Hypertension in low-income and middle-income countries (LMICs) is largely undiagnosed and uncontrolled, representing an untapped opportunity for public health improvement. Implementation of hypertension control strategies in low-resource settings depends in large part on cost considerations. However, evidence on the cost-effectiveness of hypertension interventions in LMICs is varied across geographical, clinical and evaluation contexts. We conducted a comprehensive search for published economic evaluations of hypertension treatment programmes in LMICs. The search identified 71 articles assessing a wide range of hypertension intervention designs and cost components, of which 42 studies across 15 countries reported estimates of cost-effectiveness. Although comparability of results was limited due to heterogeneity in the interventions assessed, populations studied, costs and study quality score, most interventions that reported cost per averted disability-adjusted life-year (DALY) were cost-effective, with costs per averted DALY not exceeding national income thresholds. Programme elements that may reduce cost-effectiveness included screening for hypertension at younger ages, addressing prehypertension, or treating patients at lower cardiovascular disease risk. Cost-effectiveness analysis could provide the evidence base to guide the initiation and development of hypertension programmes.
- health economics
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
What is already known?
Implementation of hypertension control strategies in low-resource settings depends in large part on cost considerations, but evidence on the cost-effectiveness of hypertension interventions from low-income and middle-income countries (LMICs) is sparse and varied across geographical, clinical and evaluation contexts.
What are the new findings?
Most interventions that reported cost per averted disability-adjusted life-year were cost-effective using national income thresholds, but gaps in evidence exist on programme elements that can affect cost-effectiveness in LMICs, such as task-sharing, risk-based treatment and standardised treatment protocols.
What do the new findings imply?
Hypertension control is found to be a cost-effective intervention for many LMICs. Gaps in evidence can be filled by economic evaluation of programme elements that include shifting some healthcare tasks to non-physician providers, integrating cardiovascular disease (CVD) risk assessment into treatment decisions and incorporating standardised CVD prevention programmes.
Hypertension in low-income and middle-income countries (LMICs) remains largely undiagnosed, untreated and uncontrolled despite being a leading factor in preventable death and disability (Chow et al, 20131; Ibrahim and Damasceno, 20122; Lozano et al, 20183; and WHO, 20134). The suboptimal treatment of hypertension in LMICs represents an untapped opportunity for public health improvement (Frieden and Bloomberg, 2018).5 Recent estimates suggest that nearly 40 million hypertension-related deaths can be avoided over the next 25 years by scaling up hypertension treatment to 70% (Kontis et al, 2019).6
Hypertension management depends on consistent and reliable access to healthcare. Areas with documented shortages of healthcare workers and with limited access to formal healthcare, such as sub-Saharan Africa, have fared the worst in addressing hypertension (Geldsetzer et al, 2019).7 At the population level, weak hypertension control and insufficient cardiovascular disease (CVD) prevention in LMICs can have broad implications that exceed the direct health consequences. For example, clustering of hypertension-related disease in younger adults, which is disproportionately more common in LMICs than high-income countries (Roth et al, 2018),8 has considerable socioeconomic effects, contributing to productivity and income losses at the household level and impeding macroeconomic growth (Bloom et al, 2011).9
While the value of addressing hypertension in LMICs has gained recognition over the past decade, resources in this area remain limited, revealing a gap between health targets and current needs (United Nations (UN), 201110;UN, 201511; and WHO, 2018a12). The transition from goal setting to actual implementation of hypertension control strategies in LMICs depends in large part on cost considerations. Although some economic modelling suggests that both population-level and clinical interventions for hypertension control can be cost-effective (Murray et al, 200313; Jha et al, 201214; Nugent and Brouwer, 201515; Bertram et al, 201816; and WHO, 2018b17), policymakers in individual countries might regard aggregate global estimates to be insufficient evidence for policy formulation in specific country circumstances. To inform policy decisions regarding hypertension approaches in LMICs, we reviewed the current evidence on costs and cost-effectiveness of hypertension interventions across LMICs. The contribution of this study is twofold. First, it provides the first comprehensive review of the evidence on cost-effectiveness of hypertension management programmes in LMICs. This review summarises the available evidence most relevant to policymakers in countries where hypertension management is currently limited or absent, and where decision-makers may be considering additions to health benefit packages without detailed cost or cost-effectiveness information. Second, this review documents the variation among existing studies across study designs and study quality. It produces a standardised quality score and explores contextual differences such as those that may arise between programmes based exclusively on pharmaceutical intervention and programmes that incorporate non-pharmaceutical components; programmes that target hypertension populations with different levels of CVD risk; or programmes applied in countries with different income levels. This too provides informative evidence to decision-makers in LMICs. The results describe a range of clinical programmes and corresponding programme cost and cost-effectiveness estimates from different settings, with varying levels of quality. We found gaps in evidence on programme elements that can affect cost-effectiveness in LMICs, such as shifting of healthcare tasks to non-physician providers, integrating CVD risk assessment into treatment decisions and standardising CVD prevention approaches.
Patient and public involvement
No patients or human subjects were involved in the process of conducting this literature review.
In March 2019, we searched for articles on economic evaluation of hypertension treatment programmes in LMICs using PubMed, the Cochrane Collaboration Database of Systematic Reviews, the Tufts Cost-Effectiveness Analysis Registry, the UK’s National Institute for Health and Care Excellence (NICE) guidelines, the University of York Centre for Reviews and Dissemination and the Disease Control Priorities (3rd Edition). To guide the search eligibility criteria, we developed a PICOTS table summarising the inclusion and exclusion criteria across the following elements: population, intervention, comparator, outcomes, time frame, settings and study design (Liberati et al, 2009)18 (see online supplementary appendix table A1). The search was performed using Medical Subject Headings (MeSH) and search terms related to hypertension and the pharmacological treatment, diagnosis, screening and management of hypertension. The list of MeSH terms can be found in online supplementary appendix table A2. We also used search terms for world regions; all low-income, lower middle-income and upper middle-income country names; newly classified high-income countries in South America, the Caribbean and the Pacific; and economic terms related to costs and cost-effectiveness. The PubMed search strategy can be found in online supplementary appendix table A3. We performed a supplemental ad hoc literature scan without MeSH terms in May 2020 to account for the lag in indexing and to capture any recent articles. The initial search identified 60 articles for inclusion in the review while the supplemental scan identified an additional 11 relevant publications. Results were not limited by publication date.
An inclusion/exclusion guide was created for reviewing the abstracts and full-text of articles (see online supplementary appendix table A4). Articles were included if they involved an intervention related to clinical screening, treatment and management of hypertension. Articles were excluded if they were designed for other diseases for which hypertension may be a risk factor or common comorbidity, or if they were for surgery patients to address acute events related to hypertension. Articles were excluded if they looked only at the cost of hypertension, with no reference to a specific intervention; only studied the prevalence of hypertension; if they did not involve any clinical setting; or, if they studied knowledge or awareness of hypertension. Studies that were conducted in high-income countries, or in territories or associated states of high-income countries (with the exception of South America, the Caribbean and the Pacific), studies that were published in a foreign language, and any article that was an editorial, review, correspondence or abstract related to study design and protocol were also excluded.
Overall, 595 references were identified: 534 from PubMed and 61 from other databases and sources. Screening abstracts identified 163 articles for full-text review of which 71 were identified as relevant for inclusion in the analysis (see online supplementary appendix table A5). Of these, 42 studies across 15 countries provided estimates of cost-effectiveness, with the rest evaluating costs only. A diagram of the search process is depicted in figure 1. Each of the 42 cost-effectiveness studies underwent a quality assessment based on a 13-question checklist informed by Drummond guidelines for economic evaluation of healthcare programmes (Evers et al, 2005).19 These studies were reviewed and assigned a total score equal to the sum of positive answers to the checklist questions.
Reported indicators included: cost per mm Hg reduction in systolic and/or diastolic blood pressure (table 1), cost per patient with controlled hypertension (table 2), cost per patient with hypertension (table 3), cost per averted disability-adjusted life year (DALY) (table 4) and cost per gained quality-adjusted life year (QALY) (table 5). Estimates were converted to constant 2017 US dollars (US$) and were adjusted to reflect annual amounts where applicable. Two studies reported estimates in purchasing-power-parity (PPP)-adjusted international dollars, which were not converted into US$ because appropriate conversion factors were not available for the blend of countries examined (Ortegon et al, 201220 and Murray et al, 2003). Studies in the above cost-effectiveness categories were further categorised according to intervention type, as follows. ‘Pharm only’ indicates interventions where pharmacotherapy is the only treatment element, encompassing various combinations of drugs and drug classes, different providers and delivery platforms. ‘Pharm plus’ indicates combination programmes that incorporate other forms of treatment for hypertension in addition to medications, such as patient education or lifestyle changes. ‘Other’ indicates interventions that did not evaluate changes in pharmacological treatment. Cost elements included costs of medication, laboratory work, labour, equipment, transportation, provider training and others.
Thirty-six of the identified studies were conducted in upper-middle-income countries (UMICs), 30 studies were from low-income and lower-middle-income countries (LLMICs) and five studies included countries of different income levels. Studies reported costs of hypertension treatment, cost-effectiveness of hypertension treatment or both. Twenty-five of the studies included only medication costs, while the remaining studies included health system costs and other services such as laboratory tests, health provider time and other screening costs. Study designs included longitudinal (seven studies), cross-sectional (four studies), modelled or simulated (22 studies), randomised control trials (seven studies) and retrospective cohort studies (two studies).
After conducting the quality assessment based on the 13-question checklist informed by Drummond guidelines for economic evaluation of healthcare programmes (Evers et al, 2005),19 the average quality score of the studies was 7.8. Modelled studies and randomised control trials tended to be higher quality, with average scores of 9.6 and 8.4, respectively. Longitudinal, cross-sectional and retrospective cohort studies were lower quality, with average scores of 5.0, 4.3 and 3.0, respectively (table 6).
Fifty-four studies described pharmaceutical-only interventions using various combinations of antihypertensive drugs and drug classes. Fifteen studies assessed pharmaceutical treatment plus at least one other component, such as providing physician training, implementing treatment guidelines or offering lifestyle advice. A small number of studies did not include pharmaceutical treatment and instead assessed cost-effectiveness of activities such as physician training, lifestyle education (Bai et al, 201321 and Jafar et al, 2011),22 or loaning out blood pressure self-measurement devices (Calvo-Vargas et al, 2001).23 Four different delivery platforms were represented across studies: community-based services; health centres providing basic medical care and staffed by a physician, nurse or mid-level healthcare provider; first-level hospitals that have the capacity to perform surgery and provide inpatient care; and referral or speciality hospitals that include general specialists and provide secondary and tertiary services. As such, care was provided by a range of providers that included physicians, nurses, pharmacists and community health workers.
Cost and cost-effectiveness evidence
Study results were reported across five outcome types: Cost per mm Hg reduction in systolic and/or diastolic blood pressure (13 studies; table 1); annual cost per patient with controlled hypertension (2 studies; table 2); annual cost per patient with hypertension (21 studies, 7 of which did not include a cost-effectiveness analysis; table 3); cost per averted DALY (14 studies; table 4); and cost per gained QALY (8 studies; table 5). Significant variability was present across studies due to cost differences even across studies with like interventions. For example, two interventions in UMICs both providing patient risk assessment, education, pharmacotherapy and adherence monitoring reported substantially different per patient costs for the intervention—US$6.19 to US$13.38 per patient in China (Bai et al, 2013) compared with US$203.85 in Brazil (Cazarim and Pereira, 2018).24 In this example, the analysis in China did not include the cost of drugs whereas the analysis in Brazil included indirect costs such as the cost of absenteeism resulting from missing work for doctor’s appointments. Across all the types of interventions, the range of estimates of the annual intervention cost per hypertension patient was wider in UMICs (ranging from US$6.2 for a non-drug intervention programme in China to US$2418 for a Pharm only programme in South Africa) than in LLMICs (ranging from US$25.6 for a Pharm only programme in Kenya public facilities to US$987 for a Pharm only programme in Kenya private facilities). Nonetheless, almost all studies in all countries yielded results below US$1000 per patient for any intervention (figure 2).
Median monthly drug costs were less than US$50 for the 23 studies with medication-specific costs of treatment by drug or drug combination group (figure 3); however, the lowest and highest monthly costs illustrate a wide range across contexts. The widest cost range was observed for monotherapy with angiotensin-converting enzyme inhibitors (ACEI) (US$0.18 to US$159 with a median monthly cost of US$11) and beta blockers (BB) (US$0.11 to US$153 with a median monthly cost of US$4.25), obtained from 13 studies for each medication type. Other commonly evaluated monotherapy plans focussed on diuretics (16 studies, with estimates ranging from US$0.12 to US$74 with a median of US$1.77), calcium channel blockers (CCB) (14 studies, with estimates ranging from US$0.79 to US$78 with a median of US$6.56) and angiotensin-II receptor blockers (ARB) (8 studies, with estimates ranging from US$1.37 to US$73 with a median of US$17). Other less common treatment plans, such as multiple-drug therapies and monotherapies involving alpha blockers, alpha-2 agonists, central acting antiadrenergics and central adrenergic inhibitors, had very limited representation with one to two studies each. Monotherapies with diuretics, BB and CCB were less costly while ACEI or ARB monotherapy incurred a higher median cost(figure 3). However, drug price variability across studies, reflecting cross-country differences in price, procurement and delivery context, prevents robust comparison of costs across treatment plans.
Of the 42 cost-effectiveness evaluations, 6 studies reported cost per averted DALY while also reporting differences across at least two CVD risk levels. Figure 4 describes the range of estimates across risk groups, in 2017 US$. Despite the wide range of cost-effectiveness estimates, most occurred below US$1000 per averted DALY. There was some indication that higher cost-effectiveness is associated with focussing on higher-risk patients (figure 4).
A common threshold for cost-effectiveness determination in LMICs is based on per capita gross domestic product (GDP), where an intervention is considered cost-effective if the cost per DALY averted or QALY gained is less than three times the annual per capita country GDP, and very cost-effective if the cost per DALY averted or QALY gained does not exceed the annual per capita GDP. Despite some limitations of the GDP threshold approach (Marseille et al, 201425 and Bertram et al, 2016),26 we used it as a guideline to compare cost-effectiveness across studies reporting DALY-based and QALY-based cost-effectiveness indicators. Hypertension interventions were found to be cost-effective in the majority of evaluations using the GDP threshold (tables 4 and 5). As figure 4 illustrates, most cost-effectiveness estimates in our review were clustered below US$1000 per averted DALY—well below the average 2017 GDP per capita for lower-middle income countries of $2188 (FRED,27 suggesting they could be very cost-effective for lower-middle income countries. Favourable cost-effectiveness levels using the GDP threshold were found for programmes in Argentina (Augustovski et al, 201828 and Rubinstein et al, 201029), Brazil (Obreli-Neto et al, 201530), China (Gu et al, 201531; Xie et al, 201832; Basu et al, 201633), South Africa (Gaziano et al, 200534), Tanzania (Robberstad et al, 200735), Vietnam (Ha and Chisholm, 201136 and Nguyen et al, 201637), India (Basu et al, 201633), Ghana (Gad et al, 202038), Thailand (Khonputsa et al, 201239), Sri Lanka (Lung et al, 201940), Ethiopia (Tolla et al, 201641), Nigeria (Ekwunife et al, 201342) and Nepal (Krishnan et al, 2019.43 A small number of studies indicated that cost-effectiveness thresholds were more difficult to meet in lower-income countries; for example, cost-effectiveness was not established for select intervention scenarios reported in Nigeria (Rosendaal et al, 201644 and Ekwunife et al, 2013) and Tanzania (Ngalesoni et al, 201645 and Robberstad et al, 2007) (table 4). Factors that were associated with not meeting the cost-effectiveness thresholds for their respective countries included treatment of patients at lower risk for CVD (Ekwunife et al, 2013 and Khonputsa et al, 2012), screening for hypertension at younger ages (for example, at age 35 vs 55, Nguyen et al, 2016), and addressing prehypertension (Chen et al, 201746 (table 5).
Several studies evaluated non-pharmaceutical interventions in addition to medication treatment. One study found a complex strategy that included community health worker home-based visits, physician education and text messaging promoting lifestyle change and medication adherence was less cost-effective than usual care (Augustovski et al, 2018). By contrast, three other studies estimated that interventions for hypertension management such as physician training were more cost-effective than usual care (Anchala et al, 201547; Jafar et al, 2011; and Wang et al, 201348).
The range of estimated costs and cost-effectiveness of hypertension programmes is wide, both across and within countries, reflecting heterogeneity in intervention design, cost components and country context. We broadly distinguished between intervention designs that involved pharmaceutical treatment only and those that included non-pharmaceutical components, such as provider or patient training, and between countries with different income levels. We did not observe clear distinctions in programme cost-effectiveness based on country group or inclusion of non-pharmaceutical programme elements; however, the large majority of interventions that reported cost per averted DALY were found to be cost-effective using national income thresholds, with costs per averted DALY not exceeding the average GDP per capita of lower-middle income nations. Some exceptions were observed in lower-income countries, where the cost-effectiveness cut-off, as defined by national GDP, is lower. This might suggest that hypertension management programmes in lower-income countries may warrant special consideration in terms of minimising costs relative to outcomes. However, the potential need to accommodate programmes in LMICs to lower cost-effectiveness thresholds is not necessarily generalisable. For example, a recent study from Nepal, a low-income country, detailed very high cost-effectiveness of a community-based hypertension management programme relative to its income threshold (Krishnan et al, 2019). Relatively higher costs per averted DALY were observed in scenarios that expanded treatment to younger age groups or to prehypertension, suggesting that more targeted treatment may improve cost-effectiveness. Median drug costs for monotherapies involving diuretics, beta blockers and calcium channel blockers appeared to be lower than those involving ACE inhibitors or combinations.
While this review did not establish a clear pattern in cost-effectiveness when comparing estimates of cost per averted DALY by patient CVD risk across studies, individual studies indicated that hypertension treatment tends to be more cost-effective when applied to populations at higher CVD risk (Ngalesoni et al, 2016; Praveen et al, 201849; Ha and Chisholm, 2011; Khonputsa et al, 2012; and Tolla et al, 2016), pointing to an important area for future research on the role of risk-tailored treatment. Hypertension treatment guidelines in LLMICs can be strengthened by further evidence translating the use of simple risk assessments based on age, smoking status and obesity into population-level efficiencies in CVD prevention (Kaptoge et al, 2019).50
In addition to the low comparability across intervention programmes in LMICs, this review is subject to a number of limitations. We did not review the economic literature for the potential of behavioural modifications such as low-sodium diet, healthy weight, physical activity and eliminating tobacco use (WHO, 2011)51 to control blood pressure. Such modifications have been promoted at the population level through national policies on taxation and/or regulation of products containing trans-fatty acids, excess sodium, tobacco and added sugar and the WHO has summarised those results in online supplementary appendix 3 of the Global Action Plan for Non-Communicable Diseases (WHO, 201752; Task Force on Fiscal Policy for Health, 201953; and WHO, 201354). Studies in this review did not specifically aim to evaluate improved access to medications; rather, they described the relative cost-effectiveness of different treatment approaches, or, less frequently, the relative effectiveness of the same treatment approach across different study groups. Three studies that compared the cost per hypertension patient with treatment relative to no treatment found, as expected, that costs increased with the initiation of treatment (Cazarim and Pereira, 2018; Gaziano et al, 2005; and Obreli-Neto et al, 2015). This review does not assess the cost-effectiveness of population-level approaches that can improve hypertension and is mostly limited to studies with health-systems perspective rather than societal perspective. Programme evaluation from the health system perspective rather than the social perspective presents a narrower view of hypertension interventions. Another limiting aspect is that many studies did not specify the type of provider involved in the intervention, precluding inferences about costs associated with different provider type or delivery platform. Comparisons of drug class combinations were limited by lack of information on underlying drivers of drug price such as generic or brand status or type of drug within a drug class.
To reduce the knowledge gap about factors that can influence the cost-effectiveness of hypertension programmes, future research can focus on programme elements that may be particularly relevant to low-resource settings, such as the uptake of healthcare tasks by non-physician providers and the assessment of patient CVD risk in treatment determination. Using community health workers (CHW) in the provision of chronic disease care has been associated with increased cost-effectiveness in the USA (Kim et al, 201655), and has been similarly regarded in LMICs (Jeet et al, 201756 and Krishnan et al, 2019), but evidence specific to hypertension care costs is mostly lacking. Additional studies focussing on the role of CHW in improving the cost-effectiveness of hypertension interventions can help inform health strategies in areas where access to care is otherwise limited. Standardisation of cost evaluation platforms can streamline economic assessment across countries. An example of a mechanism for evaluating the costs of standardised CVD prevention approaches is provided by the costing mechanism for the HEARTS package of clinical guidelines for CVD prevention in primary care (WHO, 201657 and WHO, 201758). A list of standard cost elements to track during implementation of hypertension management programmes is included in table 7, which summarises the leading cost indicators of HEARTS programme components, including establishment of treatment protocols, training of healthcare providers in lifestyle counselling and risk-based management, ensuring access to essential medicines and promoting task sharing and systems for patient monitoring. Additional evidence on the cost-effectiveness of introducing non-physician health workers in healthcare delivery can inform future approaches to address physician scarcity (Seidman and Atun, 201759; Jafar et al, 2011; and Chen et al, 200460).
Although CVD death rates have decreased worldwide since 1990, improvements have not been evenly distributed across countries, and have showed signs of slowing down (GBD, 2018). Both domestic and external financing for non-communicable diseases across LMICs remains low (IHME, 2019). The results of this review suggest that hypertension control approaches can be a cost-effective way to prevent premature CVD in LMICs across a variety of population, clinical and health system contexts.
Handling editor Lei Si
Contributors GS conducted a comprehensive literature search. All authors contributed to the analysis, drafting and editing of the manuscript.
Funding GS and RN received support from the CDC Foundation with funds provided by Resolve to Save Lives, a division of Vital Strategies. Resolve to Save Lives is funded by grants from Bloomberg Philanthropies; the Bill and Melinda Gates Foundation; and Gates Philanthropy Partners, which is funded with support from the Chan Zuckerberg Foundation. The funders had no role in the design of this study and did not have any role during its execution, analyses, interpretation of the data or decision to submit results.
Disclaimer The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon request. As a review article, this article reports data from previously published studies.