Perceptions of inhibitors and facilitators for adhering to hypertension treatment among insured patients in rural Nigeria: a qualitative study

Background Universal health care coverage has been identified as a promising strategy for improving hypertension treatment and control rates in sub Saharan Africa (SSA). Yet, even when quality care is accessible, poor adherence can compromise treatment outcomes. To provide information for adherence support interventions, this study explored what low income patients who received hypertension care in the context of a community based health insurance program in Nigeria perceive as inhibitors and facilitators for adhering to pharmacotherapy and healthy behaviors. Methods We conducted a qualitative interview study with 40 insured hypertensive patients who had received hypertension care for > 1 year in a rural primary care hospital in Kwara state, Nigeria. Supported by MAXQDA software, interview transcripts were inductively coded. Codes were then grouped into concepts and thematic categories, leading to matrices for inhibitors and facilitators of treatment adherence. Results Important patient-identified facilitators of medication adherence included: affordability of care (through health insurance); trust in orthodox “western” medicines; trust in Doctor; dreaded dangers of hypertension; and use of prayer to support efficacy of pills. Inhibitors of medication adherence included: inconvenient clinic operating hours; long waiting times; under-dispensing of prescriptions; side-effects of pills; faith motivated changes of medication regimen; herbal supplementation/substitution of pills; and ignorance that regular use is needed. Local practices and norms were identified as important inhibitors to the uptake of healthier behaviors (e.g. use of salt for food preservation; negative cultural images associated with decreased body size and physical activity). Important factors facilitating such behaviors were the awareness that salt substitutes and products for composing healthier meals were cheaply available at local markets and that exercise could be integrated in people’s daily activities (e.g. farming, yam pounding, and household chores). Conclusions With a better understanding of patient perceived inhibitors and facilitators of adherence to hypertension treatment, this study provides information for patient education and health system level interventions that can be designed to improve compliance. Trial registration ISRCTN47894401. Electronic supplementary material The online version of this article (doi:10.1186/s12913-014-0624-z) contains supplementary material, which is available to authorized users.


Background
Cardiovascular disease (CVD) is a leading cause of death globally [1,2]. Hypertension is a major risk factor for CVD [3]. Once rare in SSA, hypertension and its related complications are increasingly common in the region [4]. Treatment with medication and behavioral changes (reduction in dietary salt intake, weight reduction, moderation of alcohol intake and increased physical activity) can greatly reduce blood pressure (BP) and the risk of CVD-related mortality among people with hypertension [5]. In SSA, CVD prevention is a recent development [6], and hypertension detection, treatment and control rates are generally low [7]. The absence of affordable community-based primary care services has been identified as a major obstacle to effective hypertension treatment in the region [8]. But several countries in SSA are now developing programs that provide such services [9,10]. Hypertension is a chronic condition that requires lifelong adherence to pharmacotherapy and healthy behaviors. For many patients, long-time adherence to hypertension treatment is a problem [11][12][13][14][15][16]. There is evidence that hypertension care is more effective if it includes educational interventions that address barriers to adherence [12,17,18]. Theoretical frameworks underlying various models for patient education propose that, in order to be effective, education should be patient-centered, and tailored to patients' (cultural) views about the condition and the treatment [17][18][19]. In high income countries, where access to affordable care is generally available, many (mostly qualitative) studies have investigated patients' perspectives on adherence to hypertension treatment in different populations [13,[20][21][22]. A recent systematic review of 53 of those studies indicated that common patient-related factors for non-adherence to antihypertensive medication across countries and ethnic groups include patients' beliefs that medication is unnecessary when symptoms of hypertension or stress disappear, a dislike of medications, fear of addiction, and the experience of side effects [22]. Data from such studies have provided content for educational interventions to improve adherence and health outcomes in patients with hypertension in the USA and Europe [23][24][25]. Non-adherence to hypertension treatment can have serious consequences for the patients, in terms of lost opportunities for health improvements, and for the health care systems, in terms of wasted financial and human resources. New programs that provide universal primary care coverage are emerging in SSA [26]. So far, few studies in the region have investigated perceptions of hypertension treatment among patients who are covered by health insurance [27]. Hence, it is important to fill this gap [28]. Nigeria is one of the countries in SSA where initiatives are being taken to introduce private or public community based health insurance plans [29]. The estimated prevalence rates of hypertension in Nigeria are 19.3% for rural areas, 36.6% for semi-urban areas, 42% for urban areas, and from 12.4%-to-34.8% for the country as a whole [29][30][31][32]. Studies of patients' perceptions of hypertension treatment have been carried out in several settings and communities in Nigeria [15,16,[33][34][35][36], but not among patients who were covered by health insurance. Understandably, many of these studies concluded that patients perceived financial constraints as the major reason for not adhering to treatment [16,37,38]. The present study was conducted in the context a subsidized community based health insurance program that covers the costs of primary and limited secondary health care and medications, for low income residents of rural communities in Kwara state, the 9th poorest among Nigeria's 36 states. A recent population-based study that was conducted in Kwara reported hypertension awareness, treatment and control rates of 8%, 5% and 3%, respectively [29]. As a first stage of a larger project [39], we conducted a qualitative study among hypertensive patients who participated in the community based health insurance program in Kwara state to explore their views on hypertension management. The specific research questions of the study were as follows: 1) what are patients' views on factors that may inhibit or facilitate adherence to prescribed medications?; and 2) what are patients' views on factors that may inhibit or facilitate adherence to behavioral recommendations?

Methods
We used a qualitative research design and individual interviews with open-ended questions to explore patients' perceptions on inhibitors and facilitators for adhering to hypertension treatment [40].

Setting and participants
For this study we recruited patients who had enrolled in the Hygeia Community Health Care Plan and were treated for hypertension at Ogo Oluwa Hospital, which is located in Bacita, a small rural town in Kwara state, Nigeria.
The Hygeia Community Health Care Plan, now called the Kwara State Community Health Program, is financed by an international development organizationthe Health Insurance Fund (HIF) [41,42]. This health care plan was launched in 2007 to provide subsidized community based health insurance for low-and middle income groups in western Nigeria (Lagos) and north-central Nigeria (Kwara state). The insurance package covers primary and limited secondary care, including hypertension treatment.
Ogo Oluwa Hospital is one of the first clinics in Kwara state that joined the Hygeia Community Health Care Plan. It is a large, busy private hospital that has offered primary and secondary care to people from the Bacita area for over 25 years. In the 1990s, Bacita was a busy commercial center that hosted the then thriving but later moribund stateowned Nigeria Sugar Company. The original population of Bacita is Nupe, but the Sugar Company attracted other Nigerians of various ethnic origins, particularly the Yoruba. Although the Sugar Company went out of business, a substantial majority of the economic migrants of the time settled permanently in the Bacita area. Therefore, Yoruba is now widely spoken alongside the native Nupe. Currently considered a poor community, the predominant livelihood activities of the inhabitants are farming, petty trading, artisanship, hunting and fishing.
Ogo Oluwa Hospital is owned and managed by an experienced General Practitioner, assisted at that time by about 2 other doctors, 10 nurses, 3 pharmacy staff, 4 laboratory staff and 4 records staff. At the start of this study, about 400 of the hospital's patients were treated for hypertension and attended clinics on a monthly or bi-monthly basis. Most of them had joined the Hygeia Community Health Care Plan. About two-thirds resided within the town at an average distance of 10 to 20 minutes travel time to the hospital; the remaining third were living in some distantly located surrounding villages.
The inclusion criteria for this study were as follows: having Hygeia Community Health Care Plan; diagnosed with hypertension (systolic BP ≥ 140 mmHg and/or diastolic BP ≥ 90 mmHg); in treatment for hypertension at Ogo Oluwa Hospital for ≥ 12 months; having been prescribed antihypertensive medication for ≥ 12 months; aged 18 years or over; and being prepared to give verbal (recorded) informed consent. Since pregnancy induced hypertension has a different course and etiology, pregnant or lactating hypertensive females were excluded from participation in this study. We also sought equal participation of men and women.
In qualitative interview studies, data or thematic saturation is a criterion for determining sample size [43,44]. This means that the number of respondents is sufficient if interviews with new respondents do not yield new themes. Fifteen interviews is generally enough to achieve saturation [43,44]. On the basis of these considerations and the inclusion criteria, we selected a purposeful sample of 40 patients from the case file records of Ogo Oluwa Hospital and invited them to the study. None of the invited patients declined. Patients who agreed to participate were reminded about the appointment some days prior to date of interview. At the beginning of each interview session, patients were asked to give audiorecorded informed consent and were assured of confidentiality. Travel expenses incurred for study visits outside usual clinic days were reimbursed.

Interview guide and data collection
To explore patients' perceptions of hypertension treatment we developed a semi-structured interview guide. The guide built on instruments that were used in previous qualitative interview studies of patient perceptions of hypertension [13,45] and it was adapted to the specific context and purpose of this study. The interview guide was edited in English and translated into Yoruba and Nupe by native speakers. Eight main topic areas of the interview guide that referred to hypertension treatment were: knowledge and personal views about hypertension; access to care; the role of health insurance; sources of information about hypertension; views on (prescribed) medications; views of behavioral adaptations; satisfaction with care; and general views on CVD prevention (Additional file 1). Information on the sociodemographic characteristics of the participants was collected through close-ended questions. Information on the participants' recent health related outcomes, (blood pressure, co-morbidities), was collected from the respondents' case file records. Data collection and analysis are continuous and simultaneous iterative processes in qualitative studies [46]. In this study, we held a first set of 30 interviews between July and October 2010 and a second set of 10 interviews between November and December 2010. The specific aim of the last set of interviews was to explore a number of emerging themes that were identified through the analysis of the first 30 interviews.
The researcher (AOO) conducted the interviews at Ogo Oluwa Hospital. He is not from the study area and he did not know the participants, prior to the interview. He made his profession, researcher with a medical background (MD), known to the participants.
Each of the first set of interviews lasted 120 minutes on average, and the second set lasted about 60 minutes each. All interviews were recorded, transcribed and translated into English by the researcher and an assistant.

Data analysis
The interview transcripts were analyzed, using steps and coding procedures of the method of qualitative data analysis that was developed by Strauss and colleagues (grounded theory) [47]. Briefly, this inductive method begins with the assignment of a series of open codes to the transcripts of individual interviews, which are then grouped into clusters (concepts) in order to make them more workable. From these concepts, broader categories and subcategories are generated, through a process of constant comparison or verification. The ultimate outcome is a data-based matrix of concepts and (sub-) categories, which serve as an input for developing theoretical or conceptual frameworks.
We imported all interview transcripts into MAXQDA and we used this qualitative data analysis software to support consistent data analysis, processing, ordering and comparison of the results at different stages [48]. Underneath we provide some more detail about the process of data analysis that was adopted in the present study.
First, in each of the interview transcripts, sections containing information on respondents' ideas on hypertension and hypertension treatment (medications and behavioral changes) were identified, and inductively coded. Similar codes were assigned to text fragments that reflected the same idea. These coding activities yielded a codebook that included all the codes that were created during the analysis of the individual interview transcripts. To conclude this stage we reviewed the code book to eliminate duplicates and to inspect if the remaining codes captured similar concepts so that they could be grouped together. The goal was to establish a smaller set of mutually exclusive conceptual codes that expressed the meaning of the underlying codes without loss of the original content. The identified "concepts", and underlying text fragments were then scrutinized and grouped under four major themes referring to our research questions; inhibitors and facilitators of adherence to medication and behavioral recommendations. Subsequently, thematic coding procedures were used to group the concepts under each of the four major themes into categories and sub-categories. Results of these multilevel coding procedures were summarized into matrices. (See Additional file 2 for an example).

Trustworthiness
The researcher (AOO) selected the sections of the interview transcripts that served as an input for the data analysis. To increase reliability and trustworthiness of the analysis, the following steps were taken: Results were compared and if differences occurred (for instance with respect to the most plausible naming of a code) they were resolved through discussion and by going back to the original data. (2) To establish the (short) list of conceptual codes, two members of the team (AOO and JAH) independently reviewed the code book and created relevant concepts. Differences of opinion with respect to the clustering of codes or the naming of concepts were resolved by discussion and by going back to the original data. (3) The categorization of concepts under the main themes (inhibitors and facilitators of adherence) was done by AOO and supervised by JAH, who made sure that every decision was plausible and could be justified. (4) The preliminary results of the data analysis were based on 30 interviews. Overlooking these results, AOO and JAH acknowledged that some additional interviews (10) were needed to further confirm or explore some of the themes and concepts that had emerged from this preliminary analysis, so that thematic saturation could be reached. Examples of such themes were related to the perceived role of religion on (non-) adherence to medication and patients' ideas about "local practices" that can inhibit or enhance healthy eating habits and physical exercise.

Ethics
This study is part of a larger project that aims to develop and evaluate cardiovascular health education program for insured patients with hypertension in rural Nigeria [39]. Ethical approval was obtained on 30th March, 2010 from University of Ilorin Teaching Hospital prior to the start of the study in July. Guidelines for quality assurance of qualitative research in health from the Academic Medical Center were used to ensure appropriate conduct of the study [49]. COREQ criteria for reporting qualitative interviews were used as a guideline for reporting this study [50].

Results
We interviewed 40 patients. Half of them had controlled BPsystolic BP < 140 mmHg and diastolic BP < 90 mmHg, and the other half uncontrolled BPsystolic BP ≥ 140 mmHg and/or diastolic BP ≥ 90 mmHg (Table 1). Most were older than 50 years. All were rural dwellers, 80% earned less than 5 USD per day, and 70% had none or only primary school education, almost all (92%) lived in poor housing conditions and most (75%) were farmers, artisans or traders. All were religious, 62.5% practiced Christianity and 37.5% practiced Islam.

Factors inhibiting and facilitating medication adherence
Some patients said they did not always take their medications according to prescription. Frequently mentioned patterns of "self-regulation" included reducing the daily dosage, skipping medications under specific circumstances and discontinuing medications for several days or weeks. None had stopped using medications all together. Patients' accounts of factors that inhibited them from using their pills as prescribed could be classified into five main categories (    *n refers to the number of respondents whose perceptions contribute to the corresponding concepts.
for success for the operators of the program; the program manages my hypertension well and makes me healthier and stronger to work for my 'daily bread'.
The easy approachability of the doctor was also perceived as facilitating factor for medication adherence. A female participant with uncontrolled hypertension [ID3] respondedlast January when I was about to travel, I approached my Doctor and told him that I will be away for 2 months requesting enough drugs stock to last the period; this he willingly did, so I always take my drugs whether traveling or not. Another, a 61 year old man with controlled hypertension [ID22] respondedduring the last fasting season, I approached my doctor 2 weeks earlier to seek advice on how to use my drugs, he decided to withdraw my 'thrice a day' pill and replace this with a 'twice a day' pill, and later reverted back to the 'thrice a day' drug after the month-long fasting.

Factors inhibiting and facilitating adherence to behavioral recommendations
Managing hypertension requires a healthy behavior, including limited use of salt, weight reduction or maintenance, *n refers to the number of respondents whose perceptions contribute to the corresponding concepts.
exercise, cessation of smoking and limited use of alcohol and other stimulants. Table 4 illustrates factors which inhibit achieving such healthy behaviors, in the eyes of our respondents.
1) Salt: while most participants said they had been informed by their doctor that they need to be prudent with the use of salt, our data also suggest that some patients were aware of the impact of salt on their BP. Local food practices may inhibit salt reduction; in some communities, salt is used to preserve food, particularly meat and fish against microbial decomposition over time. When asked what local customs encourage salt use, a 70 year old man with controlled hypertension [ID19] replied in some parts of the north where I come from, we use salt to preserve food and meat. Food preparation practices can also inhibit compliance with advice to *n refers to the number of respondents whose perceptions contribute to the corresponding concepts.
reduce salt. In the study region, adding salt and salt substitutes/seasoning agents like maggi (salt plus hydrogenated oil plus monosodium glutamate -MSG), ajinomoto (an MSG product) etc. has become standard practice while cooking. The consumption of canned (processed) foods, often salty is also increasingly common in the region. The previously fairly common cultural habit of drinking salted water in attempts to treat 'undiagnosed' abdominal discomfort of all sorts was practiced by some respondents. A 50 year old woman with uncontrolled hypertension [ID6] statedit is more or less customary for people to cook food with salt in this region,… and for some people who may sometimes have stomach upset, tradition demands that one should pour salt into a cup of water and drink to treat the ailment; I used to do this too but being better aware, have now stopped the practice. 2) Weight control: participants mentioned a number of local or cultural practices that may prevent people from losing weight including the cultural perceptions; that large body sizes are associated with wealth, comfort or beauty; that weight loss is associated with disease; and that weight (fatness) is an unchangeable family trait. A 50 year old woman with uncontrolled hypertension [ID3] respondedpeople often see being fat as synonymous with affluence, comfort and peace of mind; they also sometimes associate fatness in a woman with beauty. Some of us women prefer to be fat or robust believing that by so doing; we will be more attractive to our husbands and men generally. Some men too prefer to be plump so they will be liked by women. A 63 year old man with controlled hypertension [ID30] saidin this region, people often link weight loss to disease particularly if the slimming down is getting too much, suggesting that such might be due to disease; at other times they suggest that slim people are very miserly and would rather not spend money to eat well and get robust.
Highlighting the perceived link with inheritance, a 55 year old woman with uncontrolled hypertension [ID16] respondedmy friends and neighbors often say I was too fat, but I'm not bothered about such comments as I believe I inherited the trait from my mother; we are usually fat in my family. Local food practices may also inhibit weight control. Some participants highlighted the difficulty in avoiding some of the main "fattening" ingredients of local popular dishes such as cassava, groundnut oil, (red) palm oil and meat. The first three ingredients are commonly grown by farmers in the community while meat is widely available due to popular goat breeding practice in the community whereby owning and breeding goats is a pastime that many families engage in.
3) Exercise: perceived inhibitors of exercising include 'local practices' and lack of information. Prevailing local or cultural views that exercise is "needless or useless' , dangerous, or incompatible with advancing age, may prevent people from being physically active. Exercise is sometimes regarded by people as an activity for unserious fellows, the unengaged or the miserly that rather walk habitually than pay for transportation costs to destinations. An elderly male patient with controlled hypertension [ID30] respondedwhen people see me 'walking to exercise' they often slight me and make derogatory comments that one is greedy and would rather walk long distances than spend money on transportation. Asked further how people see exercising or sporting activities, he repliedsome people see those who exercise as unserious people. Sports like playing football are only engaged by youths not elderly people like me.
Several respondents said they were not aware or informed about positive effects of exercise on BP or easy ways to exercise. Insufficient knowledge makes some patients view exercise as dangerous for their hypertension. Some others are simply unfamiliar with practicable exercises.

4) Tobacco and stimulants: besides cigarettes and
alcohol, other locally accessible stimulants like snuff (fine-ground tobacco) and kola (caffeine-containing) nuts are also used by some people in the region. Some patients perceive positive effects from using these substances, as illustrated in the following expressed views: cigarette is useful in relieving stress/tension; palm wine (local alcoholic beverage) and Ogogoro (distilled palm wine) make vision clearer; snuff stimulates work; and eating kola nuts is ideal norm for elders. A 65 year old man with controlled hypertension [ID23] respondedour fore fathers used to say palm wine makes one to see very well, but it is glaring to me now that such thing is not good. Another elderly (70 year old) man with uncontrolled hypertension [ID29] saidin this environment, we believe kola nuts and snuff aid work, though I personally don't use any of them. A 60 year old woman with controlled hypertension [ID33] said -I still eat Kola nuts although I have reduced the quantity I consume since I became hypertensive. Such views make it difficult to quit using these substances. Another 64 year old man with controlled hypertension [ID25] highlighted the social consequences of giving up smoking and drinking thusthe people and friends I used to drink and smoke with had to separate from me gradually after I stopped these habits, but this is fine with me as my health is more important.  Social support factors Gender based support (societal view) African society frowns at the habit of women smoking or using alcohol [n = 2] *n refers to the number of respondents whose perceptions contribute to the corresponding concepts.
with salt before now. The availability of substitutes for salt in local markets was similarly viewed as a facilitator. Finally 'social support' particularly the family's willingness to eat low salt meals was identified as an important motivator for reducing salt. As one respondent, a 50 year old woman with controlled hypertension [ID11] puts ityes, I was able to make use of the advice; we now reduce the quantity of salt added to common 'family pot' food generally; thereafter those in the family that desire more salt in their food can add extra salt to their portion after dishing.

Discussion
In this study we explored perspectives on treatment adherence among patients who received hypertension care in the context of a community based health insurance program in rural Nigeria. Results suggest that, having free access to previously unaffordable high quality care, from a healthcare center headed by a physician who was highly respected in the community, and to Western medication, were perceived as important facilitators for being adherent to medications, as was the perception that hypertension is a serious condition. Similarly, the performance of prayers, support from family members and other patients with hypertension and counseling from religious institutions were seen as factors reinforcing adequate medication use.
Interestingly, compared to results from similar qualitative studies on hypertension in Western countries [22] the patients in this study regarded modern "white man's" medications more positively and seemed to be less likely to alter the prescription according to their own insights. This finding seems credible in the light of a number of other Nigerian studies. In a qualitative study of explanatory models of hypertension among patients in a Lagos hospital, it was found that patients lacked "the desire to be active participants in their health care decisions" while they believed that "the doctor knew what was best for their care" [33]. Another study from urban Western Nigeria also found that patients were motivated to comply with treatment because they had access to Western prescription medications [36]. The Nigerian antihypertensive adherence trial demonstrated that medication adherence was very high among those participants who did not drop out of the study. The authors concluded that this is in line with the general experience that patients with "no or little prior contact with organized modern medical care will readily adhere to recommended therapy" [38]. At the same time, however, certain characteristics of the way care was organized could still hinder medication adherence, the most important being travel costs associated with clinic visits for those patients living far from the clinic, clinic operating hours, waiting times and the under dispensing of prescribed medications. In addition, patients' idea that hypertension is a temporary condition for which medications may no longer be required once symptoms disappear, was an inhibiting factor for taking medications regularly, or a reason for omitting clinic visits. While the practice of faith healing or the use of herbal drugs might inhibit the use of prescription drugs, in their own case, participants said they would use these remedies only to supplement antihypertensive medications and not as a substitute. Interestingly, lack of social support from the patient's immediate environment was rarely mentioned as an inhibitor for using medications.
These findings are plausible in the light of other studies that have also identified travel costs, long waiting times, clinic operating hours and limited knowledge by patients as factors limiting medication use [16,33].
While most respondents said they were aware that healthy behaviors could have a positive influence on their health and BP, adopting a healthy lifestyle seemed to be a greater challenge to them than taking their medications. Local practices and norms held by people in the community were identified as important inhibiting factors for behavioral changes including: use of salt for food preservation; negative cultural images associated with decreased body size and physical activity; and perceived (social) benefits of using palm wine, kola nuts, tobacco and snuff. Anthropologists have noted that a resistance to change (traditional) local food and eating practices and body images is common among people throughout the world, as these are often seen as an important component of an individual's cultural identity [51]. Yet, our study offers also some insight into the factors that may facilitate healthier behaviors, and tackle the inhibitors. These include sufficient knowledge or understanding of how healthy behaviors may affect hypertension, and support from family or religion, but most importantly the awareness that healthier behaviors will not require profound changes of one's usual daily life.
By using open interviews, the strength of this study is that its builds on direct experiences of the patients. We interviewed a sufficient number of patients to achieve thematic saturation, which is a criterion for sample size and internal validity in qualitative studies [43]. The iterative approach taken in this study [44] and specifically the decision to collect additional data on the basis of themes that were unclear from the preliminary analysis of data from the first set of interviews, added value to the results. We learned from these additional interviews that our informants attached significant importance to a traditional rural African way of life. And, when probed, they were readily able to describe how this way of life, not only provides inhibitors, but also opportunities for adhering to the behavioral recommendations they were given by their doctor. For example, the interviews provided much detail on what affordable food products are available in local markets to compose healthy meals (certain types of fruits, salt substitutes, vegetables and fish), and how people's normal daily activities (e.g. farming, household chores, yam pounding, walking, drawing water from the well) offer ample opportunities for getting exercise. The additional interviews also provided further insight into reasons for medication adherence. It became clear that some patients felt particularly motivated or obliged to adhere to their medications because they perceived their health insurance and, consequently, their access to good treatment as blessings from God. But at the same time, religious practices (e.g., fasting) could sometimes pose practical barriers to medication adherence.
In line with quality criteria of qualitative research [50], we undertook several efforts to reduce individual biases in the analysis and to strengthen the trustworthiness and credibility of the findings: e.g., by using MAXQDA software to ensure consistency of analytic procedures, and by involving two or more members of the research team at crucial stages of the analysis.
The study also has limitations. The participants consisted of insured hypertensive rural dwelling low income Nigerians reachable through telephone, who had continued to visit the clinic for treatment for at least one year. Consequently, our findings do not capture ideas from other groups of patients. For example, within the context of a community based health insurance program, it would also be relevant to further investigate views on treatment adherence among patients who dropped out of treatment, despite the fact that they had access to good care. In terms of data collection and analysis, our possibility to verify what patients told us against other sources (triangulation) was limited: ethical considerations prevented us from interviewing significant others to crosscheck information from participants. Similarly, since the interviews were conducted in local languages and transcribed in English, it is not impossible that some deeper meanings participants' intended to convey might be lost in the process of translation.
Poor adherence to treatment has been identified as one of the most important modifiable barriers to BP control in patients with hypertension [14,19,52]. Overlooking the accounts of the patients who participated in this study, it becomes clear that adherence can be affected by multiple spheres of influence including public policy (e.g., health insurance), institutional and organizational factors (e.g., health care system), environmental factors (e.g., availability of healthy foods), social and cultural norms and practices (e.g., food conservation and preparation practices), social networks (e.g., interactions with other patients, family members) and intrapersonal factors (e.g., patients' knowledge and awareness level).
This provides support for social ecological approaches to health promotion [53]. These approaches assume that single interventions, such as the provision of free access to care are generally not sufficient to promote healthier behaviors and that a mixture of interventions aimed at different levels is needed to accomplish this.
The framework of inhibitors and facilitators to adherence that emerged from this study (see Figure 1) suggests that, in the case of the community based health insurance program in Kwara state, a number of additional interventions may be needed to remove barriers that patients face in adhering to their treatment.
First, more attention could be paid to patient education. The findings suggest that patients with difficulty in managing their medications may benefit from further counseling or education about the nature of hypertension, why medications need to be taken regularly, what they may discuss with their doctor to make regular medication use easier (e.g., side effects, dosing, alternative medications, frequency of clinic visits), how they may deal with specific social, cultural or religious impediments encountered in taking their medications, and with the logistic challenges faced in getting refills on time.
In addition, patients who experience challenges in managing behavioral changes may benefit from further counseling or education about the positive influence of salt reduction, weight control, healthy diet and exercise on BP and health, and on how existing dietary and physical activity-related local practices may help them in implementing behavioral changes.
Community education may be one possible solution to avoid limited human resources and financial problems which patient education can pose to the already charged health care clinics. This type of education has been effectively used in the USA [54]. It can be delivered to groups of patients at the health care center or in other community settings the patients are familiar with, such as schools, churches or mosques. The education must be delivered by people who are known, trusted, culturally competent, and fluent in the language of the target community. These educators need not be health professionals, but they do need