Health beliefs and prescription medication compliance among diagnosed hypertension clinic attenders in a rural South African hospital

This study examines the relationship between health beliefs and the use of both prescribed medication and alterna­ tive healing agents among at least one year diagnosed hypertensives attending an hypertension out-patient clinic in a rural South African hospital. The sample included 33 men and 67 women, in the age range of 31 to 81 years, (M=60.7 years, SD=9.8 years). Main outcome measures included causative beliefs, health beliefs, and quality of the health care provider patient interaction. From the 100 patients studied 35% were not compliant with prescription medication. Most patients (almost 80%) had taken something else for their high blood pressure apart from prescription medica­ tion, especially those who had been non-compliant with prescription medication. Most popular were the use of home remedies and faith healing, followed by traditional healing and over-the-counter drugs. Non-compliant behaviour was associated with the use of alternative healing agents, the belief of curability of hyper­ tension by traditional and faith healers, perceived benefits and barriers of antihypertensive medication and some items of the quality of the practitioner-patient relationship such as not explaining medical problems. Results are discussed in view of improving culturally sensitive compliance behaviour among hypertensive patients.


Introduction
The South African Demographic and Health Survey (DHS) (Department of Health, 1998: 32) showed that among per sons above 15 years 11 % of men and 13% of women were found to either have a blood pressure above 160/95mmHg or were taking appropriate medication to lower their blood pressure.A calculation based on these prevalence rates from the DHS and the census figures published for the South African population 15 years and older (Statistics South Africa, 1998: 5) leads to an estimate of about 3.3 million hypertensive people in the country.Overall, fewer hypertensive men (9%) than women (23%) are aware of their condition.Similarly, fewer hypertensive men in nonurban areas know that they suffer from the condition than their urban countreparts.This highlights that non-urban hypertensive males are the group with the most undiag nosed hypertension in the country and that they should be targeted to improve the diagnosis rate.This poor level of diagnoses in men is reflected in the low rates of men ( 1 1 % compared to 28% of women) who take appropriate drugs for hypertension.Consequently, only 9% of all men with hypertension had controlled blood pressure (BP< 160/95 mmHg), compared to 23% of hypertensive women.This is still a very low level o f control and highlights the need to improve hypertension control in the country if premature death and disability are to be prevented.A more disturbing finding is that the control of hypertension in young pa tients is far worse than that achieved in older hypertensive patients.These are the hypertensive patients who require good control even more than older patients to prevent end organ damage while they are still part of the labour force of the country.For men, the worst level of control was re ported in the African group, while for women it was found in non-urban African women (Department of Health, 1998: 32).Patient drug use behaviours and compliance have been the focus of much research over the years.WHO (1993: 15f.) states that 50% of newly diagnosed patients with hyper tension fail to make a referral appointment and as many as 50% of patients seeking treatment drop out of care within a year.Nyazema (1984: 552) found in Zimbabwe that hyper tensive or diabetic patients had not complied with followup appointments since over 60% lacked comprehension of their disease and the use of the medicine prescribed to them.Esunge (1991: 293) identified the following factors in rural Cameroon that appeared to affect hypertensive pa tient compliance: free medication, free hospital visits, free transportation, open discussion with medical staff, use of a common dialect, and politeness o f medical staff.Haynes, McKibbon, Kanani, Brouwers and Oliver (1998:3f.)reviewed that although adherence and treatment outcomes can be improved by certain, usually complex interventions, full benefits o f m edications cannot be realised at currently achievable levels of adherence.It is time that additional efforts be directed towards developing and testing innova tive approaches to assist patients to follow treatment pre scriptions.Morrel, Park, Kidder Morrel, and Martin (1997: 609) state that medication adherence is governed by both beliefs and cognitive factors.Therefore, measures of health behaviours, attitudes about health and medication taking, and cognitive function need to be recorded.Among Jamaican women with hypertension it was found that the variables -perceived susceptibility, perceived se verity and 'cue to action'-identified in the Health Belief Model need to be strengthened if patient compliance and adoption o f health promoting behaviours are to be realized (Grant, 1993: 159).Anthropological research in clinical con texts has shown that differences between patient/practi tioner models of health and illness can be the source of many problems in complying with treatment and that the patient's view o f his or her own illness is important in the choices of treatment and therefore compliance (e.g.Heurtin-Roberts & Reisin, 1992:787).Compliance studies generally have not investigated the role of alternative medicines in patients' treatment decisions.People's beliefs about and evaluations of health and medical care practices affect their choices among alternative actions.Since hypertension af fects a large proportion of the black South African popula tion, it is important to learn how health beliefs and percep tions affect the use o f both formal and informal treatment practices.A theoretical framework that has been used extensively in predicting preventive health behaviours is the Health Be lief Model (HBM) (Becker, 1979:52).The model proposes that individual beliefs about severity and susceptibility of a disease and its consequences are associated with engag ing in treatment action.Subjective assessments of disease threat are assumed to provide individuals with the motivat ing force to take action.Once an individual feels substan tially threatened by a disease and its sequelae, he or she m ust decide am ong alternate actions.According to the HBM, it is at this point that individuals perform a type of cost/benefit analysis such that alternatives are subjectively evaluated in terms of their benefits and costs (or barriers).This cost/benefit analysis then results in a preferred course of action or, in this research, a preferred health behaviour (Brown & Segal, 1996:904).The purpose o f this research was to exam ine the relation ships between the health beliefs variables (health belief model, causative beliefs), quality of the health care patient relationship and drug use behaviour, i.e. compliance with prescribed antihypertensive medication, use o f home rem edies, traditional and faith healing.

Sample and procedure
The sample included 33 men and 67 women, in the age range of 31 to 81 years, (M=60.7 years, SD=9.8 years).The mean years of formal education was 7.5 years (SD=4.8years).Most (89) came from the village and some (11) were from an urban area.Fifty-nine were married or lived with a partner and 41 were single/separated/divorced/widowed.Almost all (97) belonged to the ethic group o f Northern Sotho and three to Tsonga.The major religious denomination was Apostolic and Zion Christian Church (41), followed by Prot estant type and Roman Catholic Churches (36), traditional or African religion (15), and other (9).Fourty-two rated them selves as quite well-off, 40 as not very well-off, 17 as quite poor, and one indicated to be wealthy.Permission was obtained from the University of the North Ethics Committee and the Limpopo Provincial Health and Welfare Department.Patients diagnosed with hypertension for at least one year and consecutively attending an hyper tension out-patient clinic until a sample of 100 was reached were included in the study.Before consultation the hyper tensive patient was asked for formal consent to participate in the study.Thereafter an interview was conducted and a questionnaire was face-to-face administered with the pa tient by a trained research assistant.The questionnaires were translated and back translated by bilingual experts in the major language (Northern Sotho) used in the study.The schedule was field tested before the survey and modi fied by including traditional and faith healing of hyperten sion as potential barriers/costs and benefits.The ques tionnaire was pilot tested at time 1 on five hypertensive patients and after 4 weeks re-administered to the same pa tients.The responses from the first and second interview were comparable, and this is an indication of the reliability of the interview schedule.

Measures 1
Three items on previous health-seeking behaviour (answered by "yes" or "no").(Internal consistency expressed with the Cronbach alpha coefficient for this sample was .65) 2 Four items on the curability of health care agents (rated from l=agree to 3=disagree).(Cronbach al pha for this sample was .64) 3 Two questions assessing the patients1 perceptions o f high blood pressure with respect to its identity, or signs or symptoms used to tell when blood pres sure is elevated ("Do you think you can tell when your blood pressure is up?" "How can you tell") (Meyer, Leventhal & Gutmann, 1985: 120).4 The Problem Portrait Technique on causative be liefs of hypertension (see Table 3) (MacLachlan, 1997:84-9).5 Health Beliefs: Questions developed by Brown and Segal (1996:907ff.)for the four perception elements of the Health Belief Model (HBM) were measured on a 5-point Likert-type scale anchored by 1 "Strongly Agree" and 5 "Strongly Disagree11 with 1 represent ing the strongest belief on each dimension.The four components of the HBM are as follows: • Perceived severity of hypertension (4 items) was measured in terms o f the seriousness of hyperten sion, the fear of having hypertension and the limita tion that hypertension imposes on social activities.(Cronbach alpha for this sample was .72) (3 items).Perceived benefits of antihypertensive medication (Rx) represented the degree to which re spondents believed their medication to be effective in controlling high blood pressure and preventing adverse consequences.Perceived benefits of Rx were measured in terms of the medication's ability to control high blood pressure, to prevent strokes, heart attacks and kidney disease, and to ease one's mind about having high blood pressure.(Cronbach alpha for this sample was .82) Perceived costs of hypertensive medication (4 items).Perceived costs of Rx were measured in terms o f paying for the medication (financial), forgetting to take the medication, problem obtaining refills and experiencing side effects from the medication.
(Cronbach alpha for this sample was .68) Perceived benefits of home remedies (HR)/traditional healing (TH)/faith healing (FH) (2 items for each treatment mode respectively) were measured by beliefs about home rem edies'/traditional healing'/ faith healing' effectiveness in controlling high blood pressure and by their ability to keep the body and blood balanced.(Cronbach alpha for this sample was .73) Perceived costs o f home remedies (HR (/traditional healing (TH)/faith healing (FH) (3 items for each treatment mode respectively) were measured in terms of the lack of physicians' acceptance of using HR/ TH/FH, the lack of comfort with discussing HR/TH/ FH use with their physicians, and the lack of effi cacy of HR's/TH 's/FH 's compared to prescription medication.(Cronbach alpha for this sample was .67) A 20-item health carer satisfaction questionnaire (see Table 5) (Blumhagen, 1982:323) (rated from l=strongly agree to 6=strongly disagree).(Cronbach alpha for this sample was .69)Sociodemographic data (8 items) and self-reported medication compliance.

Data analysis
Data analysis included descriptive.Chi-square, student ttest and discriminant statistics using the Staistical Package for the Social Sciences for Windows (SPSS-PC, version 10).

Results
Fourty-seven patients indicated that they had been diag nosed with hypertension in 1998,46 in 1997, five in 1999 and 2 in 1996.The study was conducted from March to June 2000.Fourty-six rated their illness with very severe, 35 with not severe, and 18 with severe.
From the 1 (X) diagnosed hypertensive patients 65 were found to be compliant and 35 non-compliant with prescription medication at their hypertension clinic visit.Compliance was defined as self-reported intake o f medication for 30 times and non-compliance 29 times and less during the last month (Brown & Segal, 1996:908).tion, especially those who had been non-compliant with prescription medication.Most popular were the use of home remedies and faith healing, followed by traditional healing and over-the-counter drugs.
Table 2 indicates the perception of the patients about the curability of hypertension by different healing agents.
M odern doctors were clearly perceived to be superior to traditional and faith healing in "curing" or controlling hy pertension.Again, the non-com pliant patients believed more in faith and traditional healing than the compliant pa tients.
Generally, most patients (83) believed that their hyperten sion is incurable, seven felt it could be cured soon with medication and ten indicated that they would not know how long the sickness will last.The names used by patients for their hypertension prob lem were in descending order of frequency: madi a magolo (litt."too much blood") (67), high blood (63), high blood pressure (13), ngope (nose bleeding for a long time) ( 1 1 ), mokola (nose bleeding when in a hot area) (5), and hyper tension (3).M adi a magolo is when the person has too much blood wondering in the body causing dizziness and headache.High blood is when blood is too hot and goes up boiling to the head so that you sweat a lotand feel dizzy.
Ngope and mokola is when there is too much blood in the body and one bleeds through the nose.Almost all patients (96) believed that they can tell when their blood pressure is up.The signs for the high blood pressure were perceived to be in descending order of im portance: headache (46), dizziness (17), heart palpitations (11), sweating-and feeling hot (6) and others (16).About half (51) felt that they could notice changes of blood pres sure (51), which were interpreted as follows: feeling well (24), headache or hotness (7), sweating (5), blurred vision (4), and other like tense blood vessels (4).Both com pliant and non-compliant participants believed that hypertension was a serious disease, fear and social limitations from hypertension were seen as a little less than serious.On an open question, more than half (61) feared stroke most about hypertension, 30 death, and nine noth ing.The participants also felt much susceptible to hyper tension such as stroke, heart attack, kidney disease and blindness.The com pliant group believed significantly stronger than the non-complient group that prescription medication was ( 1) preventing strokes and others, (2) con trolling their high blood pressure, and (3) easing their mind.The non-compliant group perceived significantly higher barriers of prescription medication, especially regarding "forgetting" and "side effects" of the medication, than the compliant group.Generally, perceived benefits of prescrip tion medication was higher rated than perceived benefits of faith healing, home remedy and traditional healing, in that order.However, perceived barriers or costs of non western treatments (faith healing, home remedy and tradi tional healing, in that order) were generally perceived as lower than that of western medication.There were no sig nificant differences regarding perceived benefits and barri ers for any of the non-western treatments between compli ant and non-compliant patients.Discriminant analysis showed that the overall percentage of cases correctly classified was 77%.O f the 65 patients of the compliant group, 50 (76.9%)were predicted correctly, while 5 out of 35 (22.9%) non-compliant cases were identi fied well.Table 5 indicates the self-rated and retrospective assess-

Discussion
This study assessed health beliefs and compliance behav iour with prescription medication in at least one year diag nosed hypertensives attending an out-patient hyperten sion clinic in a rural hospital in the Limpopo Province of South Africa.From the 100 patients studied 35% were not compliant with prescription medication.Poor adherence to long term treatment, both lifestyle modifications and phar macological therapy, has been identified as the major rea son for inadequate control o f elevated blood pressure (Kitler, 1996:5).
In this study the indigenous names used for hypertension were used interchangeably.Sim ilar results were found am ong W hite and W est Indian hypertensive patients (Morgan & Watkins, 1988: 561).However, this was differ ent to what was found among African Americans who of ten discriminated between "high blood/high blood pres sure" and "hypertension*1.P atients considered "high blood" a "blood disease" wherein excessively "hot", "thick", or "rich" blood progressively "rises in the body", whereas "high-pertension" was considered as a "disease o f the nerves" (Heurtin-Roberts & Reisin, 1992:288).Medical authorities assert that hypertension is asym pto matic, but the respondents thought in this study other wise.W hen asked, "Do you think you can tell when your blood pressure is up ?" almost all (91 %) o f the participants identified symptoms for detecting elevated pressure.Meyer et al. (1985: 121 f.) found that 71 % of the newly treated group for hypertension, 92% o f the continuing treatment group, and 94% of the re-entry group identified symptoms for de tecting elevated pressure.Signs for high blood pressure were perceived to be: headache, dizziness, heart palpita tions, pain, sweating, breathlessness and others.Morgan and Watkins (1988: 568) found among W hite and West In dian hypertensive patients that sym ptom s perceived as indicating a rise in blood pressure were in descending or der o f importance: pains or sensations in head, weakness/ tiredness, eye problems, dizziness, and feeling hot.Non-compliant behaviour was associated with the use of taking something else for high blood pressure apart from prescription medication, the belief of curability o f hyper tension by traditional and faith healers, perceived benefits o f antihypertensive medication, perceived barriers of anti hypertensive medication, and some items of the quality of the practitioner-patient relationship such as medical expla nation was not given.Especially the non-compliant patients believed that their hypertension was curable as opposed to incurable in the compliant patients.Lack of chronicity limits the importance of following a prescribed treatment (Schoenberg, 1997:174).Only a few items on the quality of the practitioner-patient relationship (general satisfaction, understandable language, expressibility, not explaining medical problems) were associated with compliance behav iour.Esunge (1991: 292) found among rural hypertensive clinic attenders in Cameroon that open discussion with medical staff, use of a common dialect, and politeness of medical staff influenced patient compliance.Interestingly, however, was that causal beliefs, perceived severity and susceptibility of hypertension, and perceived benefits of alternative healing agents were not associated with compliance behaviour.
Similar to another study in South Africa there was a pre-  (Morgan & Watkins, 1988: 561).Lit tle was mentioned about supernatural causes of hyperten sion.However, Schoenberg (1997: 179) found among Afri can-American rural elders with hypertension that the most frequently mentioned means of controlling high blood pres sure included avoiding and eating certain foods, losing weight, praying, reading the Bible, taking home remedies, and exercising.D iscrim inant analysis show ed that regarding different parametres measured that the compliant behaviour was more predictable than the non-compliant behaviour.For exam ple, regarding health beliefs 76.9% of the compliant but only 22.9% of the non-compliant group were predicted cor rectly.This seems to indicate that non-compliant behav iour is complex and cannot be reduced or isolated to spe cific factors as compared to compliance behaviour.Perceived barriers to antihypertensive medication was sig nificantly higher among the non-compliant than the com pliant patients, in particular regarding forgetting, side ef fects and costs of antihypertensive m edication.Esunge (1991: 293) also found among rural hypertensive clinic attenders in Cameroon that free medication and transporta tion positively influenced patient compliance.In this study experiencing side effects were associated with non-compli ance with antihypertensive medication.It was found from the a n a ly sis o f p h y s ic ia n -p a tie n t in te ra c tio n s w ith hypertensives that only a few doctors informed the patient about possible side effects of the prescribed medication (Peltzer, Mekwa, Khoza, Lekhuleni, Alberts, & Sethosa, 2002: 171).Lisper, Isacson, Sjoden and Bingefors (1997: 147) found from Swedish hypertensive patients that they expressed a desire to receive information at the beginning of the pharmaceutical treatment, especially concerning pos sible side-effects.Study limitations included that this was a cross sectional study representing one point in time not reflecting possi ble changes in individual perceptions, beliefs and behav iours over time.Another important limitation of this study was that compliance with prescription medication was ob tained by self-reports.However, these self-reports were cross checked by also investigating the hospital files, since patients had to space appointments in between their clinic attendance with the hospital.Moreover, the results of selfreported medication use in other studies indicate that selfreports are good estimates of actual medication taking prac tices (Brown & Segal, 1996:914).Furthermore, the hospital sample studied here was not able to include those hyper tensive cases who had dropped out from treatment previ ously and have gone elsewhere for follow-up treatment, which means that some of the truely non-compliant cases could not be included.A further study is suggested to spe cifically investigate health beliefs among those hyperten sive cases who drop out from treatment completely.

Conclusion
As a conclusion, factors that improve compliance include culturally sensitive patient education and attitude about the disease undergoing treatment (Elliot, 1994: 271).More research is needed to improve our understanding of cul tural beliefs in managing chronic illnesses.This study dem onstrated in line with a study among African American and White American hypertensives (Brown & Segal, 1996:903) that alternative healing agents such as home remedies, faith healing and traditional healing play an important role in the management of hypertension in terms of taking alternative healing modes only or concurrently with antihypetensive medication.The study findings further support that the suggestion that patient belief systems should be consid ered when developing interventions and when monitoring patient outcomes.The implication is that health and medi cal care providers can help to improve compliance with prescribed regimens by helping patients incorporate the regimens into their daily routines or by changing the medi cation so that side effects are minimized.In addition, pro viders should actively and nonjudgementally elicit patient beliefs about the use of home remedies, faith or traditional healing so that they can be aware of the potential for con current use with prescriptive medication (ibid.).

Table 3
Table 4 indicates the Health Belief Model regarding hyper tension.
indicates the causative beliefs about hypertension.The three major causes for hypertension as perceived from the Problem Portrait Technique were mental stress, salty and fatty food, whereas the three least important causes were other (old age, menopause, contraception), supernatu ral and lack of physical exercise.Case vignette examples for some of the different causative

Table 5 : Means, standard deviations and t-test for the quality of practitioner-patient relationship of the compliant and non-compliant group (scored from 1 =strongly agree to 5=strongly disagree) Quality of practitioner-patient relationship: (a) My health care provider...
***p<.001, **p<.0l,*p<.05