Study selection
The systematic search of six electronic databases resulted in a total of 2175 records. We removed 767 duplicates in Endnote and exported the remaining 1408 records to Rayyan QCRI for the title and abstract screening. We excluded 1302 ineligible titles and abstracts and retrieved the full-texts of the remaining 106. Out of 106 full-text articles that we assessed, 94 had either ineligible outcomes, were conducted among non-hypertensive or non-Nepali participants, or were reviews, case reports or conference abstracts. We did the forward and backward citation tracking for the remaining 12 studies and identified two additional records [43, 44]. No additional eligible studies were found in Nepalese journals and on websites of the Nepali Government, professional and regulatory organisation, and national and international agencies. Finally, we included 14 studies [43-56] for qualitative synthesis (see Fig 2). None of the studies were included in the quantitative synthesis of effect sizes because of the significant differences in their aims, methods, interventions/exposures and outcome measures.
Study characteristics
Most of the included studies were conducted in either province 3 (Kathmandu and its periphery, n=8) or province 4 (Kaski, n=4) or province 1 (Sunsari, n=2). The numbers of studies conducted in hospital (n=8) and community (n=6) settings were nearly equal. All the community-based studies included participants from the peri-urban (n=3) and urban (n=3) areas. Only a study conducted in Acham (hospital and community-based) included participants from a rural area [47]. Half of the studies were completed in 2015 or later; and more than half of the studies assessed the barriers and enablers of hypertension treatment and control [51-56]. They applied qualitative methods (n=3) [52-54], quantitative methods (n=4) [51, 56] and mixed methods (n=1) [55]. The majority of these studies investigated hypertension treatment as an outcome of interest (Table 1). Two studies also discussed adherence to antihypertensive medications [56]. Most studies collected data from hypertensive participants, except one study that interviewed community health workers to explore the barriers and enablers in utilising healthcare among the patients [54].
Six of the included studies that assessed the effectiveness of hypertension treatment and control strategies [45-50] were prospective comparative studies (n=2) [49, 50], randomised trials (n=2) [46, 48] and uncontrolled before and after studies (n=2) [45, 47]. Five out of these six studies reported systolic and diastolic blood pressure as outcome measures. The studies tested the effectiveness of health education [45-49] and combined, antihypertensive medication and yoga [48] interventions (Table 2).
The total number of participants in the included studies was 2854, with the individual sample sizes ranging from n=13 to n=1638. Only two studies applied multivariable analysis and reported adjusted effect sizes [46, 56]. Most of the qualitative and mixed methods studies (3 out of 4) applied a thematic analysis. Three studies [46, 47, 52] reported non-response rates, and they ranged between 9% and 27.8%.
Study quality
In quality assessment, 10 studies received three or more stars, suggesting they were of fair or good methodological quality (Additional file 2). Only two out of 11 quantitative studies [46, 56] adjusted for confounding.
Barriers and enablers of hypertension treatment and control
Analysis of qualitative studies
We identified several themes on barriers to hypertension treatment and control discussed in the qualitative studies. We categorised them into three major domains: HS excluding provider and patient (HS), providers and individual-level barriers (see Table 3). The unaffordability of health services [52], lack of human resources (primarily in rural areas) [54] and the unavailability of diagnostic tools [54] were the most often cited HS barriers.
All qualitative studies discussed the gap in communication between providers and patients that impeded medication adherence and routine follow up. Health care providers failed to deliver clear information to patients regarding medication dosage and duration, behaviour modification and need for routine monitoring. A hypertensive male participant from the study conducted by Shrestha et al. [52] complained that the doctor did not explain enough about his condition.
“….. After check-up, I was told to take medicine. …..I was not told anything. So, I asked people who have heart disease to get the information regarding what food to eat, which food increases it, and which food controls pressure.” [52]
A 55 years old participant with uncontrolled hypertension from a study in Kathmandu reported not visiting the doctor for one and half years since her last visit [55] as the doctor did not require for regular visits.
“I have not gone for follow-up. I am following the same regimen from last one and a half years. My doctor told me to visit him only if I had problems” [55]
In addition, other provider-related factors affecting hypertension treatment and control were: health worker’s lack of interest in counselling for lifestyle modifications [52-55]; long waiting hours for the appointment [52, 55]; lack of national guidelines in hypertension treatment [55]; and provider's negative behaviours [52].
A wide variety of themes evolved from the qualitative analysis of the findings on patient’s beliefs and practices that affect hypertension treatment and control (table 3). The factors impeding initiation of anti-hypertensive treatment were poor help-seeking behaviours [54] and reluctance to take medication due to perceived side effects and fear of long-term use [52, 55]. Other factors related mostly to treatment adherence and lack of routine monitoring of patients [54, 55].
Three studies discussed facilitating factors for improved hypertension treatment. These factors were: access to hypertension care [54], treatment adherence [52], patient’s attitude [52] and family support [52, 53].
Analysis of quantitative studies
Five studies reported findings based on quantitative data. The studies identified the factors associated with hypertension treatment, adherence and control (Table 4). Affordability of the medication was associated with treatment adherence [56]. At the level of the provider, prescribing more than one type of pill was associated with greater non-adherence [56]. Similarly, long waiting hours (>20 minutes vs ≤ 20 minutes) and lack of proper counselling from health workers were associated with uncontrolled hypertension [55]. All quantitative studies reported various individual (patient) level factors that were significantly associated with untreated or uncontrolled hypertension or non-adherence to medication. These were: patients’ lack of awareness on normal blood pressure target [51, 55]; lack of knowledge about possible complications of hypertension [51]; lost to follow up [55, 56]; non-adherence to medication [51, 55]; and lack of blood pressure monitoring [51].
Only one study discussed enablers separately. The study found that scores of dimensions of illness perception particularly timeline, treatment control and coherence were positively correlated with medication adherence [44]. The higher the participant perceived high blood pressure as a chronic condition (Spearman correlation coefficient (r)=0.23, p<0.05), the better was medication adherence. The more the participant believed that treatment can control the blood pressure (r=0.51, p<0.05), the higher the medication adherence score they had. The better the understanding of hypertension (coherence, r=0.22, p<0.05), the higher was the medication adherence [44].
Strategies for hypertension treatment and control
Out of six eligible studies in this category, all studies intervened at the patient level, two studies also investigated the effect of workforce strengthening (training and continuing education for integrated of NCD care) and one study evaluated the impact of new data recording systems at the healthcare centre (Table 5). The most frequently tested antihypertensive strategy was health education for hypertensive patients [45-47, 49], and three out of four studies found it effective in reducing blood pressure. Health education was applied as a component of comprehensive blood pressure management strategies [46, 47, 49] or as a single intervention [45]. Two studies [49, 50] also compared hydrochlorothiazide, enalapril and amlodipine as first-line antihypertensive drugs. The reduction in mean systolic and diastolic pressure was significantly higher with enalapril and amlodipine, compared with hydrochlorothiazide [49, 50]. The post-intervention changes in blood pressure did not vary significantly between enalapril and amlodipine groups after for weeks [49, 50]. The participant under amlodipine reported more adverse events such as peripheral oedema, shortness of breath and headache than the enalapril group.
One study studied yoga as an adjuvant therapy to medication and compared it with usual care [48]. The yoga group had a significant reduction in both systolic and diastolic blood pressure, compared with the ‘medication-only’ group after six weeks of follow-up.
One of the two studies that intervened at the level of healthcare providers found a significant improvement in blood pressure amongst their clients. The study trained and involved female community health volunteers for health education, blood pressure monitoring and referral of hypertensive cases. The intervention was effective in reducing blood pressure. The mixed-effect regression coefficient was –4.9 (95% CI: –7.8 to –2.0) for systolic blood pressure and –2.6 (95% CI: –4.6 to –0.7) for diastolic blood pressure [46]. One study implemented the intervention program at HS level, by establishing a new digital and electronic health record system along with capacity development for healthcare workers and providing health education to the patients. The intervention had no statistically significant effect on hypertension control [47].