DOI: 10.1111/j.1464-5491.2005.01778.x
Measuring beliefs about taking hypoglycaemic
medication among people with Type 2 diabetes
Medication
Original
Article
article
beliefs among
Oxford, UK
Diabetic
DME
Blackwell
0742-3071
22
Medicine
Publishing,
Ltd.people with Type 2 diabetes A. Farmer et al.
2005
A. Farmer, A.-L. Kinmonth* and S. Sutton*
Abstract
Department of Primary Health Care, University of
Oxford, Oxford and *General Practice and Primary
Care Research Unit, University of Cambridge,
Institute of Public Health, Cambridge, UK
Accepted 22 May 2005
Final Acceptance 27 August 2005
Aims Identifying patients’ beliefs about taking medication can inform interventions
to support medication taking, and their evaluation. We set out to establish the range
of these beliefs, and measure the frequency of commonly held beliefs and their correlation with intention to take medication and self-reported medication adherence.
Methods An exploratory survey among Type 2 diabetic patients aged 40 years or
older, registered in general practice, used a questionnaire measuring a range of plausible
beliefs about taking and intention to take medication developed from interviews
where belief elicitation was guided by the Theory of Planned Behaviour. The Medication Adherence Report Schedule was used as a self-report adherence measure.
Results Questionnaires were returned by 121 (61.7%) people. The majority
strongly agreed with statements about the benefits of taking medication. Negative beliefs that taking medication would ‘cause unpleasant side effects’ and
‘lead to weight gain’ were held by 24.1 and 13.9% of people, respectively.
Beliefs about benefit were strongly associated with intention to take medication
regularly. Two beliefs were associated with reduced medication adherence:
‘changes to my daily routine would make it more difficult to take my diabetes
medicines regularly’ (P < 0.001), and ‘if I were to take my diabetes medicines
regularly this would lead to my gaining weight’ (P < 0.05).
Conclusions Use of a theoretical model to elicit and identify common beliefs
about taking medication regularly underscores the importance of exploring
weight-gain concerns and how to keep taking tablets when routines change.
Beliefs associated with intention and taking medication will inform intervention
development, implementation and evaluation in randomized controlled studies.
Diabet. Med. 23, 265–270 (2006)
Keywords adherence, beliefs, medication compliance, Type 2 diabetes
Abbreviations IQR, interquartile range; MARS, Medication Adherence Report
Scale; MDQ, Medicines for Diabetes Questionnaire; TPB, theory of planned
behaviour
Background
Intensive control of blood glucose and associated cardiovascular risk factors offers the possibility of reducing the disease
burden in people with Type 2 diabetes [1–3]. A major problem
Correspondence to: Dr Andrew Farmer, Department of Primary Health Care,
University of Oxford, Old Road Campus, Headington, Oxford OX3 7LF, UK.
E-mail: [email protected]
© 2005 Diabetes UK. Diabetic Medicine, 23, 265–270
in translating this into improved outcomes is non-adherence to
medication. Across a wide range of conditions adherence may
be as low as 50% [4]. In a Tayside study, less than one-third of
patients with diabetes were dispensed their single hypoglycaemic drug sufficiently regularly to cover 90% of treatment days
in 1 year [5]. Intensive treatment regimens will not improve
outcomes if prescribed medications are not taken regularly.
Although a number of interventions have a limited impact
on adherence, for example simplified dosage regimens, self
265
266
Medication beliefs among people with Type 2 diabetes • A. Farmer et al.
monitoring of medication taking, and prompts and reminders
[6], there is limited understanding of what intervention elements
are effective. Therefore, success in transferring interventions
between different settings is limited. Interventions with the
most impact appear to be those using behavioural modification techniques and building on a participant’s own motivations within a psychological framework [7].
Qualitative studies have confirmed that people hold a range
of beliefs about their medication [8]. A range of psychological
models have been used to identify beliefs that may be related to,
and can be used to explain variations in, medication taking [9].
For example, the Health Belief Model was used to guide work
that showed that beliefs about ‘adverse consequences following
failure to take medication’ were associated with improved adherence [10]. Other studies have used an illness representation or a
self-efficacy approach [11,12]. An approach widely used in other
studies of relationships between beliefs, intentions and behaviour
is the theory of planned behaviour (TPB) [13,14]. We adopted
this model because of the tightly specified procedures for eliciting
beliefs and structuring questionnaires [15]; the focus of the model
on beliefs about a target behaviour (in this case taking medication), rather than beliefs about a disease or treatment more generally; and evidence from a large number of studies that the model
has predictive validity, explaining between 35 and 50% of the
variance in intentions and between 26 and 35% of the variance
in behaviour across a range of different target behaviours [16].
The TPB specifies three sets of beliefs that influence intentions and behaviour: behavioural beliefs, normative beliefs,
and control beliefs. This approach has a strong face validity in
the context of medication taking among people with Type 2
diabetes: behavioural beliefs are beliefs about the personal
advantages and disadvantages of taking medication (e.g.
‘taking my medication regularly would make me feel better’),
normative beliefs are beliefs about the views of significant
others (e.g. ‘my wife would approve of me taking my medication
regularly’), and control beliefs are beliefs about the factors
that may make it easier or more difficult to take medication
regularly (e.g. ‘changes to my daily routine would make it
more difficult to take my medication regularly’).
This paper describes the use of structured interviews to elicit
specific, potentially influential beliefs about taking diabetes
medications from people with Type 2 diabetes recruited from
primary care within the framework of the TPB. A cross
sectional survey was then used to establish (i) the frequency
with which the identified beliefs were held, and (ii) the extent
to which these beliefs were correlated with intention to take
medication and self-reported adherence to medication.
Participants and setting
Aylesbury Vale Research Ethics Committee approved the
study. Type 2 diabetic patients taking oral hypoglycaemic
medication, but not using insulin, aged 40 years or older and
not excluded by their general practitioner on the grounds of
frailty or inability to complete the questionnaire, were eligible.
The majority received care in the practice, with a minority attending hospital diabetes clinics.
Practice sizes ranged from 8000 to 15 000, with one serving
an urban population, and the others serving a mixed urban and
rural population. Two practices had an established nurse-run
diabetes service.
Of 697 people identified with diabetes in the three practices,
we identified 261 (37.4%) who fulfilled the inclusion criteria.
We randomly selected 200, subsequently excluding four who
had died or developed a serious illness after their initial selection, and sent 196 letters. Of those patients contacted, 121
(61.7%) returned a completed questionnaire.
Conduct of the study
Eligible patients were identified from the practice computer.
Each patient was sent a letter from their general practitioner
explaining the purpose of the study and enclosing a questionnaire,
a list of oral hypoglycaemic medication to assist in identifying
the medications that were being asked about, a consent form
and a Freepost return envelope. Reminders were sent by the
practices at 3 and 6 weeks. Following return of the consent
form, the practice-held medical records were reviewed to identify the date of diagnosis of diabetes.
Measures—self-reported measure of behaviour
The Medication Adherence Report Scale (MARS) was developed for measurement of adherence to a wide range of medication regimens [17]. The five statements comprising the scale
are: ‘I forget to take my diabetes medicines’, ‘I alter the dose of
my diabetes medicines’, ‘I stop taking my diabetes medicine for
a while’, ‘I decide to miss out a dose of my diabetes medicine’,
‘I take less diabetes medicine than instructed’. The MARS was
scored in accordance with standard practice with a maximum
score of 25 by summing the score from the five questions, each
with a five-point response scale (from ‘always true’ to ‘never’)
[17]. In previous pilot work for this study, we observed a
median score of 24 [interquartile range (IQR) 23–25] using this
measure in a similar population.
Medicines for Diabetes Questionnaire
Development
Patients and methods
Design
Beliefs, intentions and self-reported behaviour were measured
with a self-completion questionnaire sent by post to patients
registered with three Buckinghamshire general practices.
A series of brief interviews to identify salient beliefs about
taking medication for diabetes were carried out by a research
nurse. Thirty-nine non-housebound patients with non-insulinrequiring Type 2 diabetes were randomly selected from the
diabetes register of a Buckinghamshire practice and sent a
written invitation to be interviewed. Beliefs about taking medication were elicited from the participants using the standard
© 2005 Diabetes UK. Diabetic Medicine, 23, 265–270
Original article
procedures recommended for TPB studies [14]. In brief, a
structured interview ‘script’ was used to ensure that the preamble was the same for each participant. All salient beliefs were
first elicited and each belief subsequently explored to minimize
variation between interviews. The interviews were initially
taped to enable discussion with and feedback to the nurse on
the approach used. Behavioural beliefs were identified by asking
questions about the perceived advantages and disadvantages
of the behaviour of taking medicines for diabetes, normative
beliefs were identified by questions about who might approve
or disapprove of the behaviour, and control beliefs were identified by questions about what might make it easier or more
difficult to take medication regularly.
Thirty-three interviews were carried out, of which the first nine
were used, together with the results of a series of 13 preliminary
interviews, to develop a coding system for the subsequent
interviews. In the 24 subsequent interviews, the following
additional beliefs were identified: ‘it would keep my diabetes
under control’ and ‘putting my tablets out in a box would make
it easier’. Beliefs that occurred in 20% or more of the structured
interviews were carried forward into the final questionnaire.
Final questionnaire
The Medicines for Diabetes Questionnaire (MDQ) comprised
the beliefs identified above with a response scale of strongly
disagree (1) to strongly agree (5). Two additional questions
were included to assess behavioural intention using standard
phrasing [15]: ‘It is likely I will take my diabetes medicines regularly’ and ‘I intend to take my diabetes medicines regularly’.
The questionnaire was assessed for readability and clarity in a
pilot of 10 patients. All the items in the MDQ were scored on
a five-point scale from strongly disagree (1) to strongly agree
(5). The two questions measuring intention were summed to
create a measure of intention to take medication regularly with
a range from −4 to +4.
Analysis
Responses to questionnaires and information about duration of
diabetes were double entered onto computer, retaining only the
study number as a patient identifier.
The percentage of the study population agreeing or strongly
agreeing with each of the belief items is reported. Associations
between measures of intention, adherence and beliefs were
examined with a Spearman’s rank correlation coefficient calculated using the full range of responses to the categorical scales.
Differences in beliefs between groups were evaluated with a
Mann–Whitney U-test. Analysis was carried out using Stata 7.0
(Stata Corporation, College Station, TX, USA, Release 7.0, 2001).
Results
Sample characteristics
The mean age of respondents was 66 (interquartile range 57–
75) years, and 52.1% were male; 19.5% were living alone,
© 2005 Diabetes UK. Diabetic Medicine, 23, 265–270
267
66.1% lived with a partner and 14.2% lived with their family
or in other circumstances. The median duration of diabetes
was 6 years (IQR 1.7–8.2 years), and the median total number
of tablets taken per day was six (3–9).
Beliefs about taking medication
The vast majority of respondents agreed with statements
about the positive behavioural beliefs of taking medication.
Over 85% of people agreed that taking their medication regularly would ‘help them to stay well’, ‘reduce the chances of
developing complications’, ‘keep their blood sugar under control’ and ‘keep their diabetes under control’ (Table 1). Beliefs
about negative behavioural beliefs such as ‘would cause me
unpleasant side effects’ and ‘would lead to gaining weight’
were held by a minority of people (32.8 and 13.9%, respectively). The belief that taking diabetes medication regularly
would help them avoid having to inject insulin was held by
86.4% of people.
The majority of respondents also agreed with statements
about the positive normative beliefs associated with taking
medication. They agreed that their doctor or nurse (99.1%),
family (87.8%) and partner (75.7%) would approve of their
taking medication.
There was a wider range of response to the control beliefs.
Of respondents, 78.9% agreed that keeping to a regular routine
would make it easier to take medication regularly, but only
21.9% felt that changes to their routine would make taking
medication regularly more difficult. It was thought by 56.9%
of people that putting out their medication in a box would
make it easier to take it regularly. A majority of patients
(69.6%) also felt that having a regular review with their doctor
or nurse about their medication would be helpful.
The distribution of strength of individual beliefs about side
effects of medication did not differ between people taking a
sulphonylurea or metformin as medication.
There was no significant difference in distribution of the
strength of beliefs about ‘gaining weight’ between men and
women (Z = 1.7, P < 0.08). Gender differences did not explain
association between adherence and beliefs about gaining
weight with medication.
Intention to take medication and self-reported medication
adherence
The median score for intention to take medication regularly
was 3 (IQR 2– 4). Of participants, 43.3% strongly agreed with
both statements about intention to take medication regularly,
while only four people strongly disagreed. Responses to the
MARS measure of adherence were skewed with a median
score of 25 and an interquartile range of 24 –25; 43.8% scored
25, 39.5% scored 24 and 16.7% scored 23 or less. The majority of people with a submaximal MARS score indicated that it
was ‘rarely true’ or ‘sometimes true’, rather than ‘always’ or
‘often’ true that they forgot to take their medication. The
268
Medication beliefs among people with Type 2 diabetes • A. Farmer et al.
Table 1 Proportion of respondents endorsing elicited beliefs about taking diabetes medication regularly, and correlations between each of the items
with intention to take medication regularly and self-reported medication adherence
Variable
Behavioural beliefs
If I were to take my diabetes medicines regularly …
it would help me to stay well
it would reduce my chances of developing complications from diabetes
it would keep my blood sugar under control
it would keep my diabetes under control
it would help me avoid having to inject insulin
it would cause me unpleasant side effects (e.g. feeling sick or bloated)
this would lead to my gaining weight
Normative beliefs
My doctor or nurse would approve of me taking my diabetes medicines regularly
Members of my family or close relatives would approve of me taking my
diabetes medicines regularly
My wife/husband/partner would approve of me taking my diabetes medicines regularly
Control beliefs
Changes to my daily routine would make it more difficult for me to take
my diabetes medicines regularly
Putting out my tablets in a box would make it easier for me to take my
diabetes medicines regularly
Having a regular review with the doctor or nurse would make it easier
for me to take my diabetes medicines regularly
Keeping to a regular routine and being disciplined would make it easier
for me to take my diabetes medicines regularly
n†
% (95% CI)
Intention r
Adherence‡ r
108/117
108/118
108/118
101/115
102/118
38/116
16/115
92.3 (86.0–95.9)
91.5 (85.1–95.3)
91.5 (85.1–95.3)
87.8 (80.6–92.6)
86.4 (79.1–91.5)
32.8 (24.9–41.8)
13.9 (8.8–21.4)
0.59**
0.60**
0.49**
0.62**
0.51**
−0.03
−0.12
0.09
0.04
0.02
0.10
0.11
−0.08
−0.25*
115/116
101/115
99.1 (95.3–99.9)
87.8 (80.6–92.6)
0.58**
0.58**
−0.07
0.12
87/115
75.7 (67.1–82.6)
0.40**
−0.00
25/114
21.9 (15.3–30.4)
−0.18
66/116
56.9 (47.8–65.6)
0.04
−0.05
80/115
69.6 (60.6–72.2)
0.13
−0.13
90/114
78.9 (70.6–85.4)
0.25**
−0.33**
0.02
*P < 0.01; **P < 0.001.
†Number of people agreeing or strongly agreeing with statement out of those responding to question.
‡Spearman’s r uses categorical data from 5-point scale.
correlation between the intention to take medication measure
and the MARS was 0.29 (P < 0.01).
regularly would lead to weight gain was also associated with a
lower MARS score (P < 0.05).
Correlations between beliefs and intentions
Discussion
Positive behavioural beliefs about taking diabetes medications
were strongly associated with intention to take medication
(Table 1) (all P < 0.001). Negative behavioural beliefs (e.g. ‘it
would cause me unpleasant side effects’ or ‘it would lead to my
gaining weight’) were not associated with intention to take
medication. Normative beliefs were associated with intention
to take medication regularly (P < 0.001). The control belief
‘keeping to a regular routine would make it easier to take
medication regularly’ was associated with intention (P < 0.001),
but there was no association for ‘putting out my tablets in a
box would make it easier’ or ‘changes to my regular routine
would make it more difficult’.
We have used a standardized procedure to identify beliefs
about taking medication that appear to be commonly held by
people with Type 2 diabetes. Many of the beliefs supportive of
taking medication were held by the majority of respondents.
Negative beliefs about gaining weight and unpleasant side
effects were held by a significant minority. Intention to take
medication was strongly associated with positive beliefs about
clinical benefits of taking medication, but not with negative
beliefs. Beliefs about weight gain and changes to daily routine
were associated with lower reported medication adherence.
The results underscore the importance to patients of regular
medication reviews with a practitioner. They identify the
potential importance of exploring concerns about weight gain,
how to keep taking tablets regularly when routines change,
and any links made between adherence and avoiding insulin
injections.
This study was carried out in a setting and patient group
relevant to clinical practice and provided the information
about belief patterns in this population to inform an intervention, and measures of its effect, for evaluation in a clinical trial
Correlations between beliefs and adherence
Two beliefs were associated with a reduction in self-reported
medication adherence. There was a significant association
with the belief that ‘changes to my daily routine would make
it more difficult for me to take my diabetes medication regularly (P < 0.001). The belief that taking diabetes medicine
© 2005 Diabetes UK. Diabetic Medicine, 23, 265–270
Original article
269
of simple approaches to support medication adherence among
patients with diabetes in primary care.
of the drug, they are limited to assessing behaviour over only
a short period.
Potential underestimation of associations
Potential limitations of the underlying theoretical model
There are a number of reasons why we may have mis-estimated
or underestimated the strength of relationships between
beliefs, intentions, and behaviour. First, we may not have
identified commonly held beliefs among non-adherent groups.
However, few of the people approached declined to take part
in the initial interviews, and so the range of beliefs included in
the MDQ is likely to represent those held in the wider population, although the extent to which the same range of beliefs
are shared by people from other cultural groups remains to be
determined. Secondly, responders may over-represent adherent groups. The majority of those approached in the survey
also responded, but the extent to which these participants were
biased towards people taking their medication is unclear.
However, our study population does include people reporting
less than optimal adherence, and provides useful information
about their belief patterns.
Thirdly, the questionnaires measuring beliefs and behaviour
might have used wording that matched more closely. In
accordance with Ajzen and Fishbein’s [14] recommendations,
the measures of beliefs and intentions both referred specifically
to the same target behaviour (taking my diabetes medicines
regularly) to maximize the associations. However, the measure
of behaviour (the MARS questionnaire) asked about ‘not
taking’ medication rather than ‘taking’ medication and this
may have reduced the association between beliefs and
behaviour.
Fourthly, self-report measures of behaviour may underestimate non-adherence to medications, through a desire to
present oneself in a favourable light. Nevertheless, the measure
we used offers a well-tested ordinal scale to compare relative
adherence [18,19]. The desire to present oneself in a favourable light and preserve self-esteem through making the socially
desirable response of reporting high adherence, good intentions and acceptable beliefs may lead to false positive associations between beliefs and measures of intention and adherence
[20]. However, wide variation in correlations between different beliefs and intention, and absence of correlation between
many measures of belief and the MARS suggests that this was
not an overriding effect. The postal questionnaire, confidentiality
and appeal for honest responses were intended to minimize
these biases.
Alternative measures to overcome the disadvantages of self
report as an outcome measure would include use of actual
measures of behaviour or medication levels. Electronic
medication containers can be used to record each time the lid
is removed. However, knowledge that behaviour is being
observed may lead to alterations in taking medication. Neither
does removal of the lid of the container indicate that the
medication has been taken. Another potential direct measure of
adherence is medication levels, but, depending on the half-life
The TPB holds that intentions should be a strong predictor of
behaviour, at least for behaviours that are under the person’s
control. However, we found that some beliefs were associated
with intention but not with reported behaviour, and, conversely, that some beliefs were associated with behaviour but
not with intention. Participants’ beliefs about medication leading to weight gain were not associated with intention to take
medication, but were associated with reported adherence to
medication. It is possible, therefore, that, despite having an
intention to take medication, when faced with the behaviour,
concerns about weight gain lead to reduced medication adherence. Work among women with Type 1 diabetes has documented concerns about weight gain with insulin injection
omission. [21]. A cross-sectional survey suggested that up to
one-third of women omitted insulin doses and that this was
associated with disordered eating patterns and psychological
distress. However, we have not identified previous quantitative
studies showing similar findings among people with Type 2
diabetes taking oral hypoglycaemic medication.
Similarly, the belief that ‘changes to my daily routine would
make it more difficult to take medication’ correlated with
reported adherence but not with intention. This pattern could
arise from those reporting having taken their medication
regularly in the past perceiving future changes in routine as
a potential problem. Thus, apparent mismatch between our
findings and the underlying theory may have arisen from use
of a cross-sectional study design. A prospective study will be
required to see to what extent intention to take medication
regularly predicts subsequent medication taking.
The beliefs elicited in our study using the standardized methods of the TPB also encompass areas of concern identified
from previous qualitative work and from other psychological
approaches. For example, the specific concerns about the
harms and benefits of medication found in this study reflect the
general concerns identified using a ‘treatment representations’
approach [18,22]. Issues of personal control have also been
observed in cross-sectional studies of adherence to insulin
treatment [23]. Our findings are also consistent with previous
work using the TPB to examine medication adherence [24,25].
The data from this study does not indicate what type of
intervention might be effective in supporting medication
adherence. However, the data does inform the content of the
intervention. A framework where provision of tailored information is used to reinforce identified positive beliefs and problem solving used to address identified negative beliefs would
be a potential way forward. Provision of tailored information
about issues of weight gain and side effects is likely to be
particularly important. Exploratory studies will be required
to develop and evaluate potential effectiveness of such an
intervention.
© 2005 Diabetes UK. Diabetic Medicine, 23, 265–270
270
Medication beliefs among people with Type 2 diabetes • A. Farmer et al.
Conclusions
It is widely accepted that people’s beliefs about medication
should be elicited before offering them tailored information or
help [26]. We have elicited a range of beliefs about taking medication, several of which are quite commonly held. Communicating more clearly with patients and creating treatment plans
incorporating a patient perspective can usefully be informed
by these commonly held beliefs [27]. Concerns about effects of
medication and difficulties with taking medication have both
been shown to predict intention and behaviour [24,25].
We have used a predictive theoretical model to elicit and
identify common beliefs about taking medication regularly. The
model has captured beliefs that appear to be important in relation to both intention to take medication and in self-reported
behaviour. However, further work is required to explain the
discrepancy between the beliefs that correlate with intention
and those that correlate with self-reported medication adherence.
Beliefs influencing intention to take medicine will be used to
inform development and implementation of an intervention as
part of a programme of research into medication adherence in
primary care. Interventions to support adherence focusing on these
beliefs need to be evaluated in randomized controlled studies.
Competing interests
None declared.
6
7
8
9
10
11
12
13
14
15
16
17
Acknowledgements
18
We thank those patients who participated, Dr Duncan Burwood for help in working with practices, Bedgrove Surgery,
Aylesbury, Haddenham Health Centre, Cross Keys Surgery,
Princes Risborough, and Mandeville Surgery, Aylesbury. Mrs
Mary Selwood was the study research nurse. AF was funded
by an NHS R&D Primary Care Clinical Scientist award. The
study was funded by Diabetes UK. In addition to the authors,
the members of the SAM Study group are Dr S. Griffin, Dr T.
Prevost, Ms W. Hardeman, Prof. R. Horne to whom we are
indebted for wider discussions.
References
1 UK Prospective Diabetes Study (UKPDS) Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes (UKPDS 38). Br Med J 1998; 317: 703–713.
2 UK Prospective Diabetes Study (UKPDS) Group. Intensive blood
glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2
diabetes (UKPDS 33). Lancet 1998; 352: 837–853.
3 Collins R, Armitage J, Parish S, Sleight P, Peto R. MRC/BHF Heart
Protection Study of cholesterol-lowering with simvastatin in 5963
people with diabetes: a randomised placebo-controlled trial. Lancet
2003; 361: 2005–2016.
4 Sackett DL, Snow JC. Compliance in Health Care. Baltimore: Johns
Hopkins University Press 1979.
5 Donnan PT, MacDonald TM, Morris AD for the DARTS/MEMO
Collaboration. Adherence to prescribed oral hypoglycaemic medica-
19
20
21
22
23
24
25
26
27
tion in a population of patients with Type 2 diabetes: a retrospective
cohort study. Diabet Med 2002; 19: 279–284.
McDonald HP, Garg AX, Haynes RB. Interventions to enhance
patient adherence to medication prescriptions: scientific review. J Am
Med Assoc 2002; 288: 2868–2679.
Nagasawa M, Smith MC, Barnes JHJ, Fincham JE. Meta-analysis of
correlates of diabetes patients’ compliance with prescribed medications.
Diabetes Educ 1990; 16: 192–200.
Britten N. Patients’ ideas about medicines: a qualitative study in a
general practice population. Br J General Pract 1994; 44: 465–468.
Horne R, Weinman J. Predicting treatment adherence: an overview of
theoretical models. In: Myers L, Midence K, eds. Adherence to Treatment
in Medical Conditions. London: Harwood Academic 1998, 25–50.
Kelly GR, Mamon JA, Scott JE. Utility of the health belief model in
examining medication compliance among psychiatric outpatients.
Soc Sci Med 1987; 25: 1205–1211.
Hampson SE, Glasgow RE, Foster LS. Personal models of diabetes
among older adults: relationship to self-management and other
variables. Diabetes Educ 1995; 21: 300–307.
McCaul KD, Glasgow RE, Schafer LC. Diabetes regimen behaviours:
predicting adherence. Med Care 1987; 25: 868–881.
Hardeman W, Johnston M, Johnston DW, Bonetti D, Wareham NJ,
Kinmonth A-L. Application of the theory of planned behaviour in
behaviour change interventions: a systematic review. Psychol Health
2002; 17: 123–158.
Ajzen I, Fishbein M. Understanding Attitudes and Predicting Social
Behavior. Englewood Cliffs, NJ: Prentice Hall 1980.
Ajzen I. The theory of planned behaviour. Organizational Behav
Human Decision Processes 1991; 50: 179–211.
Sutton S. Determinants of health-related behaviours: theoretical and
methodological issues. In: Sutton S, Baum A, Johnston M, eds. The
Sage Handbook of Health Psychology. London: Sage 2004.
Horne R. Non-adherence to medication: causes and implications for
care. In: Gard P, ed. A Behavioural Approach to Pharmacy Practice.
Oxford: Blackwell 2001, 111–130.
Horne R, Weinman J. Patients’ beliefs about prescribed medicines
and their role in adherence to treatment in chronic physical illness.
J Psychosom Res 2001; 47: 555–567.
Enlund H, Tuomilehto J, Turakka H. Patient report validated against
prescription records for measuring use of and compliance with antihypertensive drugs. Acta Med Scand 1981; 209: 271–275.
Reiss S, Peterson RA, Gursky DM, McNally RJ. Anxiety sensitivity,
anxiety frequency and the prediction of fearfulness. Behav Res Ther
1986; 24: 1–8.
Polonsky WH, Anderson BJ, Lohrer PA, Aponte JE, Jacobson AM,
Cole CF. Insulin omission in women with IDDM. Diabetes Care
1994; 17: 1178–1185.
Horne R, Weinman J, Hankins M. The beliefs about medicines questionnaire: the development and evaluation of a new method for assessing the
cognitive representation of medication. Psychol Health 1999; 14: 1–24.
Bradley C, Gamsu DS, Moses JL, Knight G, Boulton AM, Drury J
et al. The use of diabetes specific perceived control and health belief
measures to predict treatment choice and efficacy in a feasibility
study of continuous subcutaneous insulin infusion pumps. Psychol
Health 1987; 1: 133–146.
Abraham C, Clift S, Grabowski P. Cognitive predictors of adherence
to malaria prophylaxis regimens on return from a malarious region:
a prospective study. Soc Sci Med 1999; 48: 1641–1654.
Conner M, Black K, Stratton P. Understanding drug compliance in a
psychiatric population: an application of the theory of planned
behaviour. Psychol Health Med 1999; 3: 337–344.
Vermeire E, Hearnshaw H, Van Royen P, Denekens J. Patient
adherence to treatment: three decades of research. A comprehensive
review. J Clin Pharm Ther 2001; 26: 331–342.
Heller S. Mistaken priorities in diabetes research. Diabet Med 2002;
19: 263–264.
© 2005 Diabetes UK. Diabetic Medicine, 23, 265–270