Clinical Infectious Diseases
SUPPLEMENT ARTICLE
A Qualitative Study Investigating Experiences, Perceptions,
and Healthcare System Performance in Relation to the
Surveillance of Typhoid Fever in Madagascar
Alfred Pach,1 Michelle Warren,2 Irene Chang,1,3 Justin Im,1,3 Chelsea Nichols,1 Christian G. Meyer,4 Gi Deok Pak,1 Ursula Panzner,1 Se Eun Park,1
Vera von Kalckreuth,1 Stephen Baker,5 Henintsoa Rabezanahary,6 Jean Philibert Rakotondrainiarivelo,6 Tiana Mirana Raminosoa,6
Raphaël Rakotozandrindrainy,6 and Florian Marks1
1
International Vaccine Institute, Seoul, Korea; 2University of Minnesota, Minneapolis; 3London School of Hygiene and Tropical Medicine, United Kingdom; 4Institute of Tropical Medicine,
Eberhard-Karls University Tübingen, Germany; 5Hospital for Tropical Diseases, Wellcome Trust Major Overseas Programme, Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam; and
6
University of Antananarivo, Madagascar
Background. The burden of typhoid fever (TF) in sub-Saharan Africa is largely unknown but is increasingly thought to be high,
given that water and sanitary conditions remain unimproved in many countries. To address this gap in information, the Typhoid
Fever Surveillance in Africa Program (TSAP) founded a surveillance system for TF in 10 African countries. This study was a
component of the TSAP surveillance project in Madagascar.
Methods. The study entailed a qualitative assessment of patients’ experiences and perceptions of services for febrile symptoms at
the studies’ rural and urban sentinel public health clinics. The study examined influences on the use of these facilities, alternative
sources of care, and providers’ descriptions of medical consultations and challenges in providing services. Data were collected
through semistructured and open-ended individual interviews and a focus group with patients, caregivers, and medical personnel.
Results. Thirty-three patients and 12 healthcare providers participated in the data collection across the 2 healthcare facilities. The
quality of services, cost, and travel distance were key factors that enabled access to and use of these clinics. Divergent healthcareseeking patterns were related to variability in the care utilized, socioeconomic status, and potential distance from the facilities. These
factors influenced delivery of care, patient access, and the health facilities’ capacity to identify cases of febrile illness such as TF.
Conclusions. This approach provided an in-depth investigation and understanding of healthcare-seeking behavior at the study
facilities, and factors that facilitated or acted as barriers to their use. Our findings demonstrate the relevance of these public health
clinics as sites for the surveillance of TF in their role as central healthcare sources for families and communities within these rural and
urban areas of Madagascar.
Keywords. typhoid fever; surveillance; healthcare utilization; illness experience.
Globally, it is estimated that there are 21.7 million new cases
and 217 000 deaths associated with typhoid fever (TF) per
year [1]. TF is a systemic infection caused by the bacterium Salmonella enterica serovar Typhi, which is transmitted via the
oral–fecal route. As a food- and water-borne infection, Salmonella Typhi causes a considerable disease burden in low-income
countries that lack safe water and adequate sanitation and hygiene standards [2]. The majority of our knowledge regarding
TF epidemiology arises from high-burden locations in Asia
[3]. The burden of TF in sub-Saharan Africa (SSA) is largely unknown, but anecdotally is thought to be high, given that water
and sanitary conditions remain largely unimproved in SSA
many countries. In 2004, Crump and colleagues estimated a
moderate incidence of TF in SSA [4]; however, several more
Correspondence: A. Pach, International Vaccine Institute, c/o 105 New England Ave,
T6 Summit, NJ 07901 ([email protected]).
Clinical Infectious Diseases® 2016;62(S1):S69–75
© The Author 2016. Published by Oxford University Press for the Infectious Diseases Society
of America. All rights reserved. For permissions, e-mail [email protected].
DOI: 10.1093/cid/civ892
recent population-based studies have indicated a higher incidence of the disease than earlier thought. During longitudinal
surveillance, S. Typhi was isolated from 6.4% of blood cultures
in an urban Kenyan surveillance site, resulting in an adjusted
incidence rate of 2243 per 100 000 among children aged 2–4
years. Seventy-five percent of those S. Typhi isolates in that
study were multidrug resistant [5]. Furthermore, recent data
from Pemba Island in Zanzibar estimated an adjusted incidence
rate of 110 per 100 000 cases of TF among all age groups [6].
To address the gap in information on TF across SSA, the
Typhoid Fever Surveillance in Africa Program (TSAP), established by the International Vaccine Institute (IVI), founded a
laboratory infrastructure and surveillance system for TF and invasive nontyphoidal Salmonella infections in 10 African countries. This study was a component of the TSAP surveillance
project in Madagascar coordinated by the Microbiology Laboratory of Parasitology in the Faculty of Medicine at the University of Antananarivo, Madagascar. Disease surveillance at
sentinel healthcare facilities (HCFs) is a pragmatic way to assess
the rates and distribution of disease in any given population.
Healthcare utilization surveys support estimations of the
Healthcare Clinic Experience and Performance for Typhoid Fever
•
CID 2016:62 (Suppl 1)
•
S69
incidence of an illness by quantifying the use of an HCF by the
population living within a defined catchment area. A more accurate measure of disease incidence can then be estimated using
the number of cases presenting to the facility adjusted by the
catchment population. However, healthcare surveys do not always capture the complexity and variability of household
healthcare-seeking behavior in response to illnesses [7]. For example, a healthcare utilization survey for TF combining qualitative and quantitative data collection found that respondents in
the closed-ended survey claimed they never used traditional
forms of healing. However, in open-ended, qualitative interviews, respondents revealed the use of indigenous forms of
plant and household forms of healing for enteric illnesses [8].
In some cases, patients also used alternative healthcare resources, such as purchasing medications at pharmacies. Such alternative sources of care may reduce utilization of study HCFs and
affect estimates of the burden of disease [9].
The purpose of this nested TSAP study was to explore patient
perceptions of the standard of care they received at 2 sentinel
HCFs, their healthcare-seeking patterns, and their individual
reasons for use of these facilities. Providers also described the
nature of their medical consultations, the types of febrile illnesses they see, and the challenges they face in providing care for
patients at these public health clinics. These factors influence
delivery of care, patient access, and the health facilities’ capacity
to identify cases of febrile illness such as TF.
METHODS
Study Sites
Madagascar is an island country off the coast of East Africa. It has
a population of approximately 22–23 million people in 2015. The
majority of its population (ie, 70%) lives in rural settings and engages in agricultural activity [10]. This project was conducted in 2
HCFs in Madagascar: the Isotry Primary Healthcare Center CBS
II (ie, Centre de Santé de Base), a public health clinic in the city of
Antananarivo, and the Imeritsiatosika Primary Healthcare Center
CBS II, a rural public health clinic approximately 1 hour from Antananarivo. The Isotry clinic is located in an urban environment,
but draws a large proportion of its population from nearby rural
farmers who come to the city on market days to sell their crops.
Although economic disparities exist between rural and urban
populations in Madagascar, there is a high degree of similarity
in the socioeconomic and demographic characteristics between
the populations of the 2 HCFs.
Study Respondents
Study respondents were adults (≥18 years of age) who had received medical care for a febrile disease episode affecting themselves or their children. The attending doctor solicited patients’
participation in the interview at the end of their visit. Patients
who were not enrolled in the TSAP study were enrolled in this
substudy to avoid possible biases that may have been induced
S70
•
CID 2016:62 (Suppl 1)
•
Pach et al
through differential treatment provided to TSAP-enrolled patients, which may differ from the typical standard-of-care treatment provided at the clinics. In addition to febrile patients,
administrative heads of the HCFs, physicians, nurses, and
nurse midwives were also interviewed using a different data collection instrument. Thus, this study involved a purposive, facility-based sample of a life-experience group that had febrile
symptoms and attended the primary care study clinics. There
were also complementary key informant interviews and a
focus group with the medical personnel of these facilities.
Data Collection
The interviews and focus group discussions were interactive and
flexible, allowing for participants to accurately describe what they
understood and how they felt about specific issues. Discussions
were guided by a list of topics of interest. When English was
not the primary language of the respondents, a trained local interviewer asked the questions and a translator summarized the content of the discussion, which allowed for follow-up questions for
clarification or elaborating on an issue. The interviews and a focus
group discussion were audio-recorded, and recorded notes were
taken throughout the course of the discussion. A total of 33 patients were interviewed, with 16 and 17 patients interviewed at
the urban and rural health care clinics, respectively. Six physicians
were interviewed: 2 at the Imeritsiatosika HFC and 4 at the Isotry
HCF. One nurse participated in an interview at the Imeritsiatosika
HFC, and 5 nurses participated in a focus group at the Isotry HCF.
The sample sizes of the patients are considered to be adequate
for a focused ethnographic study of the experience of a segment
of a population (ie, clinic attendees with febrile symptoms), especially as we confirmed these findings in triangulating them
with the data on the healthcare providers’ perspectives [11–
13]. The patient sample sizes also demonstrated their adequacy
in achieving a redundancy and saturation of our analytical categories, which indicated a sufficiency of information and confirmation of the findings of our qualitative analysis [14].
Interview Instruments
Data collection questions for healthcare providers inquired
about several variables including patient intake at each of the
facilities, causes of fever commonly presented, tests of febrile
symptoms, how patients typically respond to fever in families,
patient beliefs, the frequency of self-treatment for fever, the
patient load and wait times at the hospital, working conditions
at the clinic, the patient referral and record system, and the cost
of medications and medical consultations.
Patient interviews involved questions on transportation and
time required to reach a facility, symptoms prompting visits to
the HCF, steps taken to manage symptoms such as self-treatment
or seeking advice from other healthcare providers, severity of
the symptoms, preferred and typical choice of care for a family
member with fever, length of wait to see a doctor, satisfaction
with the consultation, and cost of care provided.
Data Analysis
The interviews and focus group discussions were transcribed
and translated into English. A coding dictionary was created
based on the concepts and categories of the interview guides.
A team of 3 analysts reviewed 2 transcribed interviews each.
The analysts then coded segments of the relevant data in the interviews and compared the results of the coding for comparative
reliability and completeness of the codes. Differences in coding
were discussed, and some codes were revised and others added
where needed. The texts of the interviews and focus groups were
then downloaded in the Ethnograph version 6 qualitative software program and the data were coded, segmented, and analyzed for thematic content according to key topics.
with the medical director and 2 physicians as well as a focus
group with 6 nurses.
Seventeen interviews were conducted at the rural Imeritsiatosika clinic. Sixty-six percent of cases were children with a parent.
Sixty-four percent of cases were female with a median age of 1.5
years; 87% of interviewees were female, with an average age of 32
years. The average number of years of education completed was
4.5 years (range, no schooling to university degree completed).
The occupations of respondents included farmers, laborers, vendors, and teachers. At the Imeritsiatosika clinic, we conducted
key informant interviews with the medical director and another
physician, and with 1 nurse. We also conducted informal interviews with nurse assistants in the “health room” (triage center).
Catchment Area Size and Distances to the HCFs
Location of Facilities and Travel Time
The measurements of the size of the catchment areas and the
distances from the villages to the Isotry and Imeritsiatosika
HCFs were calculated using Google Earth Pro. The analysis
identified the coordinates of the HCFs and the boundaries of
the catchment areas based on geographic information. The
size of the catchment area of each HFC was then calculated
with the location of the HFCs established as the study centers.
The distance from each village to its respective HFC was measured as the distance of the geographic centroid of the villages to
each HCF study center. The average of the distances of the villages to the HCFs was calculated to get the mean or average of
the distances of the villages, along with the high and low ranges
of the distances for each catchment area.
Using GPS locations from clinic records, it was found that the
Isotry clinic catchment area consisted of 14 villages and city sectors and spanned a radius of 2.39 km2. Patients came to the Isotry clinic within a range 0.17–1.6 km, and an average of 0.78 km
away. The Imeritsiatosika clinic had a much wider catchment
area, consisting of 38 villages across 181.9 km2 with a range
of 0.54–8 km away, and an average distance of 4.9 km.
Among patients attending the Isotry clinic, walking (71%
[10/14 individuals]) was the predominant mode of travel, although some patients and parents attended the HCF by bus
and car. The average travel time was 23 minutes (range, 5 minutes to 1.5 hours). The main mode of travel to the Imeritsiatosika clinic was also walking (75% [12/16 individuals]), with
some traveling by car and bicycle. The average travel time was
32 minutes (range, 5 minutes to 2.5 hours). These times are
consistent with the differences in size of the respective catchment areas, although there is a preponderance of people walking to both clinics, largely because of their proximity to the
clinics and because of the expense of a taking a car or bus.
Ethical Approval
This protocol was reviewed by both the IVI Institutional Review
Board and by the Human Subjects Review Committee of the
Faculty of Medicine of the University of Antananarivo. All
respondents completed a written informed consent prior to
an interview. All information was confidential. All personal
identifying information was removed from the data files and
audio recordings were discarded after transcription.
RESULTS
Demography of the Participants
Sixteen interviews were conducted at the Isotry clinic in Antananarivo. Patients aged >18 years were interviewed directly,
whereas the parents of patients aged <18 years were interviewed
about their children’s illness. Fifty percent of the cases assessed
were children with a parent. Seventy-five percent of cases were
female, and the median age was 18.9 years; 91% of interviewees
were female with an average age of 30 years (range, 18–55
years). The mean educational attainment of interviewees was
8 years of school (range, 2 years of primary education to 2
years of university education). Occupations included housewives, secretaries, students, vendors, and laborers. At the Isotry
clinic, we also conducted key informant, individual interviews
Healthcare Procedures
When patients came to either HCF, they first went to the nurses’
station where a nurse recorded their temperature, blood pressure, and weight. Babies and children were measured to assess
potential malnutrition through measurements of their age,
height, weight, and the circumference of their arms. Patients
had to bring a small carnet, which is a small notebook to record
their vital signs and to share their medical history with the doctor. Following the initial triage, the nurse placed the patient in
the queue to see a doctor ,with a typical wait of 5–30 minutes for
a consultation. Doctors examined patients and recommended
blood tests or rapid tests for malaria, human immunodeficiency
virus, or tuberculosis and, more recently, TF, and referred patients with serious illnesses to a hospital. Before the TSAP project improved the clinics’ laboratory facilities to be able to
conduct blood tests for TF, physicians had to assess the disease
based on clinical symptoms, which was complicated by a range
of other common, febrile illnesses.
Healthcare Clinic Experience and Performance for Typhoid Fever
•
CID 2016:62 (Suppl 1)
•
S71
Healthcare Facility Experience
Experiences and perceptions of the care provided at both clinics
differed among patients, and identified reasons for the use or
nonuse of these HCFs by people in the catchment areas.
Longevity and Familiarity in the Use of the Facilities
Caregivers and patients from both Isotry and Imeritsiatosika
generally stated that patients typically have utilized these facilities for a long time and have sought care for a variety of health
issues. In speaking of the Imeritsiatosika clinic, one young, female patient said, “I have come here from the very beginning,
even to have my teeth taken out as a child.” Similarly, a mother
of a child at the Isotry clinic recalled that “my parents brought
me here as a child. I came here with my mother when I was
pregnant and I still come here when my child is sick.” Thus, familiarity and regularity in the use of these clinics were mentioned as issues important to respondents.
described her 12-year-old child’s symptoms: “He had a high
temperature” and when asked “what’s the matter?” he said, ‘I
have a sore throat, pain in the chest, and the flu.’ He also had
a temperature and when the temperature didn’t stop we came
here.” Patient visits at the Imeritsiatosika HCF also often involved compounded symptoms. Another mother with an 18month-old boy described her child’s symptoms: “Yesterday
after 7:30 PM after dinner he sneezed, coughed, and vomited
all the food he took. Then he couldn’t breathe so well. He had
a high temperature last night and with that symptom we came
here.” These examples suggest that the severity of symptoms experienced can have an impact on a patient’s decision to seek
healthcare.
Patterns of Healthcare Use
Respondents described variations in patterns of healthcare utilization related to the perceived severity of the symptoms, practices of self-care, and the use of other HCFs.
Features of the Quality of Care Received
Respondents also described the quality of the care they received.
One recurrent theme was that the doctors provide medicines
that have cured them and that the consultations and medical
tests increased their confidence in the effectiveness of the care
they received. At Isotry, a mother remarked, “My first child got
cured here and that’s why I come back.” As another example
of confidence in the care provided, a young mother at Isotry
mentioned, “According to us, we find it safer for children and
adults . . . doctors weigh, take the temperature and size of the
child, and that makes parents like coming here.”
Patients at the Imeritsiatosika clinic also appreciated the quality of the care provided. In addition, patients and parents said
that they experienced a caring and hospitable attitude on the
part of the physicians. A male patient at the Isotry clinic remarked, “Not only are the doctors hospitable, but they respect
the patients, too. This is a state health facility where they take
care of patients, with enough time for the consultation.” Moreover, patients and caregivers found that the queue was generally
not long, making it easy to see a doctor. At Imeritsiatosika, patients mentioned that it took from 2 to 30 minutes to see a doctor,
although it could be longer depending on the day. One mother’s
comment at the clinic captured this attitude: “It was quite easy to
see a doctor. We had to wait 15 minutes in the queue, but it was
okay . . . we had enough time to talk [to the doctor]. I am satisfied.” These comments from these patients suggest that time to
wait for a doctor is an important factor in their experience of
using HCFs.
Reasons for Use of Healthcare Facilities
Illness Help-Seeking
People utilized the Isotry and Imeritsiatosika public health facilities as their main source of healthcare especially when symptoms became painful, prolonged, or involved multiple
complaints. For instance, one mother at the Isotry clinic
S72
•
CID 2016:62 (Suppl 1)
•
Pach et al
Alternative Sources of Healthcare
A number of caregivers and patients claimed to come to the
health clinics immediately when there was a fever or other illness. Yet, many individuals also described initially utilizing natural home remedies (eg, lemon and hot water) or medications
from local pharmacies to decrease symptoms before bringing a
child to see a doctor. One mother of a 12-year-old boy described
her efforts: “The temperature did not go down after giving him
paracetamol, and so I brought him to the doctor.” A few individuals mentioned purchasing more potent medications at
pharmacies, including Efferalgen (codeine phosphate), Nivaquine (chloroquine sulfate), amoxicillin, and other antibiotics.
As one women patient at the Isotry clinic stated, “I bought
amoxi [amoxicillin] and paracetamol because I had a sore
throat. I took 2 amoxi in the morning and 2 at night, the
same with the paracetamol . . . and I got better.” Two other respondents who had used antimicrobials during febrile disease
episodes also mentioned that they did not need to come to
the health clinics after they took the medication.
Healthcare providers at both clinics corroborated the comments of patients. A physician at the Isotry clinic remarked
that although patients use herbs and other forms of traditional
healing, they most commonly purchase medications at local
pharmacies, including antibiotics, and “do automedication at
home and come here when not cured.” This physician pointed
out that this practice can lead to inadequate or incorrect use of
antibiotics and delays in proper treatment, which creates risks
for complications in cases of TF.
Respondents did not consider going to a hospital as the first
option for primary healthcare visits. A doctor at the Isotry clinic
stated that patients have to pay for medication and tests at the
hospital, and often have to leave a family member there for
days, which can be difficult emotionally and financially for families. As she said, “It is impossible for poor people to go to the
hospital because it is expensive for those in a difficult [financial]
situation; they have to pay for the medicines and for the tests.”
Another source of alternative care was the use of private clinics. Use of these clinics necessitated having adequate financial resources to afford these facilities, and often a desire for a more
extended consultation with a physician. One mother at the Imeritsiatosika clinic described a situation that led her to visit a private
clinic. She stated that “if you don’t have money you go here [Imeritsiatosika clinic], while you can go to the private doctor when
the money is enough. We are used to going here. But at the beginning of an illness one time we gave our child paracetamol and
nivaquine but he did not improve so we brought him to the private clinic.” Another parent described the private clinics as providing more thorough and immediate care. She related
experiences in which she determined the need to go to a private
clinic: “You see there are so many patients here and it takes a long
time, and you can’t wait any longer because your child is ill . . .
just the medications and consultations are expensive.”
Healthcare Cost and Barriers to Care
Some patients complained that they often only received the
same minor medications whenever they came to the public
health clinics; access to other medications was an incentive
for visiting a private clinic. One of the most important reasons
mentioned for coming to the public health clinics was the low
cost of the medicines they provide and that the consultations
were free. Cost was spontaneously mentioned as a reason for
utilizing the Isotry and Imeritsiatosika facilities by approximately one-third of the respondents (6/16 and 6/17 people, respectively). When asked why they chose this HCF, one mother of a
patient at Imeritsiatosika said, “In the first place it is a money
problem, so that the medicines here are quite affordable.”
This patient considered this clinic within her means. Another
mother with her ill child at the Imeritsiatosika clinic stated
that “it is near to where I live, easy to see a doctor here, and
the medication cost is cheaper.”
Cost of the medications and transportation were key factors
in determining patient access to the Imeritsiatosika clinic. Patients also considered the cost of the medications at Isotry clinic
to be affordable for them. As one female patient remarked, “I
like this center because the medicine is good, and cheaper compared to other centers, and the doctors take care of patients.”
The free consultations and inexpensive medications provided
at the clinics were critical for those attending them. A doctor
at the Isotry clinic observed that most people coming to this
health facility “don’t have any money.”
While the use of private clinics may draw individuals and caregivers away from the use of public health clinics, a lack of funds
can act as a barrier to accessing any health facility. One female
patient at the Isotry clinic described this situation: “If a child is
ill, but there is no money, you go nowhere, you suffer.” An
older female patient at the Isotry clinic described the impact of
socioeconomic circumstances for her family. “They should have
seen a doctor earlier this week, but couldn’t come because no
money.” A nurse confirmed that some families with no money
feel that it is impossible to come to a clinic and receive care despite perceived need. She observed that “patients don’t come here
because they are afraid they’ll have to pay for medications and
most people who come to the clinic don’t have money.”
The government of Madagascar has a social program to provide medicines for free for the very poor and elderly. If one’s
income is above the cutoff or fluctuates, they may not qualify
for the program. This is an incentive for the very poor to utilize
these health facilities. As a doctor at Isotry remarked, “there is a
social program for people if they are poor, medication is reserved for them . . . they don’t have to pay for it.”
Physicians at both clinics referred patients to hospitals with
complicated episodes of TF that did not respond to treatment. A
physician at the Isotry clinic described this situation, saying that
“when a fever continues for 7 to 8 days or the individual comes
back after treatment . . . I send them to the hospital.” However,
for some poor patients who are fearful of the costs at the hospital,
which they cannot afford, they avoid going to the hospital even in
cases of serious illness; as a physician at the Imeritsiatosika clinic
described, they try to “treat [the illness] at home.”
Overview of Healthcare-Seeking Behavior
The nurses and doctors at the Isotry and Imeritsiatosika health
clinics all agreed that there is a growing number of people coming to these clinics with severe symptoms after other remedies
had failed to cure them. The head doctor at the Isotry clinic remarked at length on this widespread problem: “Our real problem is that there are antibiotics in the stores, and now there is
ibuprofen which people do take. They only come here when
their children and family do not get better. So patients are
often exhausted and have temperatures which they cannot reduce, so that’s why they say that they have been ill for 2 or 3
days before coming here.” This behavior represents a serious
problem for surveillance at the clinics. Accounting for varying
patterns of healthcare utilization and barriers to the use of these
clinics is a challenge in conducting effective surveillance. Another physician at Imeritsiatosika also observed that if a child’s
fever is prolonged, serious, and not responding to treatment at
home, it can increase the potential for complications.
DISCUSSION
These accounts of healthcare-seeking behaviors of patients and
healthcare providers demonstrate factors that have implications
for conducting surveillance on febrile illness. Healthcare utilization surveys are necessary tools for adjusting estimations of the
incidence of disease via passive surveillance at clinics. However,
this study suggests that there are selective factors related to both
the clinics and patient behaviors that can facilitate or limit use
of these clinics by certain subpopulations. Yet, several features
Healthcare Clinic Experience and Performance for Typhoid Fever
•
CID 2016:62 (Suppl 1)
•
S73
of these clinics make them critical sites for the surveillance of
enteric fever in Madagascar. A number of patients said that
they had utilized these clinics throughout their life and continued to access them for their children, especially for prolonged or
serious symptoms such as those that characterize TF. The majority of respondents attributed the regular use of these HCFs to
the high-quality care that they received at these clinics. They appreciated the recording of vital signs at each visit, which added
to the information available to physicians, and were very satisfied with their consultations and the wait time to see a doctor. It
is important to note that a majority of patients walked to the
clinics from nearby communities, although this may limit patients from distant communities or those who are very sick
from attending these HCFs.
Both patients and healthcare providers observed that patients
generally sought care for themselves or their children when
fever was severe, prolonged, or complicated with other symptoms and often attempted to self-treat prior to making a visit
to the clinic. Consequently, patients with mild and perhaps
moderate cases of febrile illness may not present as regularly
to these clinics as do those with more serious conditions. Moreover, adult men accounted for a small proportion of fever cases
seen at the Isotry and Imeritsiatosika clinics. One nurse observed that men rarely come to the HCF unless they are very
ill. This might be because of the opening hours that overlap
with the workday. Differential health-seeking behaviors between sexes and age groups may bias incidence estimates
based on sentinel surveillance, as well as whether disease is differentially distributed among these groups.
We noted that the cost of treatment at these HCFs encouraged or discouraged attendance depending on the respondent’s
socioeconomic status. Some respondents who did not have government-subsidized healthcare registration due to being above
the income cutoff designation described instances when their
financial status prevented them from visiting the clinic despite
a perceived need for care, whereas those with lower incomes
were able to receive free treatment. Respondents with a higher
socioeconomic status also reported using private clinics if they
felt that the public health clinic could not meet their needs.
Therefore, the poor and wealthy may be diverted from the surveillance sites during episodes of febrile illness, though the causes and avenues of diversion differ. Research on healthcare
utilization in general, and for TF specifically, shows that satisfaction with healthcare services, cost, and distance to health facilities are key factors in determining their use [15–17].
A method to address loss of cases to other facilities in surveillance at sentinel sites has been to provide incentives to private
clinics to record enteric illnesses and provide these data to study
investigators [18]. Another approach may be to reach distant
and marginal populations and increase utilization of HCFs by
complementing attendance with a mobile phone component
to the passive surveillance, in which households are incentivized
S74
•
CID 2016:62 (Suppl 1)
•
Pach et al
to call a clinic in the case of fever. When indicated, a paramedic
could be dispatched to the home of the patient to take blood
samples for laboratory assessments and case identification,
thus augmenting passive surveillance [19–20].
There were both limitations and strengths to this study. This
study involved a small and purposive sample of patients, nurses,
and doctors that came to and were present at these HCFs during
a 6-week period, which limits overall study generalizability. In
addition, data collection involved qualitative interviews and
focus groups, which did not measure the frequency or distribution of responses in a representative sample of the whole population of the catchment areas. However, the study gathered
information from an adequate sample of respondents for an
in-depth appraisal of febrile patient and caregiver perceptions
of and experience with the Imeritsiatosika and Isotry clinics
that demonstrated general agreement on their experiences and
rationale for the use of these facilities among patient respondents, which was corroborated in interviews with healthcare
providers.
CONCLUSIONS
This qualitative study of 2 sentinel public health clinic surveillance sites in Madagascar provides an in-depth perspective on
healthcare-seeking behaviors and experiences related to important contextual and patient-centered experiences. This information is useful for augmenting health utilization survey research
and disease surveillance efforts. It was critical to account for the
use and experience of these key health facilities, as well as variation in their use, in relation to the perceived quality of the services, the rationale for the use of alternative sources of care, and
the geographic and socioeconomic factors that influenced responses to cases of febrile symptoms. This study demonstrates
the importance of these 2 clinics as sites for the surveillance of
enteric fever in their role as central healthcare sources for families and communities from this area of Madagascar.
Notes
Acknowledgments. The authors thank the patients and families of those
who participated in this study for their cooperation, which made this work
possible. We also express our gratitude to the staff at the Isotry and Imeritsiatosika clinics for providing important information and for their assistance
in organizing patients and caregivers for interviews.
Financial support. This research was funded by the Bill & Melinda
Gates Foundation (grant number OPPGH5231), and this publication was
made possible through a grant from the Bill & Melinda Gates Foundation
(OPP1129380).
Supplement sponsorship. This article appears as part of the supplement
“Typhoid Fever Surveillance in Africa Program (TSAP),” sponsored by the
International Vaccine Institute.
Potential conflicts of interest. All authors: No reported conflicts. All
authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content
of the manuscript have been disclosed.
References
1. Crump JA, Mintz ED. Global trends in typhoid and paratyphoid fever. Clin Infect
Dis 2010; 50:241–6.
2. Levine MM. Typhoid vaccines. In: Plotkin SA, Oresntein W, Offit PA, eds. Vaccines. 6th ed. Philadelphia, PA: Elsevier, 2013.
3. Ochiai RL, Acosta CJ, Danovaro-Holliday MC, et al. A study of typhoid fever in
five Asian countries: disease burden and implications for control. Bull World
Health Organ 2008; 86:260–8.
4. Crump JA, Luby SP, Mintz ED. The global burden of typhoid fever. Bull World
Health Organ 2004; 82:346–53.
5. Breiman RF, Cosmas L, Njuguna H, et al. Population-based incidence of typhoid
fever in an urban informal settlement and a rural area in Kenya: implications for
typhoid vaccine use in Africa. PLoS One 2012; 7:e29119.
6. Thriemer K, Ley B, Ame S, et al. The burden of invasive bacterial infections in
Pemba, Zanzibar. PLoS One 2012; 7:e30350.
7. Biehl J, Petryna A. When people come first: critical studies in global health. Princeton: Princeton University Press, 2013.
8. Kaljee LM, Pach A, Thriemer K, et al. Utilization and accessibility of healthcare on
Pemba Island, Tanzania: implications for health outcomes and disease surveillance
for typhoid fever. Am J Trop Med Hyg 2013; 88:144–52.
9. Reddy EA, Shaw AV, Crump JA. Community-acquired bloodstream infections
in Africa: a systematic review and meta-analysis. Lancet Infect Dis 2010;
10:417–32.
10. Chabanon A, Damin M. Etude épidémiologique des pathologies et traitements
rencontrés dans le service des Maladies Infectieuses de l’Hôpital Joseph Raseta.
Antananarivo, Madagascar: 5eme ane’e pharmacie, Lyon, ISPB, Stage hospitalouniversitaire a l’Hopital Befelatanana, 2010.
11. Trotter RT, Needle RH, Goosby E, Bates C, Singer M. A methodological model for
rapid assessment, response, and evaluation: the RARE program in public health.
Field Methods 2001; 13:137–59.
12. Pelto PJ, Pelto GH. Studying knowledge, culture, and behavior in applied medical
anthropology. Med Anthropol Q 1997; 11:147–63.
13. Johnson JC. Selecting ethnographic informants. Newbury Park, CA: Sage Publications, 1990.
14. Guest G, Bunce A, Johnson L. How many interviews are enough? An experiment
with data saturation and variability. Field Methods 2006; 18:59–82.
15. Memon ZA, Pach A, Rifkin M, et al. Health care preferences for children with typhoid fever in two slum communities in Karachi, Pakistan. Southeast Asian J Trop
Med Public Health 2008; 39:1110–25.
16. NoorAli R, Luby S, Rahbar MH. Does use of a government service depend on distance from the health facility? Health Policy Plan 1999; 14:191–7.
17. Nyamongo IK. Health care switching behaviour of malaria patients in a Kenyan
rural community. Soc Sci Med 2002; 54:377–86.
18. Khan MI, Sahito SM, Khan MJ, et al. Enhanced disease surveillance through private health care sector cooperation in Karachi, Pakistan: experience from a vaccine
trial. Bull World Health Organ 2006; 84:72–7.
19. Prue CS, Shannon KL, Khyang J, et al. Mobile phones improve case detection and
management of malaria in rural Bangladesh. Malar J 2013; 12:48.
20. Campbell TC, Hodanics CJ, Babin SM, et al. Developing open source, self-contained disease surveillance software applications for use in resource-limited settings. BMC Med Inform Decis Mak 2012; 12:99.
Healthcare Clinic Experience and Performance for Typhoid Fever
•
CID 2016:62 (Suppl 1)
•
S75