How a Global Health Team Plans to Map the Hidden Playbook for Merging Diabetes and HIV Care

Across much of the world, the clinics best equipped to keep people alive with chronic disease were built for a different epidemic. In low- and middle-income countries, decades of targeted investment created remarkably effective vertical programs for HIV and tuberculosis: systems with routine follow-up, patient tracing, decentralized medication distribution, and recurring clinical contact. Now, as type 2 diabetes, hypertension, obesity, and dyslipidemia surge through the same populations, researchers are asking a deceptively simple question with enormous consequences: when health systems bolt cardiometabolic care onto these established platforms, what exactly are they doing to make the integration work? A team led by Amanda Brumwell of the University of Washington has published a detailed protocol for a scoping review designed to answer that question systematically, registered prospectively on the Open Science Framework before any data collection begins.

The scale of the underlying problem is difficult to overstate. Globally, an estimated 1.28 billion adults live with hypertension, 589 million live with diabetes, and more than one billion live with obesity, with low- and middle-income countries home to both the greatest proportion of these cases and the fastest growth. Dyslipidemia, a leading risk factor for cardiovascular mortality, affects roughly 24.1 percent of adults in the form of hypertriglyceridemia and 28.8 percent as hypercholesterolemia, though the burden is unevenly quantified in lower-income settings. Together, these four conditions account for a large and growing share of disability-adjusted life years attributable to noncommunicable disease, particularly in resource-constrained health systems. Effective treatments exist for each condition, and controlling blood pressure, blood glucose, weight, and cholesterol can prevent myocardial infarctions, strokes, and chronic kidney disease. Yet a substantial gap persists between what medicine knows and what health systems deliver.

That gap has structural roots. Health system improvements in many LMICs have been driven by policies and funding streams that prioritized individual diseases, most notably HIV and tuberculosis, along with maternal and child health. The result is vertical alignment: services from the community level to referral hospitals are primed for a specific disease, with consumables, human resources, laboratory and diagnostic systems, information systems, and funding all oriented toward eliminating one condition. Cardiometabolic diseases, whose onsets are slow and whose etiology involves overlapping socioeconomic and behavioral drivers, fit poorly into this architecture. The epidemiologic overlap, however, is striking. Hypertension co-occurs frequently with HIV, and type 2 diabetes raises susceptibility to tuberculosis, creating a biological rationale for merging services that the health system’s structure has historically kept apart.

Integrated care models co-locating cardiometabolic and infectious disease services have been increasingly documented, much of the published evidence originating from sub-Saharan Africa, though similar approaches appear across South and Southeast Asia and Latin America. But here is the technical crux that motivates the new review: the implementation strategies used to design, deploy, and sustain these integrations have never been systematically characterized. A recent scoping review by Luo and colleagues found that fewer than a third of included studies described a fully integrated service delivery model, and that reported strategies clustered narrowly around training, evaluation, and provider assistance. Strategies, in other words, remain thinly specified even where integration itself has been documented, leaving policymakers without actionable guidance on how to translate integration evidence into health systems reorganization.

The Brumwell protocol extends prior work in three deliberate ways. First, it adds obesity and dyslipidemia to the conditions of interest, which have received comparatively less attention in integration programs despite being established cardiovascular risk factors that frequently co-occur with diabetes and hypertension. Second, it treats primary care as a service platform of equal interest alongside HIV and tuberculosis, recognizing that primary care is the only one of the three whose core mandate already includes chronic disease care and that reaches adults not enrolled in any disease-specific program. Third, it specifies each implementation strategy at the granular level recommended by Proctor and colleagues, capturing the actor, action, action target, temporality, dose, implementation outcome affected, and justification, rather than simply counting broad strategy categories.

Methodologically, the review follows the frameworks of Arksey and O’Malley and the Joanna Briggs Institute, reporting in accordance with the PRISMA-ScR checklist. The team will search five databases: PubMed, Embase, Cochrane Library, Web of Science, and CINAHL. A notable technical feature is the search validation procedure. Before finalizing the search string, the investigators will assemble a benchmark set of approximately five to ten known relevant articles, deliberately varied in geographic setting, care platform, and cardiometabolic condition. The final string, developed with a health sciences librarian using database-specific controlled vocabulary such as MeSH and Emtree, will be tested against this benchmark set across each database independently. If a benchmark article is not retrieved, the team will examine whether the failure reflects database indexing limitations rather than search string deficiencies before making adjustments. A fourth group of search terms, leveraging World Bank country classifications, ensures that LMIC settings are identified through the search itself rather than only at screening.

Eligibility criteria apply the Population-Concept-Context framework from the Joanna Briggs Institute. Included studies must involve non-pregnant adults aged 18 or older receiving care for type 2 diabetes, hypertension, obesity, or dyslipidemia, and must describe a deliberate implementation action, which may include provider training, task shifting, health information system modifications, clinical workflow redesign, decentralization of care, or program scale-up. Importantly, the team will not require that articles use implementation science terminology, since integration efforts are often reported as programs, pilot projects, quality improvement activities, or ministry-led initiatives that lack such framing. To define integration depth consistently, the protocol uses Leutz’s framework, classifying each example as linkage, coordination, or full integration. Articles eligible for full-text extraction must report integration at the level of coordination or above, excluding referral and co-location alone. The content of integration will be characterized using Atun and colleagues’ framework, recording which health system functions were affected: governance, financing, planning, service delivery, monitoring and evaluation, and demand generation.

Screening and extraction proceed through staged calibration designed to quantify and control reviewer error. Four reviewers will dually screen a random 10 percent of articles at title and abstract stage, aiming for at least 85 percent agreement quantified with Cohen’s kappa, before proceeding to the remainder. Full-text review requires 90 percent agreement in its calibration sample, with a random 10 percent of singly-reviewed articles independently checked by a second reviewer. Data extraction, conducted entirely by human reviewers with no AI or automated tools, follows the same architecture: form piloting on five to ten articles, duplicate calibration on a random 10 percent sample, and a 10 percent double-extraction quality check on the remainder. Reviewers will record a verbatim quotation supporting each coded implementation strategy and will note whether mapping to the ERIC taxonomy was explicit or inferred, coding strategies only from what articles describe rather than inferring unstated strategies from plausibility.

Methodological quality will be appraised using the Mixed Methods Appraisal Tool, version 2018, selected because the inclusion criteria span qualitative, quantitative, mixed-methods, and economic designs. Consistent with scoping review methodology, appraisal results will characterize the strengths and limitations of the evidence base rather than serve as an exclusion threshold. Findings will be synthesized as a narrative and tabular evidence map characterizing implementation strategies, implementation outcomes assessed using the Proctor et al. taxonomy of acceptability, adoption, appropriateness, cost, feasibility, fidelity, penetration, and sustainability, evidence-based interventions, and contextual factors. Strategies lacking analogues in the ERIC framework will be documented as potential innovations. The team anticipates limitations, including possible thin reporting of strategies, which they will treat as a finding of the review; the exclusion of perinatal cardiometabolic conditions, which they recommend for a separate review; and language bias from restricting eligibility to English-language publications.

The timing may prove critical. Recent changes in international aid and biomedical research policies have halted infectious disease programs that relied on external support, with documented impacts on HIV and TB services in several countries. As the authors note, the combination of co-occurring hypertension and diabetes, emerging evidence on the acceptability and feasibility of integration, and disruption to vertically funded programs may place health systems at a juncture where rethinking service organization could improve the quality, efficiency, and resilience of care. The completed evidence table will be made available as supplementary material, and the team plans to share a summary of the evidence map with ministries of health and implementing partners in the settings where they work, so that the findings can directly inform decisions about integrating cardiometabolic care into established service platforms. For health systems facing a double burden of chronic and infectious disease, the review aims to supply something previously missing: a rigorous, transparent inventory of the playbook for making integration actually happen.

Subject of Research: Implementation science protocol for a scoping review of cardiometabolic care integration strategies in LMIC health systems

Article Title: Protocol for a scoping review of implementation strategies for integrating evidence-based cardiometabolic disease care for adults into siloed health services in low- and middle-income countries

Article References: Brumwell, A., Shah-Rohlfs, R., Nickel, M., Kocher, E., Ali, M., Mocumbi, A. O., Neuhouser, M., Page, S., Shapiro, A., Sherr, K., & Gimbel, S. (2026). Protocol for a scoping review of implementation strategies for integrating evidence-based cardiometabolic disease care for adults into siloed health services in low- and middle-income countries. Frontiers in Health Services, 6, Article 1961796. https://doi.org/10.3389/frhs.2026.1961796

Image Credits: AI Generated

DOI: 10.3389/frhs.2026.1961796

Keywords: cardiometabolic disease, implementation strategies, scoping review, integrated care, low- and middle-income countries, type 2 diabetes, hypertension, obesity, dyslipidemia, HIV, tuberculosis, primary care