: a review of strategies to increase participation in cancer screening
Champion V L et al. (2000)115 USA A tailored intervention for mammography among low income African American women Author (Publication year) Country Title Champion VL et al. (2002)63 USA Comparisons of tailored mammography interventions at two months postintervention Level of evidence (NHMRC) Type of cancer Nature of the intervention Participants (sex), Participants (age) Participants (cohort) Level II Int1: Usual care Breast Recruitment + baseline data collection + Women usual care (no intervention) + follow‐up > 50 data collection (2 months) No mammogram in past 15 months at baseline and no history of breast cancer. Int2: Counselling (telephone) Identified through computer lists from a Recruitment + baseline data collection + medicine clinic in St Louis and two HMOs in tailored telephone counselling + follow‐up Indianapolis data collection (2 months) Int3: Message framing Recruitment + baseline data collection + tailored mailing + follow‐up data collection (2 months) Int4: Multi‐component (Int2 + Int3) Recruitment + baseline data collection + tailored mailing + tailored telephone counselling + data collection (2 months) Level II Int1: Counselling (face‐to‐face) Breast Recruitment + baseline data collection + Women face‐to‐face in‐home counselling and 2nd 45–64 data collection + follow‐up data collection African American women recruited while (1 year) waiting for services at a community service centre. The eligibility criteria included: i) Int2: Usual care no history of breast cancer; and ii) being at Recruitment + baseline data collection + or lower than a 150% maximum poverty face‐to‐face 2nd data collection + follow‐
level equivalent to $24,666 for a household up data collection (1 year) of four. Overall compliance with mammography: Int1: baseline 59.0%; 1 year later 75.7%; diff = +17% Int2: baseline 63.8%; 1 year later 71.1%; diff = +7% Compliance controlled for pre‐intervention benefits and barriers: Int1 vs Int2: OR = 11.82 (95% CI 1.22–115) Mammography adherence at 2 months compared to Int1: Int2: OR = 1.66 (95% CI 1.12–2.46), p = 0.012 Int3: OR = 1.72 (95%CI 1.18–2.52), p = 0.0048 Int4: OR = 2.16 (95% CI 1.46–3.19), p = 0.0001 Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 101
Church TR et al. (2004)38 USA A randomized trial of direct mailing of fecal occult blood tests to increase colorectal cancer screening Level II Colorectal Women, Men ≥ 50 Random sample of residents of Wright County (a non‐urban county location just west of the Minneapolis‐St Paul metropolitan area) Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) Chang HC et al. Level III‐3 (2007)163 Cervical TAIWAN Women ≥ 30 Comparison of a Women who participated in a cervical community outreach screening program from one hospital service with catchment in Northern Taiwan between opportunity screening 1999 and 2004. for cervical cancer using pap smears Findings Int1: Usual care Of those who attended outreach, 11.7% had 1999–2001 Hospital and clinic based previously attended screening at hospital between services 1999 & 2001. i.e. up to 89% attending the outreach service were under‐screened. Int 2: Procedures (mobile clinics) 2002–04 mobile clinics with pap screenings Pap test screening rates (pre and post) after in local settings (eg. schools, local centres) introduced outreach service (i.e. Int1 + Int2): 53% (95% CI 25, 80%) increase in the uptake of screening; significant. Outreach service provided greater proportion of screening as distance from hospital increases – from 25.6% to 48.9%. Int1: Usual care Effect of direct mailing on FOBT adherence at Media campaign + baseline questionnaire follow‐up + follow‐up questionnaire (1 year) Int2 + Int3 vs Int1 Women < 65 years: Int2: Invitations and reminders (no Diff = +16.3% (95% CI 6.0–26.8) reminder) Men < 65 years: Media campaign + baseline questionnaire Diff = +17.4% (95% CI 6.3–28.1) + FOBT kits by direct mail + follow‐up Women ≥ 65 years: questionnaire (1 year) Diff = +17.2% (95% CI 2.8–30.7) Men ≥ 65 years: Int3: Invitations and reminders (reminder) Diff = +28.2% (95% CI 13.5–43.9) Media campaign + baseline questionnaire + FOBT kits by direct mail + reminders to Effect of reminder notice on FOBT adherence at non‐respondents one month later + follow‐ follow‐up up questionnaire (1 year) Int3 vs Int2: Women < 65 years: Diff = +5.0% (95% CI –9.0–19.3) Nature of the intervention VOL 3– 102 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level III‐2 Cervical Women, Men Adult Residents of transitional housing facilities at four intervention sites and two comparison sites Author (Publication year) Country Title Ciaranello AL et al. (2006)140 USA Providing health care services to the formerly homeless: A quasi‐experimental evaluation Int1: Control group Participation in interviews and brief physical examinations Int2: Community intervention HEALTH Project Integrated Service Team (medical director, nurse practitioner, medical clerk, social worker) made weekly visits to the intervention sites. They provided comprehensive health assessments at baseline, follow‐up care, social work services, health education and referrals. An advice nurse was available by telephone 24/7. Additional clinics were scheduled to provide HIV and tuberculosis testing and influenza vaccinations. Plus participation in interviews and brief physical examinations. Nature of the intervention Int2: baseline (n = 202): 34% 6 months (n = 209): 61% 18 months (n = 219): 68% Int2 vs Int1 Logistic regression (adjusted for baseline effects of the site on study outcome): 6 months (n = 77): OR = 1.47 (95% CI 0.39, 5.61) p = 0.567 18 months (n= 92): OR = 4.60 (95% CI 1.31, 16.2) p = 0.017 Mammogram in past 2 years Int1: baseline (n = 50): 60% 6 months (n = 50): 100% 18 months (n = 43): 67% Int2: Baseline (n = 202): 25% 6 months (n = 209): 50% 18 months (n = 219): 54% Int2 vs Int1 Logistic regression (adjusted for baseline effects of Pap test in past year
Int1: baseline (n = 50): 55% 6 months (n = 50): 45% 18 months (n = 43): 27% Men < 65 years: Diff = –5.2% (95% CI –19.9–8.6) Women ≥ 65 years: Diff = +17.5% (95% CI –0.5–35.6) Men ≥ 65 years: Diff = +18.0% (95% CI –2.2–38.0) Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 103
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings the site on study outcome): 6 months (n = 39): OR not reported 18 months (n = 47): OR = 0.56 (95% CI 0.043–7.38) p = 0.661 Cole SR et al. Level II Int1: Procedures (no diet) FOBT kits returned within 15 weeks: (2001)168 Colorectal Mailed stool sample kit without any diet Int1: 396/601 (65.9%) AUSTRALIA Women, Men restrictions + information sheet on bowel Int2: 321/602 (53.3%) 50–69 cancer + reminders at 5 and 10 weeks for Post hoc significance test by CHPPE chi sq = 19.729, Effect of dietary Random sample of SA residents who as of non‐responders df = 1, p = 0.000 restriction on June 1998 were registered in the Int1 vs Int2: participation in faecal Behavioural Epidemiology Database of the Int2: Procedures (diet) Difference = +12.6% (95% CI 7.1%, 18.1%) occult blood test SA DHS who agreed to participate in the Mailed stool sample kit with diet screening for Trial. restrictions (low peroxidase)+ information colorectal cancer sheet on bowel cancer + reminders at 5 and 10 weeks for non‐responders Cole SR et al. Level II Int1: Usual care Participation rates at 12 weeks (2007)39 Colorectal Standard invitation containing health‐
Int1: 237/600 (39.5%) AUSTRALIA Women, Men related messages concerning CRC Int2: 242/600 (40.3%) 50–74 prevention and the value and ease of Int3: 216/600 (36.0%) An advance From the South Australian (SA) electoral screening + FIT kit + reminder at 6 weeks Int4: 290/600 (48.3%) notification letter roll (Adelaide postcodes) for non‐responders increases Int2 vs Int1: no stat sig diffs participation in Int2: Message framing (risk) Int3 vs Int1: no stat sig diffs colorectal cancer Invitation with additional messages related Int4 vs Int1: unadjusted RR = 1.23 (95% CI 1.06–1.43), screening to CRC risk + FIT kit + reminder at 6 weeks adjusting for sex, age group for non‐responders Int3: Message framing (advocacy) Invitation with additional messages from 'like' others advocating CRC screening + FIT kit + reminder at 6 weeks for non‐
Author (Publication year) Country Title VOL 3– 104 Knowledge translation: a review of strategies to increase participation in cancer screening
Cole S R et al. (2002)189 AUSTRALIA Participation in screening for colorectal cancer based on a faecal occult blood test is improved by endorsement by the primary care practitioner Author (Publication year) Country Title Nature of the intervention Findings responders Int4: Invitations and reminders Advance notification – letter introducing standard invitation messages + followed in 2 weeks standard invitation as in Int1 Level II Int1: Usual care Participation rates at 12 weeks: Colorectal Invitation sent from central screening Int1: 192/600 (32.0%) Women, Men service without any indication that GP was Int2: 228/600 (38.0%) > 50 involved + bowel information sheet + Int3: 244/600 (40.1%) SA residents: 2 groups from GP practices in questionnaire (include consent form) + southern suburbs, and 1 group from stool sample collection kit + reminder at 6 Int2 vs Int1: chi sq = 4.45, p = 0.035 federal electoral roll based on postcode weeks for non‐responders OR = 0.77 (95% CI 0.60–0.98) (urban) Int2: Message framing (impersonal) Int3 vs Int1: chi sq = 9.74, p = 0.002 Invitation from central screening service OR = 0.69 (95% CI 0.54–0.87) and endorsed impersonally by person's medical practice + bowel information sheet Int2 vs Int3: chi sq = 0.786, p = 0.376 + questionnaire (include consent form) + stool sample collection kit + reminder at 6 weeks for non‐responders Int3: Message framing (GP endorsement) Invitation sent on medical practice letterhead indicating that screening was endorsed and signed by person's GP + bowel information sheet + questionnaire (include consent form) + stool sample collection kit + reminder at 6 weeks for non‐responders Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 105
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level II Colorectal Women, Men 50–69 SA urban residents: randomly selected from electoral roll. Nature of the intervention Findings Int1: Procedures Participation rates at 12 weeks: Mailed invitation from central screening Int1: 240/606 (39.6%) services + information sheet on colorectal Int2: 185/606 (30.5%) cancer screening + questionnaire (including Int3: 142/606 (23.4%) consent) + InSure (2 stools using Brush (simplified process) + no diet or drug Int1 vs Int2 RR = 1.297 (95% CI 1.111–1.515) restrictions) + reminders at 6 weeks for Int1 vs Int3 RR = 1.655 (95% CI 1.393–1.967) non‐responders Int2 vs Int3 RR = 1.276 (95% CI 1.059–1.527) Int2: Procedures Mailed invitation from central screening services + information sheet on colorectal cancer screening + questionnaire (including consent) + FlexSure OBT (3 stools using spatula + no diet or drug restrictions) + reminders at 6 weeks for non‐responders Int3: Procedures Mailed invitation from central screening services + information sheet on colorectal cancer screening + questionnaire (including consent) + Hemoccult SENSA (3 stools using spatula + diet and drug restrictions) + reminders at 6 weeks for non‐responders Consedine NS et al. Level III‐2 Int1: Message framing (gain) Main effect for framing condition was not significant (2007)81 Breast Gain framed telephone message (5‐minute at 6 months F(2,97) = 1.18, but there was an effect USA Women script) + mailed pamphlet + follow‐up (6 for race, F(1,97) = 6.66, p < 0.05 50–70 and 12 months) i.e. Black women reporting more screening, but was Breast screening in Low‐income, low‐screening women living qualified by interaction with income, F(1,97) = 3.66, response to gain, loss, in Brooklyn, New York – drawn from health p = 0.058 (higher income reported more screening – Author (Publication year) Country Title Cole SR et al. (2003)169 AUSTRALIA A randomised trial of the impact of new faecal haemoglobin test technologies on population participation in screening for colorectal cancer VOL 3– 106 Knowledge translation: a review of strategies to increase participation in cancer screening
Costanza ME et al. (2000)108 USA Promoting mammography Author (Publication year) Country Title and empowerment framed messages among diverse, low‐
income women Nature of the intervention Int2: Message framing (loss) Loss framed telephone message (5‐minute script) + mailed pamphlet + follow‐up (6 and 12 months) Int3: Message framing (empowerment) Empowerment framed telephone message (5‐minute script) + mailed pamphlet + follow‐up (6 and 12 months) Level II Int1: Counselling (telephone) Breast Annual screening reminder + baseline Women survey + barrier specific telephone 50–80 counselling + follow‐up data collection Members of two independent physician (3 years) associated (IPA) model HMOs serving central and south eastern Massachusetts, Int2: Education (provider) and their primary care physicians (PCPs), A five‐hour educational curriculum on who: i) never had a mammogram (never mammography counselling and clinical users); ii) had prior mammograms but breast examination none for the past 24 months (former users); or iii) had a mammogram within the Int3: Usual care prior 24 months but not another one Usual care reminders to women and within 24 months prior to that (recent primary care physicians users). Eligible PCPs were family, internal or general practitioners who: i) were either solo practitioners or members of a group of ≤ 10 physicians; ii) had a minimum of 16 eligible women; and iii) had at least one woman patient > 64 years. Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) clinics. Regular mammogram users at final survey: Int1: 235/535 (44%) Int2: 257/587 (44%) Int3: 201/481 (42%) Post hoc CHPPE significance tests Int1 vs Int3: chi sq = 0.472, df = 1, p = 0.492 Int2 vs Int3: chi sq = 0.472, df = 1, p = 0.512 Regular use was strongly associated with baseline history of utilization: 15% of never user at baseline became regular users compared to 64% of recent users. The authors concluded that the study demonstrated that Int1 is moderately effective in encouraging former users (women whose utilization had lapsed by ≥ 24 months) to become regular users (OR = 1.46, 95% CI 1.00–2.12, p = 0.05). but only for clinical breast examination and mammography. At 12 months: Int2 and Int3 increased screening rates than previously, but significant interaction between time, condition, income, Wilks = 3.97, p < 0.05, as well as condition, type of screen, time, Wilks = 2.21, p = 0.07, and between condition, type of screen and income, Wilks = 2.78, p = 0.06. Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 107
Crotta SB et al. (2008)35 ITALY Strategies to enhance participation in a screening program for colorectal cancer in a mountain community de Jonge E et al. (2008)27 BELGIUM A quasi‐randomized trial on the Participation rates: Int1: 387/1060 (36.5%) Int2: 558/966 (57.7%) Int1 vs Int2 OR = 2.40 (95%CI 2.10–2.88) RR = 1.59 (95% CI 1.50–1.71) Int1: Invitations and reminders (mailed) Preliminary letter signed by doctor + invitation letter, FOBT kit, questionnaire + instructions to return the questionnaire and FOBT + reminder telephone call at 15 days to non‐responders Int2: Invitations and reminders (face‐to‐
face) Preliminary letter signed by doctor + trained non‐health professional visited and supplied same documentation as Int1 + reminder telephone call at 15 days to non‐
responders Int1: Invitations and reminders (mailed) Mailed invitation letters containing an appointment at local municipal office where volunteers distributed kits Incremental participation rates at follow‐up Int2 vs Int1: Younger women (25–41 years) = +3% (95% CI 2.6–3.4) Participation rate: Int1: 9,708/15,780 (61.5%) National Italian survey 47% No statistical test presented Findings Nature of the intervention Level II Int1: No mailed invitation to non‐attendees
Cervical Women Int2: Invitation (mailed) 25–64 Mailed invitation to non‐attendees just Women from province of Limburg who had before the 3 year letter no Pap test taken in past 30 months Level III‐3 Colorectal Women, Men 50–74 Population of regional area. Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) Courtier RM et al. Level II (2002)47 Colorectal SPAIN Women, Men 50–74 Participation in a Attendees of a primary healthcare centre colorectal cancer affiliated to a medical insurance scheme in screening Barcelona. programme: influence of the method of contacting the target population VOL 3– 108 Knowledge translation: a review of strategies to increase participation in cancer screening
Screening rates: Int1: 36/359 (10%) Int2: 39/330 (12%) Int1 vs Int2 RR = 0.85, 95% CI 0.55–1.3 Level II Women 18–67 13 Vietnamese family names: randomly selected from electoral roll with South Brisbane addresses Cervical (n = 653) Int1: 62% Int2: 64% Adjusted OR = 1.18 (95% CI 0.82–1.70), p = 0.38 Analysis of up‐to‐date screening at follow‐up (ITT) Breast (n = 653) Int1: 50% Int2: 52% Adjusted OR = 1.16 (95% CI 0.86–1.57), p = 0.33 Older women (46–61 years) = +10.7% (95% CI 9.8–11.6) All women = +6.4% (95% CI 5.9–6.9) Findings Level of evidence (NHMRC) Type of cancer Nature of the intervention Participants (sex), Participants (age) Participants (cohort) According to Cancer Registry – assigned to either Intervention or Control Group. Baseline participation recorded in year before intervention to determine its effect. 16 age‐specific units studied. September 1994: Media campaign in region – Vietnamese radio and newspapers, flyers distributed to community venues, prevention talks in Vietnamese, GPs notified of campaign through division of GP. November 1994: Int1: Invitation (letter) Personal letter written in Vietnamese inviting them to undertake cervical screening Int 2: No letter Dietrich AJ et al. Level II Int1: Invitation and reminders (telephone) (2007)117 Breast, Colorectal, Cervical Affinity's clinical outreach program for USA Women mammography + the prevention care 40–69 management intervention (mailed Translation of an Enrolled in the Affinity (a Medicaid educational material on breast, cervical efficacious cancer‐
managed care organisation) for at least 12 and colorectal cancer screening) + short screening months, overdue for at least 1 of the telephone recommendation to discuss intervention to targeted cancer screening tests and a cervical and colorectal screening with their women enrolled in a patient of one of the six participating healthcare professional Medicaid managed Community Health Centres. Author (Publication year) Country Title effectiveness of an invitation letter to improve participation in a setting of opportunistic screening for cervical cancer Del Mar C et al. (1998) 30 AUSTRALIA Do personalised letters in Vietnamese increase cervical cancer screening among Vietnamese women? Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 109
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Colorectal (n = 309) Int1: 25% Int2: 32% Adjusted OR = 1.69 (95% CI 1.03–2.77), p = 0.04 Analysis of up‐to‐date screening at follow‐up (ONLY women reached by telephone) Breast (n = 271) Int1: 71.2% Int2: 71.6% Adjusted OR = 1.39 (95% CI 0.88–2.20), p = 0.16 Int2: Counselling (telephone) Affinity's clinical outreach program for mammography + the prevention care management intervention (mailed educational material on breast, cervical and colorectal cancer screening) + telephone assessment of barriers/assistance in overcoming barriers/scheduling assistance and appointment reminders) Cervical (n = 271) Int1: 72.0% Int2: 77.2% Adjusted OR = 1.86 (95% CI 1.08–3.21), p = 0.03 Findings Nature of the intervention Colorectal (n = 160) Int1: 27.5% Int2: 34.2% Adjusted OR = 1.84 (95% CI 0.95–3.58), p = 0.07 Dietrich AJ et al. Level IV Int1: Coaching Mammography (baseline to follow‐up) (2006)130 Breast, Colorectal, Cervical Brochure on prevention plus follow‐up Int1: 0.58 to 0.68 USA Women telephone call to answer questions + Int2: 0.60 to 0.58 50–69 received series of telephone support calls Int1 vs Int2: diff = +0.12 (95% CI: 0.06–0.19) Telephone care Ethnic minority and low‐income women from trained Prevention Care Manager management to with low screening rates who visit (Prevention Care Managers received Cervical (baseline to follow‐up) improve cancer community and migrant health centres in training addressing barriers, role plays, Int1: 0.71 to 0.78 screening among low‐ New York City. providing support etc) – average of 4 calls Int2: 0.70 to 0.70 income women – A Int1 vs Int2: diff = +0.07 (95% CI 0.01–0.12) randomized, Int2: Usual care plus brochure controlled trial Usual care plus brochure on prevention Author (Publication year) Country Title care organization VOL 3– 110 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) and single follow‐up telephone call to answer questions Nature of the intervention Findings Colorectal (baseline to follow‐up) Int1: 0.39 to 0.63 Int2: 0.39 to 0.50 Int2 vs Int1: diff = +0.13 (95% CI 0.07–0.19) Dinshaw K et al. Level III‐1 Medical social workers conducted Health Screening participation rates were presented for the (2007)197 Breast, Cervical Education Programmes at both arms, Int1 only; there was no comparative analysis INDIA Women covering risk factors, signs and symptoms, between Int1 and Int2 ≥ 35 methods of early detection as well as Determinants of Community residents in 10 Mumbai slum treatment modalities for breast and compliance in a areas cervical cancers. They also educated the cluster randomised women regarding breast self‐examination.
controlled trial on screening of breast Int1: Coaching and cervix cancer in Personal invitations to screening by trained Mumbai, India – 1. outreach workers Compliance to screening Int2: No invitation Dolan NC et al. Level II Round 1: All completed a baseline study Mammography screening rates (no differences for (1999)156 Breast questionnaire at check‐in; a screening Phases 1 and 2) USA Women mammography recommendation prompt 3 months: Int1 56%; Int2 43% (p < 0.001) ≥ 50 sheet attached to chart; at checkout they 6 months: Int1 61%; Int2 49% (p < 0.001) Impact of same‐day Consecutive patients at an urban academic were asked whether they planned to get 12 months: Int1 66%; Int2 56% (p = 0.003) screening general internal medicine practice with a the recommended mammogram and mammography hospital mammography centre located 3 where they intended to get it. Then Rates of adherence to physicians screening availability: Results of blocks away. Exclusion criteria included: i) randomised. mammography recommendations a controlled clinical presentation for an acute care visit; ii) had No mammograms in past 5 years: trial obtained a mammogram in the previous 12 Int1: Procedures Int1: 39% months; iii) had a history of breast cancer; Participants offered the opportunity to Int2: 20% iv) had an active breast symptom at the obtain the screening mammogram Int1 vs Int2: Author (Publication year) Country Title Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 111
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) time of the visit; v) did not receive a recommendation for a screening mammogram during the visit. Level II Breast Women Author (Publication year) Country Title Duan N et al. (2000)113 USA RR = 1.63 (95% CI 1.04–2.57) (p = 0.02) 1–2 mammograms in past 5 years: Int1: 57% Int2: 38% Int1 vs Int2: RR = 1.39 (95% CI 1.15–1.69) (p < 0.001) > 3 mammograms in past 5 years: Int1: 67% Int2: 59% Int1 vs Int2: RR = 1.14 (95% CI: 0.97–1.40) (p = 0.11) immediately (procedures) and if accepted then transport was available to take them to the mammography centre Int2: Usual care No further information Round 2: 2 weeks before scheduled appointments, potential participants assigned to intervention group; all completed a baseline study questionnaire at check‐in; a screening mammography recommendation prompt sheet was attached to their charts; at checkout they were asked whether they planned to get the recommended mammogram and where they intended to get it. Int1: Procedures Mailed postcard on screening mammography + notification of availability of same day screening mammography if their physician recommended it. Int2: Usual care Mailed postcard on screening mammography Recruitment, baseline survey, follow‐up survey 12 months later Non‐adherence rate at 12‐month follow‐up amongst those: Adherent at baseline: Findings Nature of the intervention VOL 3– 112 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) 50–80 Members of 30 churches (8 predominantly Latino, 12 predominanatly African American, 10 predominantly White) from the southern area of Los Angeles County. Int1: Counselling (telephone) One session of telephone counselling annually for two years by part‐time peer counsellors Int2: Usual care Nature of the intervention Findings Int1 = 15.8%, Int2 = 23.3%, Diff = 7.5%, one‐sided p = 0.029 Non‐adherent at baseline: Int1 = 34.8%, Int2 = 37.4%, Diff = 2.6%, one sided p = 0.324 Eaker S et al. Level III‐2 Three successive interventions tested: Interventions: (2004)29 Cervical (1) Modified invitation (brochure with (1) Modified invitation did not increase attendance vs SWEDEN Women standard invitation) vs standard invitation standard invitation letter (diff = +1.3% 95% CI –0.3–
25–59 letter (printed on ordinary white +2.9) A large population‐
Residents of Uppsala County who had not stationery) based randomized had a Pap test in previous 3 years and had (2) Reminder letter (same as standard (2) Reminder letter increased attendance by 9.2% controlled trial to not asked to be excluded from call‐recall invitation but included info that women (95% CI 7.9, 10.5) vs women who did not receive increase attendance system had received a prior invitation) to women reminder letter at screening for who did not attend after the first cervical cancer intervention vs no reminder letter (3) Increased attendance: 31.4% (95% CI 26.9–35.9) (3) Phone reminder (called by 1 or 2 female research assistants who described test and offered to schedule appt) to women who did not attend after reminder letter vs no phone reminder Earp JA et al. Level IV Recruitment, baseline and follow‐up Differences in self‐reported mammography use in (2002)119 Breast interview two years later past 2 years from baseline to follow‐up adjusted for USA Women age, medical visits, physician recommendation for ≥ 50 Int1: Coaching mammography and perceived susceptibility to breast Increasing Use of Sample of African American and white Recruitment + baseline data collection + cancer. Mammography residents in 5 intervention and 5 community trial using lay health advisors (i) Overall Among Older, Rural comparison counties in North Carolina. supplemented by a limited number of Int1 = +17% African American other activities + follow‐up data (2 years) Int2 = +11% Author (Publication year) Country Title Maintaining mammography adherence through telephone counseling in a church‐based trial Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 113
English KC et al. (2008)132 USA A socioecological approach to improving mammography rates in a tribal community Fang CY et al. (2007) 176 USA A multifaceted intervention to increase cervical cancer screening among underserved Level II Cervical Women ≥ 18 Women recruited from 2 Korean community centres – mainly serving low income, uninsured, recent immigrants Level III‐2 Breast Women > 50 Ramah Band of Navajo Indians in West Central New Mexico. Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) Women: Results From a Community Trial Findings Int1: Education, Coaching Pap screening rates (baseline to 6 months): 2‐hour small‐group education session Int1: 6/52 (11.5%) to 43/52 (82.7%) focussing on cervical cancer and benefits of Int2: 11/50 (22.0%) to 11/50 (22.0%) screening AND patient navigation by bilingual Korean health educators. Int1 vs Int2: Chi sq (1) = 41.22, p < 0.001 Int2: Control Group 2‐hour general health education session, Int1 vs Int2 (adjusted): diff = +7% (95% CI 0%–14%) p = 0.05 (ii) Low income Int1 = +22% Int2 = +11% Int1 vs Int2 (adjusted): diff = +11% (95% CI 2%–21%) p = 0.02 (iii) High Income Int1 = + 3% Int2 = +9% Int1 vs Int2 (adjusted): diff = 1% (95% CI –10%–,11%) p = 0.92 Community‐based participatory All the women who were approached during the Intervention which included community home visit recruitment phase (n = 30) and 9 who engagement, arranging lunch, transport, learned about the project anecdotally from project health education, appointment scheduling, staff or other community women participated in the language interpretation and social support. mammography days (i.e. were screened). Female community health representatives served as primary outreach workers to promote participation. Int2: Usual care Recruitment + baseline data + follow‐up data (2 years) Nature of the intervention VOL 3– 114 Knowledge translation: a review of strategies to increase participation in cancer screening
Level II Colorectal Women, Men 50–75 Patients of participating GPs. Guaiac = 3604 invitations (hospital 1797; GP 1 807) Immunochemical = 3716 invitations (hospital 1858; GP 1858) Federici A et al. (2005)165 ITALY The immunochemical faecal occult blood test leads to higher compliance than the guaiac for colorectal cancer screening programmes: a cluster randomized controlled trial Federici A et al. Level III‐2 (2006)166 Colorectal ITALY Women, Men 50–74 The role of GPs in Patients of participating GPs. increasing compliance Guaiac = 3604 invitations (hospital 1797; to colorectal cancer GP 1807) screening: a Immunochemical = 3716 invitations randomised (hospital 1858; GP 1858) controlled trial (Italy) Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title korean women by trained Korean health educator, including cancer screening – as well as written information included cervical and pap test free screening sites. GPs were randomly assigned to one of two types of FOBT: i) Guaiac FOBT ii) Immunochemical FOBT. Int1: Procedures Letter signed by the GP inviting them to go to either the GP's office for testing instructions Int2: Procedures Letter signed by the GP inviting them to go to the gastroenterology centre of the hospital for screening. GPs were randomly assigned to one of two types of FOBT: i) Guaiac FOBT ii) immunochemical FOBT. Int1: Procedures Letter signed by the GP inviting them to go to either the GP's office for testing instructions. Int2: Procedures Letter signed by the GP inviting them to go to the gastroenterology centre of the Nature of the intervention Screening participation: Int1: 50.2% (Guaiac = 46.0%; Immunochemical = 54.3%) Int2: 16.3% (Guaiac = 14.7%; Immunochemical = 17.9%) Int1 vs Int2: RR of participation based on the invitations issued=3.40 (95% CI 3.13–2.70) Lower compliance was obtained by GPs with > 25 patients visited/day (OR 0.74, 95% CI 0.57–0.95) and those who incorrectly recommended screening of Screening participation: Guaiac: overall = 30.4% Int1 = 46% Int2 = 14.7% Immunochemical: overall = 35.8% Int1 = 54.3% Int2 = 17.9% Guaiac vs immunochemical: RR = 1.20 (95% CI 1.02–1.44) Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 115
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention hospital for screening. Feldstein AC et al. Level IV 3 time cohorts stratified into 3 groups: Pre‐
(2009)48 Breast reminder period (2004); Post‐reminder USA Women implementation phase (2006); Post‐
≥ 42 reminder maintenance phase (Jan–Jul Effect of a Members of non‐profit group model HMO 2007). multimodal reminder operating in southern Washington and program on repeat northern Oregon (15 medical clinics). The 3 time cohorts were further stratified mammogram Criteria included: Women who had into 3 groups: screening previously had a mammogram but were 20 (1) Target group months past last mammogram. Women 50–69 to whom mammograms were recommended (reminders sent in 2006 and 2007) (2) Comparison group 1: Women 42–49 to whom mammograms were only recently recommended (3) Comparison group 2: Women ≥ 70 to whom mammograms were conditionally recommended depending on life expectancy If no appointment for mammogram made at: 21 months – received automated telephone reminder (on behalf of patient's primary care provider) and instructions on how to make an appointment. 22 months – received 2nd automated call delivered by HMO staff under a protocol that delivered appropriate intervention Author (Publication year) Country Title Target group: Year 1 – 1 month after each program step: 9.9% completed mammogram after postcard; 24% after automated call 1; 37% after automated call 2; and 47% after the live call. In pre‐intervention period 25% in target group completed mammogram by 24 months; 47% in year 1; and 53% in year 2. This is in contrast to the comparison groups who did not improve in post‐reminder periods. Women in target group were 1.51 times more likely (CI = 1.40–1.62; p < 0.001) in the post‐reminder period, and 1.81 times more likely (CI = 1.65–1.99; p < 0.001) in post‐reminder maintenance period to have mammogram when compared to comparison groups. colorectal cancer (OR 0.75 94%CI 0.59–0.97) Findings VOL 3– 116 Knowledge translation: a review of strategies to increase participation in cancer screening
Level II Cervical Women 25–69 Participants were approached at shopping areas of suburban Melbourne in postcodes with a high percentage of under‐screened women. Fernbach M (2002)212 AUSTRALIA The Impact of a Media Campaign on Cervical Screening Knowledge and Self‐
efficacy Fishman P et al. (2000)40 USA Cost‐effectiveness of strategies to enhance mammography use Level III‐2 Breast Women 50–79 HMO breast cancer screening program enrollees who did not schedule for screening within 2 months after recommended by letter. Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title Screening rates: Baseline: 81%; Phase 1: 81%; Phase 2: 82%; no stat sig diffs Findings Int1: Invitations and reminders (mailed) At 12 months, women receiving Int2 and Int3 were Mailed postcard reminder of importance of significantly more likely to have scheduled a screening and invitation to schedule mammogram than women in Int1; no significant difference between Int2 and Int3. Int2: Invitations and reminders (telephone) Reminder call acknowledging that not There was a graph in the document showing the scheduled for mammogram and that effectiveness results but no numbers were opportunity still existed presented. Authors concluded that a postcard reminder may be appropriate for those who have Int3: Counselling (telephone) previously had a mammogram, but a reminder call components. Components included computer program updates, dual‐analyst review. And mechanisms for patients and staff to provide feedback for data and program quality improvement. 23–24 months – Women in target group who had not made an appointment were given to local healthcare teams so follow‐
up live calls would be made (script and training provided to callers) – up to 2 messages left requesting call back. Int1: Media intervention Implemented in two phases; data from different samples of women collected at baseline, after phase 1 and after phase 2 Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 117
Self‐reported tracking data (i) Mammogram: Baseline 2727 women enrolled, 48% screened; 2 years later 2552 enrolled, 62% screened (58% of enrollees) (ii) Pap test: Baseline 2727 enrolled 55% screened; may be justified for those who have never been screened. Findings Int2: Patient intervention CRC screening pamphlet + videotape designed for motivating and educating low literacy patients + simplified set of instructions for completing the FOBT. Fouad MN et al. Level IV Community action plan: (2006)141 Breast, Cervical (i) Establish and sustain a core working USA Women group (CWG) in each targeted county; ≥ 40 (ii) Provide CWG members with the A community‐driven African Americans in 3 urban and 6 rural appropriate training and skills needed to action plan to counties in the Alabama Black Belt region. promote breast and cervical cancer eliminate breast and screenings among women in their Nature of the intervention Int1: FOBT completion rates in patients whose physicians: Attended NO feedback sessions: 17/95 (17.9%) Attended SOME feedback sessions: 269/891 (30.2%) p = 0.01 Int2: FOBT completion rates in patients who: DID NOT watch video: 218/728 (30.0%) DID watched video: 68/258 (26.4%); p = 0.30 Int1 vs Int2: Not presented Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Motivational Telephone Call which engaged woman in discussion re: affect, attitude, facilitating conditions, perceived breast cancer risk Fitzgibbon ML et al. Level IV Int1: Provider intervention (2007)101 Colorectal Introductory session just prior to the USA Women Men initiation of the study. Every 4 to 6 ≥ 50 months, all providers were invited to Process evaluation in Non‐compliant patients at two outpatient attend 1‐hour feedback sessions during an intervention clinics in a Veterans Affairs Medical Centre. which they received information on the designed to improve CRC screening recommendations for their rates of colorectal clinic and patient adherence to cancer screening in a recommendations. Each provider also VA medical center received confidential information on their individual recommendations and adherence in a sealed envelope. Author (Publication year) Country Title VOL 3– 118 Knowledge translation: a review of strategies to increase participation in cancer screening
Fox SA et al. Level III‐1 Level III‐2 Breast Women 50+ Women from a large city of Ohio who did Phase 2: Administer pre‐test not report mammography screening in the questionnaire, 20–30 minute face‐to‐face past 2 years education & providing informational materials, schedule appointment in 4 weeks time, provide token and $10 voucher Fowler BA et al. (2005)174 USA Collaborative breast health intervention for african american women of lower socioeconomic status Findings All participated in pre (1991) and post
Phase 6: Final contact with participants 8 weeks after phase 2. Administer post‐test.
Phase 5: Telephone contact from the project director to determine screening status Phase 4: Receipt of screening via community outreach Phase 3: Contact women within 2 weeks of the previous contact Int1 vs Int2 communities; 2 years later 2552 enrolled, 66% screened (62% of (iii) Provide coalition members with the enrollees). technical assistance needed to develop and implement mini‐grants focused on Medicare surveillance data for mammography enhancing breast and cervical cancer 1997–98: 17% disparity between African American screening among their priority audience. and white women in counties; 2001–03: 11% disparity between African American and white women in counties Int1: Multiple component interventions Mammography screening rate before (self‐reported) Phase 1: Establish initial contact through and after intervention (n = 68): 55% vs 65% influential women, local communication media & arrange follow‐up. No statistical test presented Nature of the intervention Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title cervical cancer disparity: successes and limitations Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 119
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Breast Women ≥ 64 Randomly selected group of Medicare beneficiaries who were residents of three sites in southern California (2 interventions and 1 control). Inclusion criteria included: i) non‐institutionalisation; ii) no history of breast cancer; iii) able to complete a 45 minute bi‐lingual telephone interview or mailed questionnaire. Nature of the intervention Findings (1993) test telephone and/or face‐to‐face Adjusted odds ratios from weighted logistic interviews regressions predicting mammogram use during the past two years Int1 Targeted 2‐page bilingual mailing White (n = 532): explaining the new subsidised Medicare baseline* OR = 0.60 (95% CI 0.30–1.18) screening benefit and outlining low‐cost follow‐up OR = 1.04 (95% CI 0.63–1.72) screening opportunities (1991) Black (n = 300): Int2: Usual care baseline OR = 0.40 (95% CI 0.14–1.21) follow‐up OR = 0.97 (95% CI 0.97–4.02) *Note the survey at baseline was conducted in the spring of 1991, and the Hispanic (n = 90): Medicare benefit was implemented on 01 baseline OR = 0.87 (95% IC 0.09–8.82) Jan 1991. So there was some awareness at follow‐up OR = 2.33 (95% CI 1.01–5.37) baseline. Friedman LC et al. Level III‐2 Int1 Compliance with FOBT use (2001)97 Colorectal Watched educational video, then USA Women, Men completed questionnaire prior to Int1+2: 48/110 (43.6%) ≥ 50 appointment with physician Int3: 18/50 (36.0%) Compliance with fecal Predominantly African American, low Int2 Total: 66/160 (41.3%) occult blood test income outpatients at community clinic in Completed questionnaire, watched video screening among low‐ Houston who had not had FOBT in (used African American peer educators and Not significant income medical preceding 12 months, and no history of health professional), and then completed outpatients: A CRC. portion of the questionnaire again prior to randomized appointment with physician controlled trial using a videotaped Int3 intervention Only completed questionnaire prior to appointment with physician Author (Publication year) Country Title (2001)154 USA Targeted mailed materials and the Medicare beneficiary: increasing mammogram screening among the elderly VOL 3– 120 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level III‐2 Cervical Women 15–64 Indigenous women from rural villages in Oaxaca where screening rates are low Author (Publication year) Country Title Givaudan M et al. (2008)134 MEXICO Enhancement of underused cervical cancer prevention services in rural Oaxaca, mexico Findings All groups were given FOBT kits (after appointment with physician) by one of the nursing staff. Appointments were made for return of kits and for a follow‐up visit to discuss test results. 3 months later, laboratory results were checked to see compliance rates. Int1: Community intervention Pap tests in Int1: Workshops (8 x 2hour weekly session with 1 year pre‐intervention: 1074/4865 (22%) groups of 15–25 women) – conducted by 1 year post‐intervention: 1306/4865 (27%) local women (health promoters) who received 1‐week training. The workshops Reported significant increase in screening in Int1 vs are highly interactive and include Int2: t(11) = 2.66, p=.02 discussions and role plays, presentation of BUT no individual figure was presented. information. There was also workshop for men (3x2‐hour sessions – include info on Pap test and interaction with partner. Another module for healthcare providers (2x4‐hour sessions) where partners could discuss and reflect on their role and interaction with women. Also had promotion campaign. Int2: Control group Women from 6 similar communities in region. Evaluation of program: (1) comparing number of tests recorded before and after program implementation, and (2) Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 121
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings questionnaire assessing knowledge, opinions, attitudes, behaviour of women Giveon S et al. Level III‐2 Int1: Counselling (face‐to‐face) Mammography performed at 12 months in the (2000)116 Breast Family practitioner discussed breast cancer population aged 50–74 years: ISRAEL Women screening + follow‐up survey Int1: 15/60 (25%); 50–74 Int2: 2/40 (5%) Patient adherence to Sample of attendees at a health Int2: Control group family practitioners' management organisation (HMO) primary Family practitioner did not discuss breast Int1 vs Int2 recommendations for care centre during a 12‐month period cancer screening + follow‐up survey RR = 5.38 (95% CI 1.30–22.29), p = 0.01 breast cancer screening: A historical cohort study Gotay CC et al. Level IV Int1: Community intervention Compliant with Pap screening (2000)139 Breast, Cervical Three‐year community intervention Int1: (n = 318) USA Women involving education sessions and vouchers baseline = 59%; follow‐up = 67%, significant p < 0.05 ≥ 18 for free mammograms and Pap tests for Int2: (n = 360) Impact of a culturally Hawaiian ancestry residents in a well‐
attendees and a friend. baseline = 63%; follow‐up = 64%, non significant appropriate defined Oahu neighbourhood. intervention on breast Int2: Control community Compliant with mammography and cervical screening Surveys were conducted before the Int1: baseline = 60%; follow‐up = 61% among native intervention (pre‐test) and after the Int2: baseline = 59%; follow‐up = 60% Hawaiian women conclusion of Int1 – about 3 years later (post‐test) Grazzini G et al. Level III‐2 Int1: Invitations and reminders Participation rates presented for first round (2008)36 Colorectal Invitation to undergo immunochemical (invitees = 25,428) subjects and repeat screening ITALY Women Men FOBT without dietary restrictions every round (invitees = 63,369) 50–70 two years. Int1: first round 38%, repeat round 42% Cost evaluation in a Subjects living in the Florence district. Int2: first round 15%, repeat round 18% colorectal cancer Exclusion criteria included: i) FOBT test in Int2: Invitations and reminders screening programme last two years; ii) double‐contrast X‐ray in Non‐responders to first invitation received by faecal occult blood last 3 years; iii) total colonoscopy in last 5 a reminder within 6 months. Author (Publication year) Country Title VOL 3– 122 Knowledge translation: a review of strategies to increase participation in cancer screening
Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) test in the district of years; iv) personal history of cancer or Florence adenoma under endoscopy surveillance. Gregory‐Mercado K Level II et al. Breast (2007)157 Women USA 40+ Low income, uninsured, and under‐served A combined approach people enrolled in the NBCCEDP from 1 to women's health is January 2000 through to 30 June 2003, associated with a who had a normal result on an initial greater likelihood of mammography screening during this repeat period mammography in a population of financially disadvantaged women Hancock L et al. Level IV (2001) 136 Cervical AUSTRALIA Women 18–70 Effect of a community People living in the 10 intervention group action intervention on towns and in the 10 control group towns. cervical cancer screening rates in rural Australian towns: The CART project Rescreening rate at 18‐month follow‐up: Grp1: 34/54,780 (34%) Grp2: 58/13,742 (58%) Findings Int1: Community intervention i) Distribution of pamphlets/display of posters; ii) personal visit/talk/ information session; iii) cancer awareness seminar with guest speakers, iv) media release; v) invitation to Pap tests; vi) outreach Pap test clinic; vii) Pap test reminder/checklist; viii) support of local christian women's association. The main intervention message emphasised two‐yearly Pap tests for all women ages 18 to 70. Int2: Control No community intervention in Under‐screened women (those who had a Pap test more than 2 years ago): Intervention town women more likely to have a Pap test than women in the control towns. Over‐screened women (those who had a Pap test less than 1 year ago): Intervention town women more likely to screen appropriately but no information presented for control towns. Overall no differences in screening and diagnostic test rates between control and intervention towns. Grp2 vs Grp1: Grp2: Unadjusted OR = 2.71 (95% CI 2.61–2.82) Participants of the NBCCEDP and the Well‐ Adjusted OR = 2.75 (95% CI 2.60–2.82) Integrated Screening and Evaluation for Women Across the Nation (WISEWOMAN) program Grp1: Participants of the National Breast and Cervical Cancer Early Detection Program (NBCCEDP) Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 123
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level IV Colorectal Women, Men 41–65 Employees of large industrial company in regional England Level III‐2 Breast Women 50–64 Residents in six areas of Dublin that had not been 'screened' Author (Publication year) Country Title Hart AR et al. (2003)151 UK An industry based approach to colorectal cancer screening in an asymptomatic population Hayes C et al. (1999)26 IRELAND The impact of reminder letters on attendance for breast cancer screening communities matched for demographic, population density, average summer temperature etc. 1) The personnel departments of the engineering company sent a letter to employees explaining the purpose of bowel cancer screening, and were offered a free Haemoccult test pack. 2) Scheme was advertised at worksite with posters and medical department answering any enquiries. 3) Employees who chose to participate were sent a Haemoccult pack with instructions – no dietary restrictions. 4) Kits were returned to a general hospital where they were analysed by single investigator. Letter 1: Computer generated letter inviting participants to attend for screening
Letter 2: Second letter sent 6 weeks later to those who did not respond to the first invitation Letter 3: Third letter sent further 6 weeks later to those who did not respond to the first or second invitation Nature of the intervention Letter 1: Eligible 1310 screened before second letter 795 attended (60.7%) Letter 2: Eligible 515 screened before third letter 92 attended (17.9%) Letter 3: Eligible 423 screened (unstated time period) 32 attended (7.6%) Total screened: 919 attended (70.1%) Free medical cover: Letter 1: 49%, Letter 2: 18%, Letter 3: 2%, Total: 59% Overall compliance was 25.4%, with similar rates for men (25.0%) and women (32.0%) – Chi sq = 3.0). For men, compliance was highest for 51–60 years (30.5%) Chi sq > 11.6, p < 0.001. For women, no statistical differences in different age groups Chi sq < 2.08. Managers had higher compliance rate than clerical and blue collar workers (28.6% vs 23.5%, Chi sq = 5.6, p > 0.02). Findings VOL 3– 124 Knowledge translation: a review of strategies to increase participation in cancer screening
Level III‐2 Colorectal Hou SI et al. (2004)153 Study involved:
i) Distribution of consent form, pre‐test Level IV Int1: Invitation and reminder (mailed) Cervical Community‐based approach – the Women authority (local health authority, regional 30–60 corporation or third party) invites and Patients of a country‐wide sample of family sometimes reminds the women to attend practices (n = 122). Exclusion criteria screening with a letter. included: i) test within 1 year; ii) hysterectomy; iii) already being followed Int2: Invitation (mailed) reminder (mailed up for an abnormal smear; iii) non medical or telephone) reason (social or psychosocial); iv) Family practice‐based approach – the pregnancy, recent delivery, breastfeeding. family practice invites and reminds the women with letter or phone call. Int3: Invitation (mailed) reminder (mailed or telephone) Combination approach – the authority invites the women and the family practice, and reminds them with a letter or phone call. Hermens RP et al. (2000)49 NETHERLANDS Attendance to cervical cancer screening in family practices in The Netherlands Nature of the intervention Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title Private medical insurance: Letter 1: 64%, Letter 2: 18%, Letter 3: 10%, Total:74% Post hoc CHPPE significance test: Type of CoverTotal screened: Chi sq = 24.872, df = 1, p = 0.000 Overall attendance rates ≤ 45 years: Int1 = 53% (95% CI 49%–57%) Int2 = 68% (95% CI 65%–71%) Int3 = 62% (95% CI 60%–64%) > 45 years: Int1 = 47% (95% CI 45%–50%) Int2 = 58% (95% CI 56%–60%) Int3 = 60% (95% CI 58%–62%) Impact of reminder on attendance rates Attendance after first invitation ≤ 45 years: Int2 = 60% (95% CI 56%–63%) Int3 = 51% (95% CI 48%–53%) Attendance after reminder ≤ 45 years: Int2 = 8% (95% CI 6%–10%) Int3 = 11% (95% CI 10%–13%) Attendance after first invitation > 45 years: Int2 = 51% (95% CI 49%–53%) Int3 = 49% (95% CI 47%–51%) Attendance after reminder > 45 years: Int2 = 7% (95% CI 6%–8%) Int3 = 11% (95% CI 10%– 12%) Single group of participants. Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 125
Hurdle DE (2007)89 USA Breast cancer prevention with older women: A gender‐
focused intervention study Husaini BA et al. (2002)96 USA The effect of a church‐based breast cancer screening Level IV Breast Women ≥ 40 African Americans in the metropolitan Nashville, Tenessee area. Level III‐2 Breast Women 66–75 Older women living in a north‐western American city recruited from community‐
based organisations for seniors. Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) TAIWAN Women, Men ≥ 40 Home‐administered Employees from 10 different worksites in fecal occult blood test Taiwan for colorectal cancer screening among worksites in Taiwan Int1: Full educational program 2 video presentations, Q&A session, trained personnel to teach about BSE, risk, screening, and services available. Int2: Partial program Only videos to communicate the At 2–4 week follow‐up, screening rate was 76% (n = 284), but 7 did not complete all 3 tests i.e. rate was 74% (277/375). 29.9% reported previously having FOBT. survey, screen brochure, home‐
administered kit, instructions, test result record card in Chinese. ii) Requested to complete pre‐test survey, conduct stool test using kit, and record results on result card iii) After 2–4 weeks, asked to return their test result cards and surveyed iv) Doctors and nurses were available at worksites to answer questions. Hotline from collaborating hospital was also provided for consultations and education. v) Primary outcomes were FOBT screening rate and screening intention in coming year. Int1: Education + message framing 2 one‐hour face‐to‐face sessions, held one week apart – mini lectures, demonstrations, activities, discussions, audiovisuals Int2: No intervention Baseline interviews conducted before any intervention was conducted and then follow‐up interviews at 3 months later. Mammogram obtained in last year: Int1: 67%, Int2: 70%, Recent mammography in post‐intervention survey (unadjusted for baseline rate) Int1: 18/20 Int2: 30/64 Int1 vs Int2: Chi sq 11.572, df = 1, p= 0.001. Findings Nature of the intervention VOL 3– 126 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) information Int3: Control group No educational programming Nature of the intervention Findings Int3: 63% Screening rates between baseline and 3‐month follow‐up: Int1: 36%, Int2: 45%, Int3: 24% Int3 vs Int2, significant at p < 0.05 Immonen‐Raiha P et Level II Int1: Fee for mammogram Mammogram rate before and after introduction of al. Breast A fee introduced after 1997 for those in Int1 in 1997: (2001)155 Women the age group 40–49 years and born in 40–49 age group FINLAND 40–49 and 60–69 1950–52, and those in the age group 60– Fee paying: 1750/2632 (66%) (95% CI 65%, 68%) Women living in Turku, Finland 69 years and born in 1929 and 1935 Fee exempted: 5212/5926 (88%) (95% CI 87%, 89%) Customer fee and participation in Int2: Historical control 60–69 age group breast‐cancer No fee for mammogram prior to 1997 Fee paying: 1393/1863 (75%) (95% CI 73%, 77%) screening Fee exempted: 2406/2664 (90%) (95% CI 89%, 91%) Jandorf L et al. Level III‐3 Int1: Coaching Completed FOBT after 3 months: (2005)128 Colorectal Patient navigation consisting of education, Int1: 42.1%, USA Women, Men help with participation, schedule Int2: 25.0% (p = 0.086) ≥ 50 appointments, address barriers and Use of a patient Patients attending a primary care practice concerns regarding the procedures and navigator to increase in East Harlem, New York City, a prompt the patient to make appointments colorectal cancer predominantly medically under‐served, and return FOBT cards screening in an urban ethnically diverse, low SES area. Exclusion neighborhood health criteria included: i) had a FOBT in the past Int2: Usual care clinic year; ii) a sigmoidoscopy or barium enema within the past 3–5 years; or iii) a colonoscopy within the past 10 years. Jean S et al. Level III‐2 Int1: Invitations and reminders (mailed)
There was a statistically significant difference (2005)19 Breast (i) Letter of invitation signed by a regional between the observed and expected screening rates. Author (Publication year) Country Title education program on mammography rates among African‐
American women Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 127
Kidder B (2008)90 USA P.O.W. (protect our women): Results of a breast cancer prevention project targeted to older African‐American women. Percentage up‐to‐date with screening tests (i) Pap test Pre (1992) Int1 = 54.4% Int2 = 43.6% p = 0.002 Post (1996) Int1 = 47.7% Int2 = 37.0% p = 0.002 Pre vs Post Int1 p = 0.052 Int2 p = 0.052 (ii) Mammogram Pre (1992) Int1 =52.6% Int2 =46.6% p = 0.203 Post (1996) Int1 =55.1% Int2 =45.6% p = 0.039 Pre vs Post Int1 p = 0.606 Int2 p = 0.828 All: Pre‐workshop– 112 women identified Int1: as target population received mailed 72% had completed mammogram packet of information about risk reduction, early identification, and treatment Control group: (culturally tailored) and a save‐the‐date 22% had mammogram notice for upcoming event. The increases were statistically significant due to the large numbers included in the study. The rates varied with age and region of residence. Overall the observed expected cumulative probability of screening 12 months after the mailing of the invitations was 20.3 and the observed cumulative probability was 30.2. program physician; (ii) Reminder letter two months later if no screening mammogram has taken place. Comparison with baseline (expected) rate Int1: Community intervention (media education) Pre‐test survey 1992 + phase 1 of media campaign + follow‐up survey 1993 + phase 2 of media campaign +,follow‐up survey 1994 + phase 3 of campaign + post‐test survey 1996 Int2: Control Pre‐test survey 1992 + post‐test survey 1996 Findings Nature of the intervention Level II Breast Women ≥ 60 Older African American women who a panel of recognised community leaders believed could benefit from breast health education. Int1: Workshop 62 from target population attended Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) CANADA Women 50–69 Screening Residents of Quebec who were eligible to mammography have a screening mammogram between participation and May 1998 and the end of June 2000. invitational strategy: the Quebec Breast Cancer Screening Program, 1998–2000 Jenkins CN et al. Level III‐2 (1999)213 Breast, Cervical USA Women Cervical ≥ 18; Breast ≥ 40 Effect of a media‐led Vietnamese residents in two intervention education campaign counties in northern California and two on breast and cervical control counties in southern California. cancer screening among Vietnamese‐
American women VOL 3– 128 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Lam TK et al. Level III‐2 (2003)123 Cervical USA Women ≥ 18 Encouraging Vietnamese Americans recruited through Vietnamese‐American agencies in San Jose, California women to obtain Pap tests through lay health worker outreach and media education Lewis CL et al. Level II (2008)177 Colorectal USA Women, Men 50–75 The uptake and effect Patients at an academic primary care of a mailed multi‐
practice with a number of attending and modal colon cancer resident physicians who care for those 50 screening and older. Those who were patients of intervention: A pilot attending physicians AND had not had a controlled trial colonoscopy in last 10 years, flexible sigmoidoscopy in 5 last years, or FOBT in Author (Publication year) Country Title Ever had a Pap test among those who had not had a Pap test at baseline: Int1: 14/57 (47.8%); Pre‐post comparison p < 0.001 Int2: 33/69 (24.6%); Pre‐post comparison p < 0.01 Post hoc CHPPE statistical significance test; chi sq = 7.224, df = 1, p = 0.007 Findings Int1: Multi‐component intervention
Screening rates: Mailed package – reminder letter signed by Int1: 20/137 (15%) patient's physician, survey (screening Int2: 4/100 (4%) history, personal and family history of polyps etc) to be completed prior to Int1 vs Int2: diff = +11% (95% CI 3%, 18%) p = 0.01 watching video (Decision Aid), Decision Aid (35‐mins long), 2nd survey (interest and Cost per additional patient screened was estimated acceptability of intervention) to be at $94. Uptake was low. completed after watching decision aid, info encouraging return of all materials especially the 1st survey, instructions and workshop. Completed evaluation form re: awareness of early detection. 1 month after workshop women were mailed a gift for participating and postcard inquiring about date of most recent mammogram. Int2: Control group No workshop; mailed evaluation tool with inquiry about most recent mammogram. Media campaign undertaken in the area Int1: LHWO (lay health worker outreach) + ME (media education campaign) Int2: ME (media education campaign) Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 129
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) last 11 months. Each physician excluded patients who deemed too ill to benefit. Lipkus IM et al. (2005)77 USA Increasing colorectal cancer screening among individuals in Level II Colorectal Women Men 50–75 Employed and retired carpenters receiving health benefits through the New Jersey Carpenters Fund from 1996 to 1998. Lipkus IM et al. Level II (2000)86 Breast USA Women ≥ 50 Can tailored Members of the Kaiser Foundation Health interventions increase Plan of North Carolina from 5 sites who mammography use had ≤ 2 mammograms in a 36‐month among HMO women? period. Exclusion criteria included women who: i) did not speak English; ii) had a history of breast cancer that had resulted in a double mastectomy; or ii) currently had breast cancer. Author (Publication year) Country Title postage for returning package. Note: instructions for obtaining screening test without an office visit – i.e. patients could access testing directly. Int2: Control :Waiting list patients (similar in age, sex, race) All received invitation and consent form + pre‐intervention assessment Int1: Usual care Usual care (mailed reminders every year) Int2: Education (mailed) Usual care + 2 mailings of tailored print communications in the form of a two‐
colour booklet with graphic images 1 year apart Int3: Counselling (telephone) Usual care + 2 tailored telephone counselling sessions 2 years apart Int1: Recruitment/consent + baseline interview + mailed non‐tailored basic risk factor information + 3mth, 12mth, 24mth follow‐
up interview Int2: Nature of the intervention (i) Screening rates (ITT) Year 1: Int1 = 54%; Int2 = 57%; Int3 = 61%; Int4 = 64% (no stat sig diffs) Year 2: Int1 = 41%; Int2 = 41%; Int3 = 43%; Int4 = 44% (no stat sig diffs) Year 3: Int1 = 30%, Int2 = 44%, Int3 = 35%, Int4 = 36% (Int 2 stat sig p < 0.05) On schedule for mammography screening 1995 (n = 1099): Int1 = 72%; Int2 = 71%; Int3 = 73% (no stat sig diffs) 1996 (n = 1059): Int1 = 61%; Int2 = 67%; Int3 = 71% (p = 0.020) Int3 vs Int2 p < 0.006 1997 (n = 1070): Int1 = 70%; Int2 = 69%; Int3 = 72% (no stat sig diffs) Findings VOL 3– 130 Knowledge translation: a review of strategies to increase participation in cancer screening
Author (Publication year) Country Title the carpentry trade: Test of risk communication interventions Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Inclusion criteria included: i) not currently in treatment or awaiting any cancer treatment; ii) no history of CRC; iii) never had a sigmoidoscopy or colonoscopy for routine screening; iv) no FOBT within 15 months prior to the baseline interview. Findings Recruitment/consent + baseline interview + mailed non‐tailored comprehensive risk ii) Repeat screening rates (ITT) factor information + 3mth, 12mth, 24mth Years 1–2: Int1 = 38%, Int2 = 35%, Int3 = 40%, follow‐up interview Int4 = 42% (no stat sig diffs) Years 2–3: Int1 = 28%, Int2 = 32%, Int3 = 28%, Int3: Int 4 = 31% (no stat sig diffs) Recruitment/consent+ baseline interview + Years 1–3: Int1 = 28%, Int2 = 29%, Int3 = 27%, mailed tailored basic information + Int 4 = 30% (no stat sig diffs) counselling (phone call) ~2 weeks after mailing + 3‐mth, 12‐mth, 24‐th follow‐up i) Screening rates (observations) interview Year 1: Int1 = 60%, Int2 = 60%, Int3 = 68%, Int4 = 74% (Int4 stat sig p < 0.05) Int4: Year 2: Int1 = 52%, Int2 = 54%, Int3 = 57%, Int4 = 59% Recruitment/consent+ baseline interview + (no stat sig diffs) mailed tailored comprehensive Year 3: Int1 = 41%, Int2 = 59%, Int3 = 49%, Int4 = 51% information + counselling (phone call) ~2 (Int2 stat sig p<0.05) weeks after mailing + 3‐mth, 12‐mth, 24‐ mth follow‐up interview ii) Repeat screening rates (observations) Years 1–2: Int1 = 48%, Int2 = 46%, Int3 = 52%, Int4=56% (no stat sig diffs) Years 2–3: Int1 = 39%, Int2 = 44%, Int3 = 39%, Int4 = 44% (no stat sig diffs) Years 1–3: Int1 = 40%, Int2 = 42%, Int3 = 42%, Int4 = 47% (no stat sig diffs) However in both forms of analysis, for: Participants aged approx ≤ 59, initial, yearly, and repeat screening rates were higher when they received basic (Int1 & Int3) rather than comprehensive risk information (Int2 & Int4). Participants aged > 59, initial, yearly, and repeat Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 131
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings screening rates were higher when they received comprehensive (Int2 & Int4) rather than basic risk information (Int1 & Int3) Lipkus IM et al. Level III‐2 All: 15‐minute baseline telephone survey + 6‐month follow‐up, 31% (n = 32) reported having had (2003)76 Colorectal scheduled to come to Medical Centre's lab FOBT; 8% reported having a SIG. Logistic regression USA Women, Men (within 1–2 weeks of call). All participants analysis predicting FOBT as a function of the ≥ 50 were provided with basic information interventions revealed: Chi sq (3) = 7.9, p < 0.05. Manipulating Men and women who had not had FOBT regarding the function of the colon and perceptions of within last 2 years, residing in 3 counties in rectum, CRC, symptoms of CRC and Logistic regression colorectal cancer North Carolina. Participants were different screening tests. Then Int2 vs Int1: OR = 1.9 (95% CI 1.5, 27.3); p < 0.05 threat: Implications recruited via local newspaper. randomised to receive additional Int3 vs Int1: OR = 1.0 (95% CI 0.7, 12.0); p > 0.05 for screening information. Int4 vs Int1: OR = 1.3 (95% CI 0.8, 16.4); p > 0.05 intentions and behaviors Int1: Control – no additional information provided Int2: CRC risk information + interview Int3: CRC severity information + interview Int4: CRC risk + severity information + interview Lopez VA et al. Level III‐2 Int1: Coaching Pap screening: (2006)129 Breast Cervical Culturally tailored cancer prevention Int1 vs Int2: OR = 0.69 (95% CI 0.41–1.19) USA Women intervention that focused on health ≥ 18 education and cancer screening using lay Cancer intervention attendance Participation and Members of 14 participating churches health workers who are also members of Int1 vs Int2: OR = 1.12 (95% CI 0.91–1.37) program outcomes in considered to be low income, less the participating church communities. a church‐based acculturated Hispanic women Mammography Screening: cancer prevention Int2: Control Int1 vs Int2: OR = 0.82 (95% CI 0.44–1.56) program for Hispanic Participants in the family mental health women program. Cancer intervention attendance: Int1 vs Int2: OR = 1.31, (95% CI 0.99–1.74). Loss J et al. Level II Int1: Community intervention FOBT screenings within 12 months following Int1: Author (Publication year) Country Title VOL 3– 132 Knowledge translation: a review of strategies to increase participation in cancer screening
Author (Publication year) Country Title (2006)103 GERMANY The effects of promoting colorectal cancer screening on screening utilisation: Evaluation of the German campaign "Aktiv gegen darmkrebs" (action against colorectal cancer) Luckmann R et al. (2003)111 USA A randomized trial of telephone counseling to promote screening mammography in two HMOs Mann BD et al. (2000)105 USA Screening to the converted: an educational intervention in African American Int1: Counselling (telephone): Mailed reminders + tailored telephone counselling Int 2: Usual Care Mailed reminders Int1: Education Five‐week, low cost, easily replicable model for cancer education designed to effect a change in healthcare attitudes by encouraging participants to obtain cancer screening through their healthcare coverage entitlements. No control Level II Breast, Colorectal, Cervical Women, Men > 18 Parishoners of three African American churches; one from each of three local communities considered 'partner' neighborhoods of the MCP Hahnemann School of Medicine in Philadelphia. Public awareness campaign consisting of public information sessions and leaflet distribution. Data on screening was collected by doctors for 12 months following campaign. Included where screening was prompted by campaign, in order to determine the proportion against the total number of screenings. Nature of the intervention Level IV Breast Women 50–80 Continuously enrolled in two HMOs in central Massachusetts Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Colorectal Women Men ≥ 50 2 Bavarian districts Pre‐test: Screened within last 2 years (survey response rate ~ 20%) breast 84%, cervical 78%, colon 62%, non‐
compliant 120 (28%), compliant 72% Follow‐up via telephone interview (time to follow‐up was not given): 58/120 (48%) self‐reported that they had been screened; i) Screening rates at 12 months Int1: HMO1 54.6%, HMO2 63.2% Int2: HMO1 49.7%, HMO2 60.1% ii) Int1 effects compared to Int2 HMO1: adjusted OR = 1.26 (95% CI 1.00–1.58), p = 0.048 HMO2: adjusted OR = 1.15 (95% CI 1.01–1.30), p = 0.03 225/3531 (8.7%) 102/225 (45.3%) motivated by info sessions 120/225 (53.4%) motivated by leaflets 3/225 (1.3%) motivated by both However, the impact decreased during the data collection period over 12 months (screenings decreased) Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 133
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings 19/58 (33%) because of program; 25/58 (43%) said “would have sought screening anyway”; 14/58 (24%) had no information reported. Marcus AC et al. Level II Int1: Compliant with screening at 14‐month follow‐up (2005)70 Colorectal Brief educational message + single All Female Male USA Women, Men untailored mail‐out of print material Baseline 20% 20% 22% ≥ 50 (comparison) Int1 42% 40% 50% The efficacy of English speaking callers to 9 Cancer Int2 44% 45% 51% tailored print Information Services (California, Coastal, Int2: Int3 51% 52% 39% materials in Heartland, Mid‐Atlantic, Mid‐South, New Brief educational message + single mail‐out Int4 48% 50% 52% promoting colorectal York, North Central, South Central, of tailored print material Comparative assessment for: cancer screening: Southeast) who would be eligible for CRC ALL: Int1 vs Int3: p = 0.03 Results from a screening at 14 months follow‐up and did Int3: Female: Int1 vs Int3: p = 0.03 randomized trial not call about CRC or CRC screening. Brief educational message + four mail‐outs, involving callers to Eligible = 5987; baseline interview spanning 12 months, of tailored print Age 50–59 60+ the National Cancer completion = 4014; 6‐month follow‐up = material based on baseline information Baseline 17% 24% Institute's Cancer 2740; 14‐month follow‐up = 2224 Int1 34% 50% Information Service Int4: Int2 46% 45% Brief educational message + four mail‐outs, Int3 53% 47% spanning 12 months, of tailored print Int4 50% 57% material based in the first mail‐out on Comparative assessment for 50–59 years age group baseline information and then re‐tailored Int1 vs Int2 (p = 0.048); Int1 vs Int3 (p = 0.003) in line with information received at a 6‐
Int1 vs Int4 (p = 0.008) month interview Ever screened: baseline = 33%; 14 Months Int1 = 47%; Int2 = 49%; Int3 = 56%; Int4 = 59%; stat sig diffs Int1 vs Int4 (p = 0.04) Never screened: Baseline = 0%; 14 Months Int1 = 35%; Int2 = 41%; Author (Publication year) Country Title churches VOL 3– 134 Knowledge translation: a review of strategies to increase participation in cancer screening
Level II Breast Women 65–80 Medicare recipients in North Dakota who had not had a mammogram in two‐and–a‐
half years. McCaul KD et al. (2002)23 USA The effects of mailed reminders and tailored messages on mammography screening Int1: Invitations and reminders (mailed) Mailed simple reminder message Int2: Invitations and reminders (mailed) Mailed reminder accompanied by a persuasive communication emphasising personal risk Int3: Invitations and reminders (mailed) Mailed postcards + follow‐up reminder accompanied by a message tailored to the participants' chief barriers mentioned in returned postcard Int4: Invitations and reminders (mailed) Level II Recruitment + study entry mammogram + Breast baseline telephone survey Women Int1: Reminder (mailed) 50–74 Reminder letter from mammography Patients referred for a mammogram to one facility of the six participating facilities by one of the participating doctors. Inclusion criteria: Int2: Reminder (mailed) (i) no history of breast cancer; (ii) negative Reminder letter from the physician who test result for entry mammogram, (iii) had referred for the previous mammogram
consented to participate, (iv) spoke either English or Spanish. Int3: No reminder Mayer JA et al. (2000)21 USA Patient reminder letters to promote annual mammograms: a randomized controlled trial Nature of the intervention Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title Int3 = 41%; Int4 = 44%; no comparisons reported. Intention‐to‐treat analysis Screening adherence (return rates): Int1: 46.6%; Int2: 47.7%; Int3: 28.3%; chi sq = 51.3, df = 2, p < 0.001 Bonferoni pair‐wise comparison indicated: Int3 vs Int1: stat sig diff Int3 vs Int2: stat sig diff Int1 vs Int2: not stat sig Logistic regression Probability of returning for a mammogram: Int1 vs Int3: OR = 2.37 (95% CI 1.76–3.20) Int2 vs Int3: OR = 2.24 (95% CI 1.66–3.01) The authors conclude that there is no significant impact on mammography screening by type of intervention. Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 135
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level III‐2 Cervical Women ≥ 18 Enrolled members of the Cherokee and Lumbee tribes of North Carolina Level II Breast Women 59–80 Under users – those who had never had a Author (Publication year) Country Title Messer L et al. (1999)124 USA Early detection of cervical cancer among Native American women: A qualitative supplement to a quantitative study Messina C et al. (2002)112 USA Effectiveness of Int1: BSTC + CME Barrier‐specific telephone counselling (BSTC) for participants and no continuing medical education (CME) for doctors Mailed postcards + follow‐up reminder accompanied by National Cancer Institute brochure because no returned postcard Int5: Control: No letter Int1a: Pre‐test + community based, individualised health education program involving two home visits to each participant by lay health advisors + post‐test Int1b: Community‐based, individualised health education program involving two home visits to each participant by lay health advisors + post‐test Int2a: pre‐test + post‐test Int2b: post‐test Nature of the intervention Proportion of women having a Pap test within past twelve months Lumbee With pre‐test: Int2a = 65.1%, Int1a = 71.0% (stat sig) No pre‐test Int2b = 62.5%, Int1b = 76% (stat sig) Cherokee With pre‐test Int2a = 69.1%, Int1a = 74.6% (stat sig); No pre‐test Int2b = 66.8%, Int1b 74% (not stat sig) Percent of regular mammography users at follow‐up stratified by under‐user status (i.e. never user or ever user) Ever user: Int1: 46/115 (40%) Findings VOL 3– 136 Knowledge translation: a review of strategies to increase participation in cancer screening
Michielutte R et al. (2005)85 USA Intervention to increase screening mammography among women 65 and older Findings Int2: 57/63 (57%) Int3: 29/104 (39%) Int4: 29/84 (34%) Int2 vs Int4: p = 0.003 Never user: Int1: 3/34 (9%) Int2: 0/29 (0%) Int3: 7/50 (14%) Int4: 1/31 (3%) Results indicate that the BSTC is effective for ever‐
users (confirmed by logistic regression controlling age and insurance coverage) and CME is effective for never‐users although the small numbers in this group mean this result should be treated with caution. Level IV Int1: Education (mailed)+ counselling Inaccuracies in the practice records meant Breast (telephone) participants were divided into: Women Stage 1 – Education: CME for doctors a primary group i.e. those had not had a ≥ 65 (mailed) mammogram in the past 15 months; and Random samples of patients from 21 Stage 2 – Education: mailed information on a maintenance group i.e. those who had intervention group practices and 22 breast cancer and screening for breast mammogram in the 12 months preceding the control group practices from 15 central cancer for eligible patients approximately 4 baseline data collection and western North Carolina locations. months later, Patient eligibility criteria included: no Stage 3 – Counselling (telephone) Mammography screening: history of breast cancer, have not had a approximately 4 months later Int1 vs Int2: mammogram in the past 15 months, have No significant differences at Stage 1 & 3 for both no serious physical or cognitive problems groups that ruled out screening for breast cancer. Int2: Control Group BUT Stage 1 – Education: CME for doctors on Int1 vs Int2: Significant at Stage 2 Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) women's telephone mammogram and those who had some Int2: BSTC only counseling and previous mammography experience but physician education did not screen regularly– in four townships Int3: CME only to improve in Nassau and Suffolk Counties on Long mammography Island, New York. Exclusion criteria: i) Int4: no intervention screening among women who currently had breast cancer or women who underuse a history of breast cancer; ii) did not speak mammography English or Spanish; iii) had 2 mammograms in the last 4 years with a recommended interval between them. Doctors: Doctor’s women named as their 'doctor' in the recruitment phase. Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 137
Level II Colorectal Women, Men ≥ 50 Outpatients of a university‐affiliated, community‐based internal medicine outpatient practice Miller Jr. DP et al. (2005)104 USA Using a computer to teach patients about fecal occult blood screening: A randomized trial Level III‐3 Cervical Level II Colorectal Women, Men 55–74 Residents in selected postcodes in Melbourne, Adelaide and Mackay Millard FB (2006)37 AUSTRALIA Bowel cancer screening in Australia Morrell S et al. (2005)28 Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title Findings Both groups: Int1 (14.4%) vs Int2 (10.2%); diffs = 4.3% p = 0.020 Primary group: Int1 (13.0%) vs Int2 (8.6%); diffs = 4.4% p = 0.026 Maintenance: Int1 (17.7%) vs Int2 (13.3%); diffs = 4.4% p = 0.148 National Bowel Cancer Screening Pilot Response rate = 45% Program (NBCSPP). The multicultural diversity of the Australian Int1: Invitations (mailed) with the FOBT to population presents challenges to any health approximately 56,907 people program, particularly in groups with a history of poor uptake of health services, the culturally and linguistically diverse, and Aboriginal and Torres Strait Islander communities. In these groups, the response to postal invitations was reduced, despite targeted liaison and educational program, with wrong address, literacy and language problems identified as barriers. Int1: Education (computer) Test kits returned within 30 days: Multi‐media educational computer Int1: 5/93 (62%) program + FOBT kit Int2: 58/93 (63%) Int1 vs Int2: diff = –1% (95% CI –15%, +13%) p = 0.89 Int2: Counselling (face‐to‐face) Nurse counselling (usual care) + FOBT kit Those most likely to return kits were females (71% vs 51% males, p = 0.006, and patients with prior CRC screening (79% vs 54% with no screening history; p < 0.001). Int1: Reminder (mailed) Pap test rates within 90 days of mail‐out: Int1: (n = 59,780) 4.44% skin cancer screening (mailed), Stage 2 –‐ Education: mailed information on skin cancer Stage 3 – Telephone interview without counselling. Nature of the intervention VOL 3– 138 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Women 20–69 Women who had not had a test in 48 months who were on the NSW Pap test registry. Int2: No Reminder Nature of the intervention Findings Int2: (n = 29,919) 2.9 % Int1 vs Int2: 1.53 (95% CI: 1.42–1.65), significant for all age groups except those aged 20–24 years, rate ratio increased with age Multivariate regression analysis Int1 vs Int2: 1.54 (95% CI: 1.43–1.67) By location: Hazard ratios were significantly higher for women outside of Sydney (HR = 1.0), 'other metropolitan (HR = 1.2), large rural centre (HR = 1.2) and other rural area (HR = 1.3). By time: Letters sent out just after Xmas/New Year break were significantly more likely to be associated with test than mail‐out just prior to break (HR = 1.3, p < 0.0001). Mullins R et al Level II Media campaign: On three free to air Negative binomial regression to assess the effects of (2007)144 Cervical commercial television stations in the campaign on the number of Pap tests conducted AUSTRALIA Women programmes rated as appealing to women per week from the first week in June 2002 to the last > 40 in the target range which was broadcast week of 2004. Campaign had a significant positive Encouraging the right Victorian women who were overdue for a for nearly four weeks (July–August 2004). effect on screening behaviour, with an 18% increase women to attend for Pap test (i.e. under‐screened) during the campaign (coeff –0.169, p = 0.029) cervical cancer compared with other periods. screening: results from a targeted television campaign in Victoria, Australia Myers RE et al. Level III‐2 Recruitment, baseline telephone survey , Intention‐to‐treat multivariate analysis Author (Publication year) Country Title AUSTRALIA How much does a reminder letter increase cervical screening among under‐screened women in NSW? Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 139
Level of evidence (NHMRC) Type of cancer Nature of the intervention Findings Participants (sex), Participants (age) Participants (cohort) Colorectal midpoint survey 12 months after Women, Men randomisation, endpoint survey 24 months Int2 vs Int1: OR = 1.84 (95% CI 1.3–2.50); p < 0.0001 50–74 after randomisation Int3 vs Int1: OR = 1.69 (95% CI 1.24–2.29); p = 0.001 Patients of a large urban practice located Int4 vs Int1: OR = 2.08 (95% CI 1.54–2.83); p < 0.0001 at Thomas Jefferson University in Int1: Usual care Philadelphia. Inclusion criteria: i) no prior No statistically significant differences in pair‐wise diagnosis of colorectal neoplasia or Int2: Invitations and reminders (mailed) comparisons between: Int2 & 3; Int2 & 4; Int3 & 4 inflammatory bowel disease; ii) at least CRC invitation letter, information booklet, one visit to the facility within the previous Insure Faecal Immunochemical Test (SBT) + 2 years; iii) provided the facility with reminder letter; two times – 12 months complete contact information including apart address and telephone number; iv) had not undergone recent CRC screening. Int3: Invitations and reminders (mailed) + message framing (mailed) Int2 + 2 tailored message pages addressing personal barriers to SBT and flexible Sigmoidoscopy (mailed) Int4: Invitations and reminders (mailed) + message framing (mailed) + reminder (telephone) Int3 + reminder (telephone) Navarro AM et al. Level II Primary participants: educational Self‐reported mammography rates (2007)93 Breast, Cervical interactive group sessions run by trained Primary participants: USA Women community health advisors Pre (63.3%) vs Post (70.0%), not stat sig ≥ 40 Learning partners: Diffusion of cancer Latino women from local community in San Learning partners recruited by primary Pre (66.7%) vs Post (68.3%), not stat sig education Diego participants: friends and/or family information through recruited by primary participants with a Latino community whom they shared the cancer info they Self‐reported Pap test rates: health advisor received. Primary participants: Author (Publication year) Country Title (2007) 34 USA A randomized controlled trial of the impact of targeted and tailored interventions on colorectal cancer screening VOL 3– 140 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Findings Women who had not had a mammogram (never‐
screened) did so during the 16–18 week follow‐up Ever had a Pap test Int1: pre (77.5%) vs post (84.2%); diff = +6.7% (p < 0.001); Int2: pre (73.9%) vs post (70.6%); diff = –3.3% (p > 0.05), Change Int1 vs Int2: (p < 0.001) Had a Pap test in the last year Int1: pre (64.9%) post (70.4%); diff = +5.5% (p < 0.5) Int2: pre (59.1%) post (53.1%); diff = –6.0% (p > 0.05) change Int1 vs Int2 (p < 0.001) Pre (92.3%) vs Post (97.9%), not stat sig Learning partners: Pre (88.3%) vs Post (92.8%), not stat sig Int1: Community intervention Expected number of Pap tests and mammograms Personal health record booklets (mailed) to were determined from data for the 19 quarters (3 all adults in the towns + family months) before the intervention. practitioners consented to support and promote the intervention + media Observed/expected ratios 3‐month post intervention campaign were not different between Int1 and Int2 for mammogram and Pap tests Int2: Usual care Nature of the intervention Level III‐2 Breast Cervical Women 20–60 Residents of 10 matched pairs of rural towns in New South Wales. Inclusion criteria for towns: i) non‐metropolitan postal codes; ii) total population of 1000–
5000; iii) one family practitioner; iv) closest town > 30 km distance away; v) not currently or imminently involved in similar preventive healthcare trial projects. Nguyen TT et al. Level III‐2 Community‐based participatory research (2006)135 Cervical involving the development and USA Women implementation of a community action Not stated plan funded for four years. Community‐based Vietnamese living in Santa Clara County, participatory research California (intervention community) and Int1: Community intervention increases cervical Vietnamese living in Harris County, Texas i) media campaign; ii) involved lay health cancer screening (control community) worker outreach; iii) continuing medical among Vietnamese‐
education; iv) restoration of the breast and Americans cerivcal cancer control program; v) Vietnamese pap clinic and patient navigator; vi) Pap registry and reminder system. Int2: Usual care Oliver‐Vãzquez M et Level II All participants received a first interview al. Breast and pre‐test, the health education Newell SA et al. (2002)182 AUSTRALIA Can personal health record booklets improve cancer screening behaviors? Author (Publication year) Country Title program Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 141
Level of evidence (NHMRC) Type of cancer Nature of the intervention Participants (sex), Participants (age) Participants (cohort) Women sessions, a second interview and post‐test ≥ 65 to compare knowledge and belief. Then: Clients of senior centres. Inclusion criteria Int1: included: i) not having performed at least Received external support including one of the breast cancer early detection medical appointments for CBE and practices; ii) completion of the pre‐test and mammograms, and reminders about post‐test; iii) attendance at two or more appointments educational sessions; and iv) possession of Int2: the mental and auditory capacity to Received no external support participate in the program. Both groups participated in a follow‐up interview at 16–18 weeks to assess impact on breast cancer screening Ore L et al. Level II All took part in a telephone interview (2001)78 Colorectal approximately 2 months after the initial ISRAEL Women, Men mailing 50–74 Screening with faecal Residents of Haifa and members of the Int1: Mailed FOBT kit occult blood test Kupat Holim Clalit health maintenance (FOBT) for colorectal organisation (HMO) Int2: Message framing (risk) cancer: assessment of Int1 + leaflet describing risk and the two methods that importance of early detection attempt to improve compliance Int3: Procedure Mailed request card to be sent to the Carmel Medical Centre asking for a kit to be mailed Int4: Message framing Int3 + leaflet describing risk and the Author (Publication year) Country Title (2002)122 PUERTO RICO Breast cancer health promotion model for older Puerto Rican women: Results of a pilot programme Compliance assessed 5 months after the initial mailing using the HMO FOBT screening program database Women: Int1+Int2 (kit) = 21.3% Int3+Int4 (card) = 14.7%, p = 0.007 Men: Int1+Int2 (kit) = 18.5% Int3+Int4 (card) = 17.1%, p = 0.55 Total: Int1+Int2 (kit) = 19.9% Int3+Int4 (card) = 15.9%, p = 0.02 The authors indicate that the inclusion of the leaflet (Int2 and Int4) had no 'observable impact on period: Int1: 3/3 Int2: 0/12 Findings VOL 3– 142 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention importance of early detection Page A et al. Level II Int1: Control group (2006)41 Breast No invitation (45–49 years who are not AUSTRALIA Women invited to BreastScreen New South Wales 50–54 program) Recruitment to All women on electoral roll aged 50–54 mammography during March 2004 who have never Int2: Invitation (mailed) screening: A attended BSNSW for mammogram in two Standard practice – mailed personalised randomised trial and screening and assessment service invitation to make appointment for free meta‐analysis of catchment areas screening invitation letters and telephone calls Int3: Invitations and reminders (mailed) Mailed invitation + send reminder by mail after 6 weeks Int4: Invitation (mailed) and reminders (telephone) Invitation letter + follow‐up (telephone) all to non‐responders after 6 weeks. Outcome was women in each group who attended for mammogram after 12 weeks.
Page A et al. Level III‐2 Media campaign: Italian radio scripts and (2005)196 Breast newspaper advertisements. AUSTRALIA Women 50–69 Phase 1 – May to August 2003 Mammography Italian speaking women in NSW Phase 2 – September to December 2003 screening participation: Effects Each phase compared the media Author (Publication year) Country Title Mammogram rates for Italian women Phase 2 vs (January 1999–August 2003): no stat sig Phase 2 vs (January 2004–June 2004): no stat sig 23% cited Italian radio advertisement, and 17% cited Italian newspaper advertisements as prompts to have screening. Mammogram at 12‐week follow‐up (ITT) Int1: 11/788 (1.4%) Int2: 43/786 (5.5%) Int3: 67/785 (8.5%) Int4: 61/785 (7.8%) Int1 vs Int2: OR = 0.24 (95% CI 0.13–0.48) Int3 vs Int2: OR = 1.61 (95% CI 1.08–2.40) Int4 vs Int2: OR = 1.46 (95% CI 0.97–2.18) compliance.’ Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 143
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings intervention (four weeks) and a follow‐up monitoring period (12 weeks). The most common prompt cited to attend screening was the BreastScreen NSW letter (66%), followed by information or recommendation from GP (44%). Paskett ED et al. Level II Int1: Community intervention Within guidelines mammogram screening rates: (1999)138 Breast, Cervical Surveys of providers and women in target Int1: baseline = 31%, follow‐up = 56%; diff = +25% USA Women group + in‐reach (health centre based) and (p = 0.049) ≥ 40 outreach (community‐based) activities + Int2: baseline = 33%, follow‐up = 40%; diff = +7% (not Community‐based Residents in low income housing follow‐up survey of women in target group stat sig) interventions to communities in one intervention and one (~2.5 years after baseline) Int1 vs Int2: diff = +18%; p = 0.04 unadjusted Wald improve breast and comparison city in North Carolina Chi sq test cervical cancer Int2: Control screening: Results of Surveys of providers and women in target Pap test rates within the last 3 years the Forsyth county group + follow‐up survey of women in Int1: baseline = 73%, follow‐up = 87%, diff = +14% cancer target group (~2.5 years after baseline) (no stat) screening(FOCAS) Int2: baseline = 67%, follow‐up = 60%, diff = –7% project (no stat) Int1 vs Int2: diff = +21% unadjusted Wald Chi sq test p = 0.004, Pignone M et al. Level II Int1: Education (video, brochure) Colon cancer screening test completed 3 months (2000)180 Colorectal Viewed educational video on colon cancer after initial scheduled visit: USA Women, Men and then chose one of three colour‐coded, Int1 (n = 125) 50–75 patient directed brochure to indicate their FOBT = 28.5%; sigmoidoscopy = 17.6%; Videotape‐based Patients scheduled to be seen for a new or interest in screening and a matching card either = 36.8% decision aid for colon ongoing health problem by one of the was attached to the patient's chart + cancer screening participating providers in three community completed 2nd questionnaire; saw Int2 (n = 124) primary care practices in central North provider then completed 3rd questionnaire FOBT = 20.2%; sigmoidoscopy = 4.8%; either = 22.6% Carolina. Exclusion criteria were patients who: i) reported a personal or family Int2: Control Int1 vs Int2: history of colon cancer; ii) reported FOBT Viewed video on car safety and received a FOBT: diff = +8.3% (95% CI –2.4–+18.9) Author (Publication year) Country Title of a media campaign targeting Italian‐
speaking women VOL 3– 144 Knowledge translation: a review of strategies to increase participation in cancer screening
Potter MB et al. (2009)178 USA Offering annual fecal occult blood tests at annual flu shot clinics increases colorectal cancer screening rates Potter MB et al. (2009)43 USA Improving colorectal cancer screening: A partnership between primary care practices and the American cancer society Author (Publication year) Country Title Level of evidence (NHMRC) Type of cancer Nature of the intervention Participants (sex), Participants (age) Participants (cohort) in the past year or flexible sigmoidoscopy, standard brochure on automobile safety + colonoscopy or barium enema in the past 5 completed 2nd questionnaire; saw years; iii) were judged to be too ill; or iv) provider then completed 3rd questionnaire
had scheduled appointments for laboratory blood work only. Level III‐1 Int1: Usual care Colorectal Women, Men Int2: > 50 Poster presenting the menu of options for Patients of primary care clinics at the colorectal cancer screening in a University of California, San Francisco multilingual format placed in each examination room Int3: Int2 + reminder (telephone) for those from whom tests were ordered + option of further telephone call 2–4 weeks after the first Level II Int1: Procedure Colorectal Flu shots plus FOBT kits (9 intervention Women, Men sessions) + reminders (telephone: 3 and 6 50–79 weeks if FOBT kit not returned) Patients of the Family Health Center at San Staff knew in advance who was needed for Francisco General Hospital – low SES, FOBT. On arrival patients received ethnically diverse – who attend annual flu colourful multilingual education sheet with shot clinics run by multilingual medical messages about flu prevention and colon assistants and health workers. cancer prevention. Int2: Control Flu shots only (8 control sessions) FOBT rates at the end of flu season:
Int1: baseline = 54.5%, follow‐up = 84.3%; diff = +29.8% (p < 0.001) Int2: baseline = 52.9%, follow‐up = 57.3%; diff = + 4.4% (p = 0.07) Being up‐to‐date for screening at the conclusion of study among patients initially due for screening Int1: 68.0% Int2: 20.7% Multivariate analysis: Int1 vs Int2: OR of becoming up‐to‐date = 11.3 (95% CI 5.8–22.0); Multivariate analysis Up‐to‐date screening at baseline: Int2 vs Int1: OR = 1.05 (95% CI 0.79–1.39) Int3 vs Int1: OR = 1.36 (95% CI 1.02–1.80) Due for screening at baseline: Int2 vs Int1: OR = 1.04 (95% CI 0.81–1.34) Int3 vs Int1: OR = 1.49 (95% CI 1.16–1.90) p < 0.001 sigmoidoscopy: diff = +12.8% (95% CI 9.1–20.4) either: diff = +14.2% (94% CI 3.0–25.4) Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 145
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Powe BD et al. (1999)98 USA An intervention to decrease cancer Level II Colorectal Women Men ≥ 50 Attendees at senior citizen centres in a rural southern state who were mentally Powe BD Level II (2002)94 Colorectal USA Women ≥ 50 Promoting fecal African American, who attended senior occult blood testing in citizen centres in a rural southern state, rural African and who were mentally oriented to time American women and location. Author (Publication year) Country Title p < 0.001 OR of remaining up‐to‐date = 5.8 (95% CI 1.5–22.0); p < 0.05 Screening participation rates at baseline and 12‐ month follow‐up: Int1: 65%; 63% Int2: 43%; 34% Int3: 10%; 7% Int1 vs Int3: p < 0.0001 Int2 vs Int3: p < 0.0001 Int1 vs Int2: p = 0.0052 Significant predictors of participations were: family history of cancer (p = 0.006); average number of visits to providers (p = 0.036); and intervention group (p < 0.0001). Age and education were not significant predictors. Findings Int1: Data collection and distribution of FOBT kits + cultural and self‐empowerment consisting of a five stage intervention (n = 5 centres) including: i) videotape designed to address cancer fatalism; ii) 12‐month calendar containing information on colorectal cancer, iii) wall sized poster , iv) culturally appropriate brochure, v) one‐
page colour handout showing the correct method for using the FOBT kits. Int2: Education Data collection and distribution of FOBT kits and cultural intervention consisting of the first stage of the five stage intervention (n = 5 centres) Int3: Control group Traditional group (n = 5) data collection and distribution of FOBT kits only Int1: Education (video) FOBT screening participation: Video ‘Telling the story. To live is God's Int1: 60% will’ + FOBT kits Int2: 68% (Chi sq = 0.50, p = 0.48) Int2: Control group Education: Video (American Cancer Society Nature of the intervention VOL 3– 146 Knowledge translation: a review of strategies to increase participation in cancer screening
Rakowski W et al. (2003)66 USA Powell ME et al. (2005) 106 USA Increasing mammography screening among African American women in rural areas Powe BD et al. (2004)95 USA An intervention study to increase colorectal cancer knowledge and screening among community elders Author (Publication year) Country Title fatalism among rural elders Level of evidence (NHMRC) Type of cancer Nature of the intervention Findings Participants (sex), Participants (age) Participants (cohort) oriented to date and location, and agreed colorectal cancer video) + researcher to participate. discussed foods and medications to avoid while completing the FOBT and how to complete the FOBT kits Level II Int1: Education 12‐month follow‐up on screening participation: Colorectal Video ‘Telling the story..To live in God's Int1: 33/54 (61%) Women, Men will’ + calendar, poster and brochure Int2: 18/39 (46%) ≥ 50 Int3: 5/41 (12%) Attendees at the centres who were Int2: Education oriented to time and location and reported Video ‘Telling the story. To live in God's Post‐hoc test done by reviewers that they had not participated in a study will’ Int1 vs Int3: chi sq = 23.235, df = 1, p = 0.000 on colorectal cancer within the past year. Int2 vs Int3: chi sq = 11.252, df = 1, p = 0.001 Int3: Control Int1 vs Int2: chi sq = 2.046, df = 1, p = 0.153 Standard treatment watched the ACS video 'Colorectal cancer: The cancer no one talks about' Level II Int1: Education Mammography screening up‐to‐date at baseline and Breast Full program: group educational session + at follow‐up: Women in‐home visit from a Home Health Educator Int1: 45.3%; 62.7%; diff = 17.4% ≥ 40 + financial incentive Int2: 64.8%; 70.4%; diff = 8.6% Attendees at African American churches in Int3: 56.8%; 61.4%; diff = 8.2% Greene County, Alabama Int2: Education Int1 vs Int3: p < 0.001 Partial program: group educational session + financial incentive Additional analyses (not shown) indicated that the effect of the program on mammography attainment Int3: Control did not vary by age, educational level, insurance Delayed program + financial incentive status, family history of breast cancer, or frequency of church attendance. Level III‐2 Reminder letter sent about 2 months Results indicated no substantial differences in repeat Breast before next mammogram due + phone call screening among the four intervention groups. The Women + baseline interview and information about authors conclude that: Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 147
Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) 50–74 randomisation group to all participants, Reminder letter, Inclusion criteria included: i) scheduled then: tailored stepped‐care, mammography appointment designated as and self‐choice screening at a Rhode Island HMO and a Int1: Invitations and reminders (mailed) comparison for North Carolina radiology clinic of a Mailed reminder letter 2 months before repeat university medical system; ii) never had next mammogram mammography breast cancer; iii) were not a facility employee; iv) were in the health plan; v) Int2: Invitations and reminders (mailed) had not been in a prior mammography Mailed follow‐up tailored material 2 intervention; vi) did not have more than months after keeping scheduled one mammogram a year; and vii) spoke mammogram (i.e. 10 months before next English. mammogram due) Int3: Invitations and reminders (mailed) Mailed follow‐up tailored material 10 months after keeping scheduled mammogram (i.e. 2 months before next mammogram due) Int4: Self‐choice of above 3 interventions Let women select any of the 3 contact strategies. Reeves MJ et al. Level IV Int1: Invitations and reminders (mailed) (1999)31 Breast, Cervical Mailed reminder from own doctor within USA Women the past 12 months to schedule either an Pap test ≥ 18; Mammogram ≥ 40 appointment for a Pap test, or a Use of patient Respondents in the Wisconsin Behavioral mammogram. reminder letters to Risk Factor Survey promote cancer Outcomes measured as: screening services in Compliant: had screening test in past 2 Pap screening test within the last 3 years Remembered receiving letter (n = 103): compliant = 94.3%, past due = 4.8%, never‐screened = 0.9% Did not remember receiving letter (n = 702): compliant = 78.1%, past due = 18.2%, never‐screened = 3.7%; Chi sq = 19.8, df = 2, i) for women who had already kept a scheduled mammography appointment, a simple reminder letter was as effective as a more complex intervention; ii) the results are consistent with studies that have shown that women, even those of lower income, respond favourably to simple mailed and telephone reminders. Findings VOL 3– 148 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level II Breast, Colorectal, Cervical Women, Men Older adults Older adults living in continuing care retirement community Author (Publication year) Country Title women: A population‐based study in Wisconsin Resnick B (2003)92 USA Health promotion practices of older adults: Testing an Findings p = 0.000 Multivariate logistic regression model – Women who remembered receiving a reminder letter vs women who did not remember receiving letter: OR = 4.36 (95% CI 1.79, 12.0) Mammogram test within past 2 years Remembered receiving letter (n = 54): compliant = 81.7%, past due = 10.6%, never‐screened = 7.7% Did not remember receiving letter (n = 411): compliant = 59.4%, past due = 16.3%, never‐screen = 24.3%; Chi sq = 11.1, df = 2, p = 0.004 Multivariate logistic regression model– Women who remembered receiving a reminder letter vs women who did not remember receiving letter: OR = 2.86, (95% CI 1.15, 7.07) Education: Testing individualised care Pre‐post analysis: approach to health promotion (geriatric No change in screening for FOBT (p = 0.07), nurse practitioner model) provides formal mammograms (p = 0.69), or cervical smear tests and informal education about health (p = 0.12). In fact, there was a decrease in those promotion activities to group of older interested in cancer screening. adults in continuing care retirement community. Health behaviours were years (mammography) or 3 years (Pap test). Past due: had at least one screening test in past but not currently compliant. Never‐screened: never had screening test. Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 149
Rimer BK et al. (2002)102 USA Rimer BK et al. (2001) 67 USA The short‐term impact of tailored mammography decision‐making interventions Author (Publication year) Country Title individualized approach Reuben DB et al. (2002) 214 USA A randomized clinical trial to assess the benefit of offering on‐
site mobile mammography in addition to health education for older women Nature of the intervention Level II Breast Women Int1: Education Int3: Message framing + counselling Int2 + Counselling (telephone) Two time periods measured: 12 and 24 months for the 3 groups: measured at baseline and follow‐up 12 months later. Level II Recruitment, eligibility assessment, Breast baseline assessments, outcomes Women assessment (by telephone) at 3 months 60–84 Attendees at 60 community‐based senior Int1: Education + reminders (mailed) citizen sites. Exclusion criteria: i) Education – Senior Women's Health Day in mammogram within past year; ii) inability English and Spanish + reminder (mailed) 2 to speak English or Spanish; iii) limited weeks later cognitive capacity. Int2: Education + procedure Offer of on‐site mobile after the presentation mammography Level II Int1: Usual care Breast Reminder letter on woman's birthday, Women letters to providers identifying their 40–44 and 50–54 patients in need of mammograms, Random sample stratified by age and physician training manual about mammography adherence status, enrolled mammography to all providers with the Personal Care Plan of Blue Cross Int2: Message framing and Blue Shield, North Carolina. Usual care + tailored print booklets Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Mammogram rates at 12 and 24 months: Logistic regression Int1 vs Int3: OR = 0.7 (95% CI 0.5–1.0) p = 0.059 At 12‐month follow‐up mammogram in past 15 months: Bivariate analysis: Int1: 63% (n = 412) Int2: 61% (n = 392) Int3: 69% (n = 323) not statistically significant p = 0.066 Multivariate analysis: Int2 vs Int1: OR = 0.97 (95% CI 0.71–1.32, p = 0.852) Int3 vs Int1: OR = 1.41 (95% CI 1.01–1.97, p = 0.045) Mammography rates within 3 months: Bivariate analysis: Int1 = 91/228 (40%) Int2 = 129/235 (55%), p = 0.001 Multivariate controlling cluster: Int2 vs Int1: OR = 1.83 (95% CI 1.22–2.74) Multivariate controlling for cluster and education: Int2 vs Int1: OR = 1.81 (95% CI 1.21–2.71) Findings VOL 3– 150 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) 40–44 and 50–54 2 samples of women recruited from those enrolled in the Blue Cross and Blue Shield, North Carolina Risi L et al. (2004)87 SOUTH AFRICA Media interventions to increase cervical Level II Cervical Women 35–65 Residents in peri‐urban community of 350,000–500,000 people near Cape Town –
Rimer BK et al. Level II (1999)71 Breast, Cervical, Colorectal USA Women, Men ≥ 18 The impact of tailored Stratified random sample of Lincoln interventions on a Community Health Centre clients who had community health visited the centre within the previous 18 center population months. Author (Publication year) Country Title Effects of a mammography decision‐making intervention at 12 and 24 months Findings Int1: Education (photo‐comic) Received photo‐comic ‐ 'Nokwhezi's Story' + media program (radio) Int2: Control Received 'placebo' photo‐comic Int3: Message framing + counselling (telephone) Int2 + tailored counselling (telephone) 2 weeks later
Int2: Message framing Tailored print communications mailed which were sent to patients around the time of their birthdays All participated in baseline telephone interviews and follow‐up interviews at 16 months Int1: Reminders/prompts Provider prompting intervention which generated printed physician prompts that were attached to the patient's chart Int3: Usual care Int2: Int1 + counselling (telephone) Self‐reported Pap screening uptake 6 months after distribution of comics: Int1: 18/269 (7%) Int2: 25/389 (6%); p = 0.89 Did recall hearing the radio drama: 9/ 58 (17%) Pap test within last year at follow‐up: Int1: 56% Int2: 52% Int3: 64% (p = 0.05) Mammogram within past 2 years: Int1: 86% Int2: 82% Int3: 85% (not stat sig) Tailored print materials (mailed) – tailored Int2 vs Int3: OR = 1.4 (95% CI 0.9–2.1) p = 0.283 booklet sent in 1st year (based on info from baseline interview). After 12‐month follow‐up interview, newsletter designed to reinforce, and update content in booklet mailed Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 151
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) from stratified sample of census areas. ii) Multivariate analyses of intervention effects at 24 i) Pap tests at baseline and 24‐month follow‐up: Baseline 24 months Pap tests control 57.6% 45.3% Pap tests int. 61.9% 47.3% Mammograms control 75.9% 64.5% Mammograms int. 71.4% 67.0% FOBT control 22.1% 12.6% FOBT int. 35.9% 28.2% ii) Multivariate analyses of intervention effects at 12 months Pap tests: OR = 1.57 (95% CI 0.92–2.64), p = 0.096 Mammograms: OR = 1.62 (95% CI 1.07–9.78), p = 0.023 FOBT: OR = 2.56 (95% CI 1.65–4.01), p = 0.000 12 months 48.2% 62.4% 71.1% 75.7% 11.9% 40.1% Did not recall hearing the radio drama: 19/429 ( 4%); p < 0.001 educational information on finances; no healthcare messages + media program (radio) i) Up‐to‐date with screening Baseline Pap tests control 57.6% Pap tests int. 61.9% Mammograms control 75.9% Mammograms int. 71.4% FOBT control 22.1% FOBT int. 35.9% Findings Nature of the intervention Level II Int1: Prompts Breast, Colorectal, Cervical Int2: Usual care Women, Men 50–75 Established patients of 8 community health centres in Hillsborough County, Florida. Exclusion criteria included: i) women with personal history of breast cancer were excluded from mammogram analysis; ii) women with a personal history of cervical cancer or those who had a hysterectomy were excluded from Pap test analysis; iii) patients with a personal history of colon cancer and those who had received a colonoscopy or double‐contract barium enema in the previous 10 years were excluded from FOBT analysis. Roetzheim RG et al. Level III‐1 Int1: Prompts (2005)186 Breast, Colorectal, Cervical Int2: Usual care USA Women, Men 50–75 Long‐term results Established patients of 8 community health from a randomized centres in Hillsborough County, Florida. controlled trial to Exclusion criteria included: i) women with increase cancer personal history of breast cancer were screening among excluded from mammogram analysis; ii) attendees of women with a personal history of cervical Author (Publication year) Country Title screening uptake in South Africa: An evaluation study of effectiveness Roetzheim RG et al. (2004)185 USA A randomized controlled trial to increase cancer screening among attendees of community health centers VOL 3– 152 Knowledge translation: a review of strategies to increase participation in cancer screening
Sarfaty M et al. (2006) 137 USA Choice of screening modality in a colorectal cancer education and screening program for the uninsured Rutter DR et al. (2006)24 UK An implementation intentions intervention to increase uptake of mammography Author (Publication year) Country Title community health centers Level of evidence (NHMRC) Type of cancer Nature of the intervention Findings Participants (sex), Participants (age) Participants (cohort) cancer or those who had a hysterectomy months were excluded from Pap test analysis; iii) Pap tests: OR = 0.88 (95% CI 0.68–1.15), p = 0.34 patients with a personal history of colon Mammograms: OR = 1.26 (95% CI 1.02–1.55), cancer and those who had received a p = 0.03 colonoscopy or double‐contract barium FOBT: OR = 1.17 (95% CI 0.92–1.48), p = 0.19 enema in the previous 10 years were excluded from FOBT analysis. Level III‐3 Int1: Invitations and reminders Mammography rate at follow‐up Breast Invitation (mailed) – postal questionnaire + Int1: 80.2% Women National Health Service Breast Screening Int2: 78.9% 50–64 Program (NHSBSP) invitation Int3: 80.3% not stat sig All members of two screening cohorts from Kent in the UK. 34.4% were being called Int2: Invitations and reminders Statistically significant effects of screening history: for their first round of screening (first Invitation (mailed) – postal questionnaire 89% of recall women, 69.6% of first timers timers) and 65.6% for routine three‐yearly containing section on implementation follow‐up (recalls). intentions to overcome 3 barriers + NHSBSP invitation Int3: Control NHSBSP invitation Level IV Education + clinical program Screening rate using any modality (i.e. FOBT, Colorectal Education program: 1st Year – media FOBT/endoscopy OR colonoscopy): Women Men campaign (English and Spanish) included baseline = 13% ≥ 50 broadcast, print, transit advertising. follow‐up = 52% (n = 866) Low income, uninsured in diverse county in Postcards mailed to 63,000 residents. Maryland – offered free colonoscopy 2001–03. Trained Spanish speaking health Over 90% who screened were minorities. Despite workers conducted education sessions free colonoscopy, 20% chose FOBT. (brief stand‐up and longer formal sit‐down) at diverse local venues. Also, primary care physicians were sent education packet, with ‘direct referral’ forms, and Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 153
Saywell Jr. RM et al. (2003)74 USA The cost effectiveness of 5 interventions to increase mammography adherence in a managed care population Sauaia A et al. (2007)120 USA Church‐based breast cancer screening education: impact of two approaches on Latinas enrolled in public and private health insurance plans Author (Publication year) Country Title Nature of the intervention Findings presentations to hospital staff. Clinical program: physician referral, clinical screening, medical treatment, case management. Also cancer information phone line, lab handling FOBT kits, reporting, individual follow‐up. Level II Tepeyac Project: a six year health quality Mammogram screening rates for the baseline period Breast improvement projects that began in 1999 (January 1998 – December 1999) and the follow‐up Women using a community‐participatory approach. period (January 2000 – December 2001) Not stated Int1: 2979/5130 (58%); 3338/5708 (58%) Latinas living in each of the intervention Int1: Education Int2: 316/536 (59%); 359/590 (61%) areas in Colorado Tailored breast‐health promotion packages were mailed to Catholic churches GEE model adjusting for age, income, urbanicity, disability and insurance type: Int2: Coaching Int2 vs Int1: p = 0.03 [GEE parameter estimate= 0.24] Trained peer counsellors (Promotoras) delivered the health promotion message personally on a one‐to‐one basis Level III‐2 Study Period: 1994 to 1999. Mammography adherence at 6 months post‐
Breast intervention Women Baseline information obtained and then Logistic regression: 50–85 randomisation to: Int2 vs Int1: OR = 1.72 (95% CI 0.964–3.054) p = 0.067 Women on the records of a midwestern Int1: Control group Int3 vs Int1: OR = 2.66 (95% CI 1.529–4.610) p = 0.001 HMO who had never had breast cancer No reminder letter or counselling Int4 vs Int1: OR = 2.02 (95% CI 1.165–3.514) p = 0.012 and had not had a mammogram in past 15 Int2: Counselling (telephone) Int5 vs Int1: OR = 2.21 (95% CI 1.282–3.824) p = 0.004 months. Int3: Counselling (face‐to‐face) Int6 vs Int1: OR = 4.27 (95% CI 1.919–5.801) p < 0.001 Int4: Message framing Subgroup analysis: Physician reminder letter (mailed) Int6 had significant effect on all subgroups; all other Int5: Counselling (telephone) plus message interventions did not have stat sig effect. Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) VOL 3– 154 Knowledge translation: a review of strategies to increase participation in cancer screening
Level II Colorectal Women, Men Unknown Internal medicine, family practice, general practice, and gastroentrology physicians in North and South Carolina. Intervention Group: 12 Counties in North Carolina and 17 Counties in South Carolina
Comparison Group: 36 Counties in North Carolina and 17 counties in South Carolina Schenck AP et al. (2006)184 USA Increasing colorectal cancer testing: Translating physician interventions into population‐based practice framing Physician reminder letter (mailed) Int6: Counselling (face‐to‐face) plus message framing Physician reminder letter (mailed) Nature of the intervention Int1: Education Telephone‐based continuing medical education to physicians and office‐based tool kits. Int2: Invitations and reminders (mailed) Telephone based continuing medical education to physicians and office‐based tool kits + piloted small groups of physicians: personalised letters sent by physician to patient, and distribution of free FOBT kits in physician offices. Int3: Comparison group No intervention (usual care) Schneider TR et al. Level II Recruitment, completed questionnaires (2001)80 Breast before and after viewing videos + follow‐
USA Women up by mail at 6 months and 12 months for > 40 those not reporting a mammogram at 6 The effects of Patients attending an inner‐city community months message framing and health clinic or residing in public housing Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title Self‐reported mammogram at 6‐month follow‐up: Int2 + Int4 (loss framed): 50% Int1 + Int3 (gain framed): 36% Loss vs Gain: OR = 1.81, p < 0.01 BUT: Int3 (41%) vs Int4 (36%): OR = 1.22, p > 0.10 Contemplators: OR = 2.66 (95% CI 1.109–6.366) Pre‐contemplators: OR = 4.27 (95% CI 1.903–9.579) History of mammogram: OR = 3.11 (95% C 1.726–5.598) No history of mammogram: OR = 7.00 (95% CI 1.044–46.94) No stat sig effects on the use of FOBT Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 155
Sequist TD et al. (2009)33 USA Segura JM et al. (2001)44 SPAIN A randomized controlled trial comparing three invitation strategies in a breast cancer screening program Author (Publication year) Country Title ethnic targeting on mammography use among low‐income women Level of evidence (NHMRC) Type of cancer Nature of the intervention Participants (sex), Participants (age) Participants (cohort) developments in the same neighbourhood. Int1: Message framing Gain framed, multicultural video + flyer Int2: Message framing Loss framed, multicultural video + flyer Int3: Message framing Gain framed, Latino targeted video + flyer Int4: Message framing Loss framed, Latino targeted video + flyer Level II Int1: Invitations and reminders (mailed) Breast Invitation (mailed) from program director Women 50–64 Int2: Invitations and reminders (mailed) People living in the Raval Nord Invitation (mailed) from primary neighborhood of Barcelona. Exclusion healthcare team criteria were: i) women who reported having had a mammogram within the past Int3: Invitations and reminders (face‐to‐
12 months, ii) women with a prior breast face) neoplasm or terminal illness, iii) women Direct contact by non‐health professional who could not be identified because of who handed the letter personally to errors in the census data. women at home and explained the procedures of the screening program, commenting on the letter contents and all advantages of early detection within a program. Level II PATIENT GROUP Colorectal Int1: Invitations and reminders: invitations Women Men (mailed) Screening rates for patients Int1: 44.0% Int2: 38.1% Screening rates: Int1 = 165/317 (52.1%) Int2 = 183/329 (55.6%) Int3 = 216/340 (63.5%) Int3 vs Int1: RR = 1.22 (95% CI 1.07–1.39) Int3 vs Int2: RR = 1.14 (95% CI 1.01–1.29) Subgroup analysis indicates that Int3 resulted in higher rates in women with less education and women who had completed primary education, but not in the highest education group. At 12‐month follow‐up: No stat sig differences Findings VOL 3– 156 Knowledge translation: a review of strategies to increase participation in cancer screening
Nature of the intervention i) cover letter from chief medical officer advising overdue, and dates for screening; ii) educational pamphlet with screening options; iii) FOBT kit with 3 coloscreen stool cards and return envelope; iv) dedicated telephone number to schedule flexible sigmoidoscopy or colonoscopy. Int2: Control: usual care PHYSICIAN GROUP Int3: Reminders/prompts Received electronic reminders during visits with their patients who were overdue for screening plus prior to intervention received education session, re: use of reminders. Int4: Control Prior to intervention received education session, re: use of reminders. Sheeran P et al. Level II Int1: Invitations and reminders (2000)69 Cervical Reminder (mailed) standard postal UK Women reminder from their medical practitioner 20–67 indicating that they should attend for the Using implementation Patients registered at a single medical cervical smear test within the next 3 intentions to increase practice in rural England who were due for months. They were then sent a attendance for a cervical smear test during a three‐month confidential postal questionnaire cervical cancer period. concerning their views of the cervical screening smear tests with an implementation intention specifying when, where and how they would make an appointment for the Author Level of evidence (NHMRC) (Publication year) Type of cancer Country Participants (sex), Participants (age) Title Participants (cohort) 50–80 Patient and physician Patients and physicians at healthcare reminders to promote centres in Massachusetts. Patients colorectal cancer overdue for screening. screening a randomized controlled trial Cervical cancer screening rates: Int1: 92% Int2: 69% Chi sq = 9.20, df = 1, p < 0.002 Int1 vs Int2: diff = +5.8% (95% CI 4.5–7.1); p < 0.001 Effect increased with age: Int1 vs Int2 50–59: 42.1% vs 38.4%; diff = +3.7 (95% CI 2.0–5.5) 60–69: 45.4% vs 38.0%; diff = +7.3 (95% CI 4.5–10.1) 70–80: 47.4% vs 37.3%; diff = +10.1 (95% CI 7.0–13.2) Int3 vs Int4 (41.9% vs 40.2%); p = 0.47 But electronic reminders tended to increase rates among patients with 3 or more visits (59.5% vs 52.7%); p = 0.07 Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 157
Level II Breast Level II Breast Women ≥ 50 Attendees at influenza clinics in Litchfield County (a semirural area in northwestern Connecticut) who had not had a mammogram in the preceding year. (5 control and 4 intervention clinics) Shenson D et al. (2001)45 USA Improving access to mammograms through community‐
based influenza clinics: A quasi‐
experimental study Simon MS et al. (2001)25 Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Author (Publication year) Country Title Findings test Int2: Control: invitations and reminders Reminder (mailed) standard postal reminder from their medical practitioner indicating that they should attend for the cervical smear test within the next 3 months. They were then sent a confidential postal questionnaire concerning their views of the cervical smear tests without the implementation intention. Media campaign to increase influenza Intention‐to‐treat analysis at 6‐month follow‐up: clinic attendance Int1(a&b): Int1a: Had mammogram: 35% Willing to receive a call from a local Had not had mammogram: 36% hospital based radiology to schedule a Mammogram status unknown: 29% mammogram + call from facility + 6‐month follow‐up survey Int2: Int1b: Had mammogram: 15% Unwilling to receive a call from a local Had not had mammogram: 48% hospital based radiology facility + 6‐month Mammogram status unknown: 37% follow‐up survey Int2: Sensitivity analysis based on the number of Participants were asked if they would mammograms obtained by women whose status was accept a call in 6 months concerning a few unknown: brief questions about the use of preventive Int1 vs Int2: RR = 1.6–2.1; Chi sq = 8.51–12.2 care in the community + 6‐month follow‐ p < 0.001 up survey Int1: Invitations and reminders + provider Intention‐to‐treat prompt Clinic 1: Int1 = 19%, Int2 = 20%, Int3 = 17%, p = 0.743 Nature of the intervention VOL 3– 158 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Women ≥ 39.5 Enrollees in two Detroit Health Department primary care clinics who became due for mammography screening during the course of the project and who had not had a prior abnormal mammogram. Sowjanya AP et al. (2009)160 INDIA Level III‐2 Cervical Women Slater J S et al. Level IV (2005)20 Breast USA Women 40–64 Effect of direct mail Low‐income, underinsured women in as a population‐based Minnesota who are eligible for the free strategy to increase screening services. mammography use among low‐income underinsured women ages 40 to 64 years Author (Publication year) Country Title USA The effect of patient reminders on the use of screening mammography in an urban health department primary care setting Findings Reminder letter to prompt physician visit Clinic 2: Int1 = 11%, Int2 = 14%, Int3 = 11%, p = 0.376 to get a referral for a mammogram (physician referral) + reminders attached Int1 vs Int3: OR = 1.1 (95% CI 0.7–1.6) to medical records to prompt physicians Int2 vs Int3: OR = 1.3 (95% CI 0.9–1.8) Int2: Invitations and reminders Clinic1 vs Clinic2: p < 0.001 Reminder letter to prompt at call to the BCCP to arrange a visit (direct access) Int3: Usual care No letter reminder Before intervention mail‐outs, letter was 1 year post‐intervention screening rate: sent to all primary care physicians to Int1: 232/11513 (2.02%) inform them of study and seek their Int2: 313/11513 (2.72%) support in terms of encouraging patients Int3: 194/11514 (1.68%) to service. Int1: Invitations and reminders: invitation Int1 vs Int3: diff = +0.23 (95% CI 0.09–0.37); p < 0.005 (mailed) Int2 vs Int3: diff = +0.75 (95% CI 0.58–0.92); p < 0.001 Folded card with brief message about the Int2 vs Int1: diff = +0.52 (95% CI 0.32–0.72); p < 0.001 free mammogram and prompt to call toll‐
free number Int2: Invitations and Reminders + Economic Incentive: invitation (mailed) Int1+ $10 incentive if completed mammogram within ~ 1 year. Int3: Control (No mailing) Women in a subset of villages in the In the 8 villages where women were recruited to Community Access to Cervical Health Study provide a home‐based self collected sample were recruited at two separate time points regardless of previous participation, Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 159
Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) ≥ 25 for: Suitability of self‐
Residents in eight villages in the Ranga Int1: Clinic‐based self collection collected vaginal Reddy District of Andhra Pradesh State Int2: Home‐based self collection samples for cervical All women were provided with verbal and cancer screening in printed diagrammatic instruction for peri‐urban villages in collecting the vaginal swab by nurse. Andhra Pradesh, India Sox CH et al. Level II Int1: Education (provider) (1999) 164 Cervical Training for Community Health Aides in USA Women intervention villages to undertake Pap 18–75 tests Improved access to Residents in four intervention villages in women's health the Yukon‐Kuskokwim Delta region of Int2: Usual care services for Alaska south‐western Alaska. Village inclusion natives through criteria: i) estimated minimum of 90 target community health group women in the village; ii) at least two aide training female Community Health Aides (CHAs) who had been on the village clinic staff for at least two years and who were interested in additional training to provide the target services; iii) the village tribal council indicated their interest in the participation of the CHAs of their village; iv) clinic facility offered privacy appropriate for genital examination; v) for budgetary reasons, the round‐trip airfare to Bethel (the main centre) was less than $225. Residents in four comparison villages were paired to intervention villages in terms of size and distance from Bethel. Stein K et al. Level I Int1: telephone call from a nurse Cervical screening rate within 90 days of intervention Proportion of women with Pap test based on Pap registries: 12 months prior to project vs 12 months since baseline Int1: 0.44 (95% CI 0.36–0.52); 0.48 (95% CI 0.38–
0.58) Int2: 0.42 (95% CI 0.34–0.50); 0.39 (95% CI 0.35–
0.43) Int1 vs Int2: p = 0.37 Int1: 257/564 (45.6%) Int2: 324/564 (57.5%) Post hoc significance test by CHPPE chi sq = 15.933, df = 1, p = 0.000 Findings VOL 3– 160 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Cervical Women 39–64 Random sample of Devon residents identified from NHS screening records as having no smear for at least 15 years. Level II Breast Women 50–80 Participants in five research projects undertaken by the Breast Screening Consortium funded through a NCI interactive grant in Washington, New York, Los Angeles, Massachusetts and North Carolina. Stokamer CL et al. Level II (2005)190 Colorectal USA Women Men ≥ 50 Randomized Outpatients of primary care clinics at VA controlled trial of the New York Harbor Healthcare System in impact of intensive New York City, who had FOBT ordered by patient education on their physician, and were referred to a Author (Publication year) Country Title (2005)83 UK Improving uptake of cervical cancer screening in women with prolonged history of non‐
attendance for screening: a randomized trial of enhanced invitation methods Stoddard AM et al. (2002)109 USA Effectiveness of telephone counseling for mammography: results from five randomized trials Regular mammography use at follow‐up: Int1 vs Int2 Controlling for design factors and stage of adoption at baseline: OR = 1.08 (95% CI 0.91, 1.27) Controlling for design factors, stage of adoption at baseline, race, insurance: OR = 1.08 (95% CI: 0.91, 1.28) Return rates of FOBT cards within 6 months: Total: 462/788 (58.6%) Int1: 261/396 (65.9%) Int2: 201/392 (51.3%) Int1 vs Int2: p < 0.001 Int1: Education (face‐to‐face) Intensive Patient Education – One‐on‐one session by primary care nurse on importance of screening, how to collect specimens for FOBT, and 2‐page handout on CRC screening Int2: Usual care FOBT cards and written instructions from Int1: 1.4% (95% CI 0.38%–3.6%) Int2: 1.8% (95% CI 0.57%–4.0%) Int3: 4.6% (95% CI 2.5%–7.7%) Int4: 1.8% (95% CI 0.57%–4.0%) No stat sig diffs between groups Findings Baseline survey, 12‐month follow‐up survey Int1: Counselling Barrier‐specific counselling (telephone) Int2: Usual care Int2: letter from well‐known celebrity Int3: letter from local NHS Cervical Screening Commissioner Int4: Control Group No intervention Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 161
Taylor VM et al. (2002)100 USA A randomized controlled trial of interventions to Findings Prior to randomisation, 1965/3743 (52%) women scheduled a mammogram within two months of the 'mailed recommendation to schedule a mammogram'. Mammograms completed at 12 months by women randomised to the interventions Never‐screened: Int1: 18/163 (11%) Int2: 42/160 (26%) Int3: 39/165 (24%) Ever‐screened: Int1: 181/427 (45%) Int2: 261/425 (61%) Int3: 255/425 (60%) All women: Int2 vs Int1: adjusted HR = 1.9 (95% CI 1.6–2.4) Int3 vs Int1: adjusted HR = 1.9 (95% CI 1.5–2.2) Level II Introductory letter, baseline face‐to‐face Pap testing between randomisation and six‐month Cervical survey, randomisation, follow‐up survey 6 follow‐up: Women months after randomisation Int1: 50/129 (39%) 20–69 Int2: 35/139 (25%) Chinese women in Seattle (Washington) Int1: Coaching Int3:20/134 (15%) and Vancouver (British Columbia). High intensity outreach worker based, Inclusion criteria included: i) under‐utilisers including: i) introductory mailing; ii) home Int1 vs Int3: p < 0.001 Level of evidence (NHMRC) Type of cancer Nature of the intervention Participants (sex), Participants (age) Participants (cohort) primary care nursing for patient education manufacturer on how to collect specimens and distribution of FOBT kits. for FOBT. Taplin SH et al. Level II Int1: Invitations and reminders (2000)42 Breast Mailed recommendation to schedule USA Women mammogram + mailed postcard reminder 50–79 Testing reminder and Stratified random sample of enrollees in Int2: Invitations and reminders motivational the breast cancer screening program in an Mailed recommendation to schedule telephone calls to HMO on Puget Sound, Seattle who met the mammogram + telephone reminder increase screening following selection criteria: i) no history of mammography: a breast cancer; ii) no prior involvement in Int3: Invitations and reminders randomized study mammography recruitment studies; iii) Mailed recommendation to schedule residence in the regions served by two mammogram + motivational telephone call
screening centres; iv) due for a mammogram; and v) English speaking. Author (Publication year) Country Title compliance with fecal occult blood testing VOL 3– 162 Knowledge translation: a review of strategies to increase participation in cancer screening
Level II Breast Women Taylor V et al. (1999)175 USA Nature of the intervention Int1: CME for doctors + multiple clinic‐
based and patient follow‐up components visit (video, motivational pamphlet, educational brochure, fact sheet tailored, counselling); iii) logistical assistance offered at outreach worker's discretion; and iv) follow‐up telephone call. Int2: Education Low intensity direct mailing (cover letter, video) motivational pamphlet, educational brochure, fact sheet) Int3: Control Taylor VM et al. Level II Int1: Coaching (face‐to‐face) (2002)125 Cervical Baseline survey plus multi‐faceted USA Women intervention delivered by bilingual, ≥ 18 bicultural Cambodian women including: Evaluation of an Cambodian American residents who i) home visits, group meetings in outreach intervention participated in a community‐based survey neighbourhood settings; and to promote cervical in the southern and central regions of ii) logistic assistance accessing screening cancer screening Seattle. Exclusion criteria included: services + follow‐up survey. among Cambodian personal history of invasive cervical cancer American women and/or hysterectomy, participation in an Int2: Control group earlier qualitative study about cervical Baseline survey and follow‐up survey cancer, did not live in close proximity to other respondents and could not be assigned to a neighbourhood group. Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) of Pap testing; ii) spoke Cantonese, Mandarin or English; iii) no history of invasive cervical cancer; iv) not had a hysterectomy. Under‐utilisers were defined as women who had not had Pap testing in the 2 years prior to the baseline survey and/or did not intend to have a test in the next 2 years. Author (Publication year) Country Title promote cervical cancer screening among Chinese women in North America Follow‐up 12 months after baseline survey: Ever had a Pap test (baseline; follow‐up) Int1: 108/144 (75%); 121/144 (85%); diff = +10%; p = 0.011 Int2: 112/145 (77%); 122/145 (84%); diff = + 7%; p = 0.059 Pap test in last 12 months (baseline; follow‐up) Int1: 64/144 (44%); 87/144 (61%); diff = +17%; p < 0.001 Int2: 74/145 (51%); 90/145 (62%); diff = +11%; p = 0.027 Int1 vs Int2: not stat sig Completed mammography within 8 weeks of follow‐
up: Ever‐screened: Int2 vs Int3: p = 0.03 Int1 vs Int2: p = 0.02 Multiple regression controlling for variables such as age and education Int1 vs Int3: OR = 3.5 (95% CI 1.9–6.6) Int2 vs Int3: OR = 2.0 (95% CI 1.1–3.7) Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 163
Findings Int1: 63% Int2: 23% p < 0.05 Never‐screened: Int1: 37% Int2: 21% p < 0.01 Bivariate comparisons of mammography completion rates among subgroups indicated the intervention effect was significant for both 50–59 and 60–74 age groups, African American and white women, patients with Medicare coverage and the uninsured. Thompson B et al. Level IV Int1: Community intervention Follow‐up at approximately 2.5 years after the (2006)143 Breast Colorectal Cervical Baseline survey + community intervention baseline survey and adjusted for baseline USA Women Men aimed at increasing screening rates and characteristics: ≥ 40 breast; ≥ 50 colorectal; > 18 cervical lifestyle practices (vegetable consumption No stat sig differences between Int1 and Int2 in Celebremos la salud! Residents of 20 rural agricultural area and smoking) lasting approximately 30 either the Hispanic or non‐Hispanic or white survey a community communities (Yakima Valley) of months + follow‐up survey respondents in screening randomized trial of Washington State and three communities cancer prevention in Columbia Basin. Int2: Control Group (United States) Baseline and follow‐up survey Thomson RM et al. Level II Int1: Mobile breast screening unit Participation rate amongst eligible attendants to (2009)158 Breast mobile breast screening unit: NEW ZEALAND Women 2005: 185/189 (97.9%) 50–64 Post hoc CHPPE sig test: Improving Rural, coastal, predominantly Maori chi sq = 0.027, df = 1, p = 0.870 participation in breast community in the Eastern Bay of Plenty 2007: 207/212 (97.6%) screening in a rural general practice with Author Level of evidence (NHMRC) (Publication year) Type of cancer Nature of the intervention Country Participants (sex), Participants (age) Title Participants (cohort) 50–74 Int2: Control group A clinic‐based Established patients of a firm of the Adult Usual care mammography Medicine Clinic at the Harborview Medical intervention targeting Centre in Seattle. inner‐city women VOL 3– 164 Knowledge translation: a review of strategies to increase participation in cancer screening
Author (Publication year) Country Title a predominately Maori population Tifratene K et al. (2007)167 FRANCE Colorectal cancer screening program: cost effectiveness of systematic recall letters Tilley BC et al. (1999)152 USA The Next Step Trial: Impact of a worksite colorectal cancer screening promotion program Adjusted compliance (i.e. all recommended examinations during the 2‐year trial period) and coverage (at least one recommended examination during the 2‐year trial period) : Compliance (self‐reported; confirmed) Int1: 36%; 23% Int2: 35%; 19% Int1 vs Int2 (self‐reported): OR = 1.46 (95% CI 1.1–2.0) p = 0.006 Int1 vs Int2 (confirmed): OR = 1.71 (95% CI 1.1–2.7) p = 0.012 Coverage (self‐reported; confirmed) Int1: 61%; 47% Int2: 61%; 44% Int1 vs Int2 (self‐reported): OR = 1.33 (95% CI 1.1–1.6) p = 0.002 Int1 vs Int2 (confirmed): OR = 1.57 (95% CI 1.2, 2.0) p < 0.001 Int1: Company screening program + enhanced health promotion program offered each year approximately 6 weeks before the screening program was offered at the site. Int2: Control Company screening program (digital rectal exam, FOBT, flexible sigmoidoscopy at recommended intervals) Level II Colorectal Women, Men Unstated Past and present employees who had worked in the pattern and model making areas of the automotive industry at 20% effort (i.e. worked in the area for at least 20% of the time) for at least 2 years Findings FOBT rates Int1: 489/3508 (14.2%) Int2: 271/3450 (7.8%) Int1 vs Int2: p < 0.00001 No differences between groups’ completion of the tests in terms of age and gender. Nature of the intervention Level II Int1: Invitations and reminders (mailed) Colorectal Letter of invitation to contact GP + mail Women Men FOBT kit directly to non‐responders 50–74 Non‐responders to the medical phase at six Int2: Invitations and reminders (mailed) months who were resident in the 10th Letter of invitation to contact GP + recall arrondissement of Marseille letter with an order coupon to request a free FOBT test kit Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 165
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Level IV Colorectal Women, Men 50–74 Patients of participating general practices in New South Wales. Exclusion criteria included: i) poor English; ii) significant cognitive impairment; iii) serious physical or mental illness; iv) residents of nursing homes; v) personal history of colorectal cancer; vi) completed an FOBT, sigmoidoscopy or colonoscopy in the previous 2 years; vii) strong family history of colorectal cancer. Tu SP et al. Level II (1999)64 Breast USA Women ≥ 40 Breast cancer Enrollees in the Group Health Cooperative screening by Asian‐
of Puget Sound (an HMO in Washington American women in a state) Breast Cancer Screening Program managed care (BCSP). Inclusion criteria included no prior environment history of breast cancer. Trevena LJ et al. (2008)88 AUSTRALIA Randomized trial of a self‐administered decision aid for colorectal cancer screening Author (Publication year) Country Title All completed a baseline survey to determine their breast cancer (BC) risk factors; all received a BC Screening Program (BCSP) recommendation for a screening mammogram between 1 May 1988 and 30 April 1994. Int1: Mailed consumer version of Australian guidelines and questionnaire + 1‐month telephone follow‐up Int2: Mailed appropriate version of decision aid (DA) booklet and questionnaire + 1‐month telephone follow‐up Nature of the intervention Korean vs non‐Asian: Adjusted OR = 0.43 (95% CI 0.18–1.02); no stat sig ii) Mammogram participation in Chinese modelled separately by age group: 40–49 yrs Adjusted OR = 1.23 (95% CI 0.84–1.80) Vietnamese vs non‐Asian: Adjusted OR = 1.63 (95% CI 0.93–2.88) Japanese vs non‐Asian: Adjusted OR = 1.07 (95% CI 0.82–1.41) None specific Asian vs non‐Asian: Adjusted OR = 1.13 (95% CI 0.88–1.44) i) Mammogram participation with BCSP Asian names vs non‐Asian names: Adjusted OR = 0.95 (95% CI 0.80–1.14) Proportion of people who had completed FOBT at one‐month follow‐up: Int1: 6.6% Int2: 5.2%, p = 0.64 No effect on 'intention to screen' Findings VOL 3– 166 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Tuominen A et al. Level III‐3 (2009)162 Cervical FINLAND Females ≥ 30 Self‐sample HPV test ‐ Non‐responders to the initial invitation to Increasing screening cervical cancer screening in 2008 in Espoo. coverage among non‐
responders in Finland? Twinn S Level II (2001)91 Cervical HONG KONG Women Adult The evaluation of the Chinese women living in Hong Kong effectiveness of health education interventions in clinical practice: A continuing methodological challenge Author (Publication year) Country Title Community outreach program run by registered nurse involving: i) needs assessment; ii) development of community outreach program; iii) health education sessions for community groups; and iv) mail out of cervical cancer and screening to identified housing estates. Int1: Procedure Mailed self‐sampling devices Int2: Reminder (mailed) Usual care (mailed reminder) Nature of the intervention Analysis of records maintained by service providers for changes in attendance rates from 4 weeks prior to commencements to 6 weeks after the end of the project Family Planning Association: Jan–Mar: 2233 Apr–Jun: 2869 Jul–Sep: 3327 Oct–Dec: 3391 Department of Health: Jan–Mar: 975 Apr–Jun: 926 Jul–Sep: 1215 Oct–Dec: 917 No stat test. 50–64 yrs Adjusted OR = 0.66 (95% CI 0.44–0.97); stat sig ≥ 65 yrs Adjusted OR = 0.43 (95% CI:0.43–0.90); stat sig Int1: 314/1133 (27.7%) Int2: 787/3031 (26.0%) Int1 vs Int2: Not stat sig Findings Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 167
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Level II Baseline survey (questionnaire and Breast, Cervical telephone interview) Women 52–69 Int1: In‐reach: Coaching (face‐to‐face) Members of an HMO servicing the Tailored in‐reach intervention delivered at Portland‐Vancouver metropolitan areas of the site of a primary care visit (patients Oregon and southwest Washington. intercepted post visit to reinforce any Inclusion criteria: i) health plan members clinician suggestion regarding screening for 3 years; ii) no record of a hysterectomy and facilitate scheduling a screening or bilateral mastectomy; iii) no record of a appointment) mammogram in the prior 2 years and a Pap test in the prior 3 years. Int2: Outreach: message framing + counselling (telephone) Tailored, sequential mail‐telephone outreach (tailored letter and information at the beginning of the intervention + telephone call to participants who had received one or none of the services at six months) Int3: Combined Combined In‐reach (Int1) and outreach (Int2) Int4: Usual care No intervention contact with the study Van Harrison R et al. Level IV Int1: Message framing (mailed) 73
(2003) Breast Personally addressed letter from the USA Women Medical Director of Michigan Medicare Valanis BG et al. (2002)72 USA Screening HMO women overdue for both mammograms and pap tests Author (Publication year) Country Title All mammograms (screening and diagnostic) at 14 months post‐intervention: Total sample: Services obtained by 14 months after implementation of the interventions Int1: No services 62%; Mammogram only 4%; Pap test only 8%; Pap + mammogram 26%; Int1 vs Int4 p = 0.511 Int2: No services 51%; Mammogram only 5%; Pap test only 3%; Pap + mammogram 39%; Int2 vs Int4 p = 0.006 Int3: No services 50%; Mammogram only 9%; Pap test only 4%; Pap + mammogram 32%; Int3 vs Int4 p = 0.05 Int4: No services 68%; Mammogram only 6%; Pap test only 6%; Pap + mammogram 19% Findings VOL 3– 168 Knowledge translation: a review of strategies to increase participation in cancer screening
Level II Cervical Women mean age of 62 years Women from rural Crete Vivilaki V et al. (2005)127 GREECE Are health education meetings effective in recruiting women in cervical screening programmes? An innovative and inexpensive intervention from the island of Crete Vogt TM et al. (2003)181 USA The safety net: A cost‐
effective approach to improving breast and Level IV Breast, Cervical Women Breast 40–70; Cervical 18–70 Members of Northwest Kaiser Permanente (NWKP) HMO in Portaland Oregon who had ≥ 3 years of continuous membership Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) ≥ 70 Medicare beneficiaries living continuously in Michigan. Inclusion criteria included: i) no Medicare claims for a mammogram; ii) no claims for diagnoses or procedures associated with medical conditions that may affect screening; iii) not enrolled in managed care during the study period. Author (Publication year) Country Title Personalized targeted mailing increases mammography among long‐term noncompliant Medicare beneficiaries: a randomized trial Findings Int1: Invitations and reminders (mailed) Letter on health plan stationery signed by two of the investigators and a brochure that answered frequently asked question about breast or cervical cancer screening + second letter if no record of screening 6 weeks after the first letter . Mammogram Int1: 24% (95% CI 19–29) Int2: 51% (95% CI 44–59) Int3: 50% (95% CI 44–57) Int4: 10% (95% CI 7–14) Int1 vs Int4: OR = 2.82 (95%CI 1.74–5.55) p < 0.0001 Int2 vs Int4: OR = 9.63 (95%CI 5.98–15.51) p < 0.0001 Int1: 8.1% Int2: 5.2% Int1 vs Int2: diff = +2.9%; OR = 1.60 (95% CI 1.15, 2.21); p < 0.005 Urban black: Int1: 9.2% Int2: 4.9% Int1 vs Int2: diff = +4.3%; OR = 2.09 (95% CI 1.04–
4.22); p ≤ 0.05 Int1: Group educational discussion meeting 16 women participated in the education session + invitation to organise a group visit to a recruited 32 non‐participants to attend screening, rural primary healthcare centre 52.1% of these women had not been screened previously mailed with informational materials Int2: Control Usual care Nature of the intervention Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 169
Level of evidence (NHMRC) Type of cancer Nature of the intervention Findings Participants (sex), Participants (age) Participants (cohort) and not received a: a) mammogram; or b) Int3 vs Int4: OR = 9.22 (95%CI 5.79–14.68) p < 0.0001 a Pap test during the three years. Int2: Invitations and reminders (mailed and Exclusion criteria for the mammography telephone) Pap test study included previous diagnosis of cancer Same initial letter as Int1 + phone call if no Int1: 22% (95% CI 17–28) or benign breast disease, bilateral record of screening 6 weeks after the first Int2: 54% (95% CI 47–61) mastectomy. Pap test exclusion criteria letter Int3: 50% (95% CI 42–58) included prior hysterectomy, previous Int4: 17% (95% CI 13–22) history of cervical cancer, cervical dysplasia Int3: Invitations and reminders (telephone) Int1 vs Int4: OR = 1.37 (95%CI 0.89–2.12) p < 0.16 or hysterectomy. Two telephone calls; the second coming 6 Int2 vs Int4: OR = 5.57 (95%CI 3.64–8.53) p < 0.001 weeks after the first if no screening record Int3 vs Int4: OR = 4.77 (95%CI 3.08–7.39) p < 0.0001 Int4: Usual care Routine system and environmental reminders Wikstrom I et al. Level II Invitations and Reminders (mailed) Returning rate of self‐sampling device (SSD) (2007)161 Cervical Sent letter saying that they were going to = 106/183 (58%) SWEDEN Women be offered the possibility of self‐sampling 35–55 of vaginal smear at home and that the Attitudes to self‐
People registered on the database of the collected material was to be sent to a sampling of vaginal Department of Cytology, University laboratory for analysis of the presence of smear for human Hospital at Uppsala who had not attended high‐risk HPV, the reason for their papilloma virus the organised gynaecological screen for selection, and the importance of HPV analysis among over 6 years, who had not had a infection for the development of women not attending hysterectomy and who could be contacted premalignant cell alterations on the cervix. organized cytological by mail. One week later a self‐sampling device was screening sent. Two months later all women received a questionnaire and reminder. Williams‐Piehota P et Level II All participants received: Overall mammogram rate at: al. Breast i) Message framing (telephone) 6 months: 35.4% (2005)65 Women Women completed baseline telephone 12 months 57.8% Author (Publication year) Country Title cervical cancer screening VOL 3– 170 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Nature of the intervention Findings Participants (sex), Participants (age) Participants (cohort) ≥ 41 survey, including assessment of monitor‐ (Chi sq 91, n = 232) = 3.06, p < 0.05. Women who called the Cancer Information blunter coping style, then listened to short, Service, New England. randomly assigned message promoting Matched message vs mismatched message for the Criteria: i) had not previously participated mammography (tailored to monitor or coping style of participants at 6 months: in a study; ii) did not call about themselves; blunter coping style). Overall: Chi sq(1) = 9.69, p < 0.01 iii) were not cancer patients; iv) received < ii) Message framing (mailed): Blunters: Chi sq (1, n = 147) = 5.77, p < 0.05 50% of recommended mammograms for After a few days, they were mailed packet: Monitors: Chi sq (1, n = 120) = 0.39, ns. their age tailored brochure (consistent with type of Matching message to coping style no longer had message delivered over phone), similarly differential influence on utilisation at 12 months tailored fridge magnet, survey to complete, plus $10 compensation when returned survey. Williams‐Piehota P et Level III‐2 Int1: Message framing (telephone and Overall mammogram rate at: al. (2003)68 Breast mailed) 6 months: 35.4% USA Women Telephone message prompting 12 months 57.8% ≥ 41 mammography (tailored to those with high (Chi sq 91, n = 232) = 3.06, p < 0.05 Matching health Women calling the Cancer Information Need for Cognition (NFC) plus similarly Mammogram in individuals low in NFC who were messages to Service, New England. tailored pamphlet 1 month later, and exposed to message tailored to: information‐
Criteria: i) not calling for themselves; ii) not fridge magnet, and survey to complete Low NFC: 36.6% processing styles: called previously during study period; iii) at after reading pamphlet High NFC: 37.9% Need for cognition least 41 years; iv) not current cacner Mammogram in individuals high in NFC who were and mammography patients for waiting for results etc; v) Int2: Message framing (telephone and exposed to message tailored to: utilization received < 50% of mammograms for their mailed) Low NFC: 29.1% age. Telephone message prompting High NFC: 39.3% mammography (tailored to those with low NFC) plus similarly tailored pamphlet 1 Authors stated confirming message‐matching month later, and fridge magnet, and hypothesis survey to complete after reading pamphlet
Wood RY et al. Level III‐2 Int1: Education (video)
Mammogram within 8 months post‐study: 99
(2004) Breast Baseline interview + video + follow‐up Int1: 36/75 (48%) USA Women interview 2 weeks later + telephone call 8 Int2: 24/68 (36%) Author (Publication year) Country Title USA Matching health messages to monitor‐
blunter coping styles to motivate screening mammography Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3– 171
Worden JK et al. (2002)131 USA Evaluation of a community breast screening promotion program Young RF et al. (2002)172 USA A breast cancer education and on‐site screening intervention for unscreened African American women Author (Publication year) Country Title Video breast health kits: Testing a cancer education innovation in older high‐risk populations Level II Educational program offered in primary Breast care sites: Women ≥ 40 Int1: Education (face‐to‐face); Economic African American, low‐income women who (mobile van) were patients at medical centres in Detroit, 1) Targeted cancer education (1 hour) with poor screening histories. tailored for African‐American women, and taught by African‐American cancer educator; and 2) Appointment for free on‐site mammography (mobile unit parked outside each site). Int2: Control group Observational assessment that included Screening rates at 3‐month follow‐up: (amongst 86% of women i.e. not intention to treat) Int1: 66% Int2: 38% Int1 vs Int2: p < 0.05 Level of evidence (NHMRC) Type of cancer Nature of the intervention Findings Participants (sex), Participants (age) Participants (cohort) ≥ 60 months later to assess mammogram Chi sq = 1.7, p ≤ 0.09 Subgroup of non‐compliant women (had compliance not had mammography in last 15 months or more at baseline) recruited to field test Int2: Control an age and ethnically sensitive video Baseline interview + follow‐up interview 2 program designed to enhance breast weeks later + telephone call 8 months later cancer screening in older black and white to assess mammogram compliance women. Level II Comprehensive community program No evidence that the breast screening promotion Breast involving: i) public education, promotional program achieved higher screening levels Women events, and messages in the mass media; Adults ii) physician education; iii) mammography Residents in one intervention area and two registry and training for radiologists; iv) comparison areas in Florida. support for low‐income women VOL 3– 172 Knowledge translation: a review of strategies to increase participation in cancer screening
Level of evidence (NHMRC) Type of cancer Participants (sex), Participants (age) Participants (cohort) Nature of the intervention Findings questionnaire administered by telephone (history, barriers, knowledge etc.) Zhu K et al. Breast Community‐based intervention: No significant differences in mammography between (2002)121 Women the 2 groups. USA ≥ 65 Int1: Coaching (face‐to‐face) Single (never married, widowed, divorced, Women from 5 public housing complexes An intervention study separated) African American women received education in their home, including on screening for residing in public housing complexes in how to overcome barriers or problems breast cancer among Nashville, with no history of breast cancer. from an African American lay health single African‐
educator, plus received brochures on American women breast health and psychological aged 65 and older adjustment. Int2: Control group Women from 5 public housing complexes 3 time points (self‐report): pre‐
intervention (baseline), post‐intervention (12 months after intervention), follow‐up (12 months after post‐intervention) Author (Publication year) Country Title Knowledge translation: a review of strategies to increase participation in cancer screening VOL 3–173