Herbal and Alternative Medicine Use in Tanzanian Adults Admitted with Hypertension-Related Diseases: A Mixed-Methods Study

Background Hypertension is increasingly common in sub-Saharan Africa where traditional medicine use is also common. We conducted a hospital-based, mixed-methods study to determine prevalence, pattern, and correlates of herbal and alternative medicine use in Tanzanian adults hospitalized with hypertension. Methods A standardized questionnaire was administered. In-depth interviews were performed on a subset of participants. Factors associated with herbal medicine use were determined by logistic regression. The association between traditional medicine uses and allopathic medication adherence was determined using ordinal logistic regression. Qualitative data were analyzed according to grounded theory. Results Of 213 adults enrolled, 52 (24.4%) reported using herbs during the previous month and 47 (22.1%) reported concurrent use of herbs and allopathic medicines. Lower educational level, nonprofessional employment, and lack of health insurance were significantly associated with herbal medicine use. Alternative medicines use was not associated with lower medication adherence. Qualitative interviews identified several important themes including reasons for herbal medicine use. Conclusion The use of traditional medicines is very common among patients with hypertension. Adults from low socioeconomic status, those with misunderstandings about hypertension, and those without health insurance were more likely to take herbs. Open, nonjudgmental communication between healthcare workers and patients regarding use of traditional medicines must be encouraged in Africa.


Background
The burden of hypertension is increasing in sub-Saharan Africa (SSA) [1], with an age-standardized prevalence of ∼40% in adults over 25 years of age [2]. Of adults with hypertension in Tanzania, only ∼35% are aware of their diagnosis, ∼5% are on treatment, and ∼1% are controlled [3]. At our own hospital in Tanzania, hypertension is the second 2 International Journal of Hypertension in a more specific term include herbs, herbal materials, herbal preparations, and finished herbal products that contain active ingredients parts of plants. In the current study we used terms "herbal" to describe traditional medicine that is derived from plants and "alternative medicine" to describe other traditional medicines.
Traditional medicine use in hypertension and its related diseases can lead to adverse outcome, related to delay in seeking proper treatment, leading to increased severity or complications of the disease, drug interactions, and poor adherence to allopathic medicines that are known to control the disease [5]. Although traditional medicine use is common in sub-Saharan Africa (SSA) [5][6][7][8][9], little is known about its overlap with hypertension. In our own, recent metaanalysis, we found only 4 published studies reporting data regarding herbal and alternative medicine use in adults with hypertension in Africa. These 4 studies indicated that 25-65% of adults with hypertension reported use of herbal and alternative medicines [10].
To the best of our knowledge, no published data exists regarding the pattern of herbal and alternative medicine use among adult inpatients or the association between herbal and alternative medicine use and antihypertensive medications adherence. Specifically, it determined (1) the prevalence and pattern of herbal and alternative medicine utilization, (2) factors associated with herbal medicine use, and (3) whether herbal or alternative medicine use was associated with medication nonadherence. We hypothesized that herbal and other alternative medicine use is common among Tanzanians hospitalized with hypertension-related diseases and is associated with nonadherence to allopathic medicines.

Study Setting.
This study was conducted in the adult medical wards of Bugando Medical Centre (BMC). BMC is a tertiary and teaching hospital that serves the Lake region in the northwestern part of mainland Tanzania.

Study Design and Recruitment of Participants.
To provide better understanding of the problem of herbal and other alternative medicines use, the current study was conducted prospectively by a hospital-based mixed quantitative and qualitative research design and was conducted concurrently between April and October 2014. The subjects for this study were selected from patients hospitalized at the medical wards of BMC. All adults (>18 years of age) admitted with hypertension-related diagnoses were approached within 24 hours of admission and invited to participate in our study.
Blood pressure measurement was done at least twice (once on each arm) after 5 minutes of sitting using a mercury sphygmomanometer. If either or both first 2 blood pressure measurements are elevated, a 3rd blood pressure measurement was taken and the average of the 2nd and 3rd measurements was considered for the diagnosis of hypertension.
Diagnoses of hypertensive diseases, that is, hypertensive stroke, hypertensive heart failure, or hypertensive renal failure, were made per WHO International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) version for 2010. Adults with altered mental status were excluded since they could not provide adequate or reliable information.

Data Collection and Analysis.
A structured researcheradministered questionnaire with both closed ended and open-ended questions was used to collect quantitative data from eligible patients. The questionnaire consisted of sections that assessed patient information such as age, gender, marital status, education level, and duration of disease. Responses on patients' knowledge about hypertension, use of herbal and other alternative medicines, reasons for herbs use, sources of herbs, and adherence to antihypertensive medications were obtained.
Adherence to antihypertensive medications was measured using a Morisky Medication Adherence Scale, the 4item scale (MMAS-4). MMAS-4 is a self-reported commonly used adherence-screening tool. It is composed of 4 yes/no questions about past medication use patterns and is validated in antihypertensive patients in several countries [11,12].
Data were coded and entered into Microsoft Excel 2010 software (Microsoft Corporation, Redmond, WA, USA) and analyzed using STATA version 12 (StataCorp, Collage Station, Texas). Categorical variables were described as proportions (percentiles) and continuous variables were described as medians [interquartile ranges]. Differences between patient groups were calculated along with 95% confidence intervals (CI). To examine potential factors associated with the current use of herbal and alternative medicines, univariate and multivariate logistic regression were conducted. Factors associated with adherence to allopathic antihypertensive medications were explored using ordinal logistic regression. The Brant test was used to assess the proportional odds assumption. A twosided p value < 0.05 was considered significant. Since this was an exploratory analysis, we did not make any adjustment for multiple comparisons.
12 study subjects who used herbal or other alternative medicines were selectively chosen and underwent in-person, semi-structured interviews. The primary investigator conducted these interviews in a private space on the medical wards and they were ∼20 minutes long. Interviews were conducted in Kiswahili, recorded, transcribed, and then translated into English by a professional translation service.
The analysis process had two levels. We started collecting qualitative data through open-ended questions. As we reviewed the data collected, repeated ideas, concepts, or elements become apparent and were eventually tagged with inductive and deductive codes, which were extracted from the data. The second level involved collection of data, and as data were rereviewed, codes were grouped into themes and then into categories. These categories were the basis for new theory. Two independent researchers (AL and RP) identified themes and supporting quotations were selected.
In-depth interviews were performed on 12 adults admitted with hypertension. After breaking data into conceptual components, these concepts were theorized, integrated, and refined, and four themes/theories emerged: (i) lack of understanding about hypertension, (ii) the exact identity of herbal remedies being often unknown, (iii) contradictory opinions on risks of concurrent use of allopathic and herbal medications, and (iv) 3 common reasons for herbal medicine use.

Ethical Approval.
Ethical approval for the study was obtained from the Bugando Research Ethics Committee in Tanzania and the Institutional Review Board of Weill Cornell Medical College. Each participating patient provided written informed consent. Consent to participate in this study did not affect the clinical care of the patients.

Enrollment.
Between April 1 and October 1, 377 adults were hospitalized to the medical wards of Bugando Medical Centre with hypertension-related diseases. Of these, 164 were excluded for the following reasons: 147 were unable to give a reliable history due to altered mental status (mostly due to stroke or hypertensive encephalopathy), 9 declined to participate in the study, and 8 were discharged within 24 hours (before enrollment could occur). The remaining 213 adults consented and were enrolled.

Outcomes.
Of the 213 Tanzanian adults in patients with hypertension that participated in the study, 161/213 (75.6%) reported using herbal and alternative medicines at some point in their life. 69/213 (32.4%) used herbal and alternative medicines specifically to manage hypertension at some point in their life and 52/213 (24.4%) used herbal and alternative medicines in the past month before admission to the hospital. 38/213 (17.8%) of the respondents admitted to having stopped allopathic medications for herbal and alternative medicines and 47/213 (22.1%) reported to the concurrent use of herbal and alternative medicines and allopathic medications. Only 20/213 (9.4%) reported having been asked about traditional medicine use by their admitting doctor. Herbal and other alternative medicines use did not differ significantly by gender. The other outcomes related to the utilization of herbal and alternative medicine are presented in detail in Table 2. Table 3 displays the results of our logistic regression analysis to determine factors associated with herbal medicine use for hypertension in the past month before admission. Patients with higher categories of education (secondary education to college education) were less likely to consume herbal medicine compared to their counterparts ( < 0.007; OR = 0.49; CI 0.29-0.82). Professionals (salaried workers), businessmen (business owners), and government employees were also less likely to consume herbal medicine compared to farmers ( = 0.002; OR = 0.27; CI 0.11-0.61). Study subjects with medical insurance were also less likely to use herbal medicines ( = 0.001; OR = 0.24; CI 0.10-0.57). Stroke was the only comorbid condition that was significantly associated with herbal medicine use ( = 0.008; OR = 3.60; CI 1.40-9.21). Table 4 displays our ordinal logistic regression analysis of factors associated with poor medication adherence. Higher education level (secondary school and above) was associated with better medication adherence ( = 0.04; OR = 0.63; CI 0.40-0.99 for lower adherence). Study subjects who believed that hypertension could not be cured with a short course of traditional or allopathic medicines also had better medication adherence ( = 0.001; OR = 0.34; CI 0.18-0.63 for low adherence). In addition, study subjects who were diagnosed with hypertension in places other than a health facility were less likely to be adherent to the antihypertensive medications ( = 0.02; OR = 3.80; CI 1.21-11.85). Of note, neither herbal and alternative medicine use nor attendance to traditional healers was significantly associated with nonadherence. In fact, none of the variables for herbal and alternative medicine utilization were significantly associated with nonadherence.

Reasons for Herbal Medicine Use.
The 69 study subjects who had used herbal medicines specifically for the treatment of hypertension were asked to provide reasons for herbal medicine use: 22/69 (31.9%) reported using herbal medicines because they still believed in traditional ways of treatment and customs of healthcare; 12/69 (17.4%) were convinced by friends or people who have claimed to have been cured by herbs; 11/69 (15.9%) were looking for other options for treatment; 7/69 (10.2%) reported pressure from a spouse or family member; 5/69 (7.3%) reported that they did not want to go to hospital because of uncaring attitude of hospital staff; 3/69 (4.4%) patients used herbal medicines because they did not have funds to go to hospital; 3/69 (4.4%) patients wanted

Discussion
Herbal and other alternative medicines use is highly prevalent among Tanzanian adults admitted with hypertension-related conditions and many of those adults were using herbs specifically for hypertension. More than 75% of the adults in our study reported using herbal medicines in the past and nearly 70% had attended a traditional healer. Approximately 25% of adults were currently using herb specifically for management of hypertension. This study is the first to describe the pattern of herbal and alternative medicine in African inpatients. This pattern is broadly similar to findings from hypertension clinics in South and West Africa [6,7,13,14], as well as International Journal of Hypertension 7  International Journal of Hypertension community-based studies in Eastern and Western African countries [8,9]. In all of these settings, approximately onequarter of African adults with hypertension reported the use of traditional medicines and most of these were using herbal medicines. Because herbal medicines in particular are such an important part of hypertension self-management throughout Africa, collaboration with traditional healers should be explored as a possible route for improving hypertension care [9].
In the current study, nearly 20% of adults reported stopping their allopathic medicines for herbal medicine and >20% were concurrently using both herbal and allopathic medications. More than 70% of patients believed that there is no problem to use herb together with allopathic medications. These findings are concerning. Although some of the herbs used by our study subjects are likely harmless, most patients did not know what they were taking, and concurrent use of some herbs could hamper the effectiveness of allopathic antihypertensive drugs such as the interaction between St. John's Wart and calcium channel blockers or the interaction between garlic and thiazide diuretics [15,16]. Interactions between herbs and cardiovascular drugs could also increase the toxicological effects of cardiovascular drugs such as the interaction between garlic and warfarin or the interaction between Ginseng and Digoxin [5,17,18]. Sodium-sparing herbal aquaretics such as dandelions, for example, may offset antihypertensive effects of thiazide diuretics [16]. The use of ginseng with digoxin is associated with increased digoxin concentration in the blood and hence digoxin toxicity [19]. African adults with hypertension who are using herbs clearly require counseling not to stop their allopathic medicines for herbs as well as counseling as to possible interactions between the herbs they are using and their cardiovascular medications.
Factors that were significantly associated with the use of herbal medicines for hypertension included lower education level, nonprofessional occupation, lack of medical insurance, and history of comorbid stroke. The lower education level was also found to be associated with the use of herbal medicines in Nigeria [14]. Occupation has also been shown to be associated with herbal medicine use in another study from West Africa [9]. It is possible, though, that individuals of higher education level and occupation status are simply less willing to report herbal medicine use and this deserves further investigation. Inability to afford allopathic medicines was associated with herbal medicine use in Ghana [6]. The association between herbal medicine use and stroke may be related to the cultural perception in Tanzania that stroke may have an extra-biological (i.e., "magical") origin, a perception that has been associated with herbal medicine use in other studies [9]. These data indicate populations for whom more intensive counseling should be provided regarding herbal medicines. Providing insurance may have a beneficial effect.
Among patients who were aware of the herbal preparations that they are taking, garlic was the most frequently used herb. Garlic (Allium sativum) has been closely studied. Garlic has demonstrated multiple cardiovascular benefits that include lowering blood pressure, inhibiting platelet aggregation, enhancing fibrinolytic activity, and reducing cholesterol and triglyceride levels. However, despite all this evidence the routine use has not been advised because of methodological shortcomings of the published literature [20] and extreme caution is advised if garlic is taken concomitantly with CYP2E1 substrates [15] or in patients taking warfarin [16]. Studies are needed to determine possible benefits and harms of native African herbs that are used by traditional healers for hypertension.
Over 90% of study subjects reported that their doctors did not ask them about the use of either herbs or other alternative therapies for hypertension. These findings are consistent with other studies from West Africa which indicate (1) that the overwhelming majority of healthcare workers are not aware of the herbal medicine use of the hypertensive adults in their care [6,14] and (2) that a leading reason for this is fear and lack of inquiry of the part of the healthcare worker [6].
In our study, current use of traditional medicines for hypertension was not associated with lower compliance to allopathic medicines, as it has been in prior studies [6,13]. It is therefore important that doctors not only ask about herbs and other alternative therapies among adults admitted with hypertension but also provide careful, specific, and nonjudgmental recommendations. The authors suspect that more patients will be willing to listen and follow advice if clinicians apply this careful, discerning approach. Further studies are needed to determine the best way to ask and educate about herbal medicine use in the African context.

Conclusion
The use of herbs and other alternative therapies are very common among patients with hypertension. Nearly a quarter of the hypertensive adults in our study reported recently taking herbs specifically for the treatment of hypertension. Adults from low socioeconomic status, those with misunderstandings about hypertension, and those without insurance were more likely to take herbs. Study subjects often chose to take herbs due to poverty, family pressure, and a desire for a complete cure. Use of herbal or alternative therapies was not associated with noncompliance. Their doctors had asked very few patients about herbal medicine use. Open, nonjudgmental communication between healthcare workers and patients regarding use of herbs and other alternative medicines must be encouraged in Africa. Future investigations should focus on identifying the magnitude of the problem in the community level, traditional healer's knowledge about hypertension, and the timing of initiation of herbal use and alternative therapies.

Conflicts of Interest
Anthony Liwa was supported by United States Agency for International Development (USAID) leadership training program. Luke Smart and Robert Peck have each received grants from the National Institutes of Health Fogarty International Center (Grant nos. R25TW009337 and K01TW010281, resp.). Other authors report no specific funding in relation to this research and have no conflicts of interest to disclose.