Comment on: “Impact of home-based follow-up care on hypertension control among patients attending Thika Level Five Hospital, Kiambu County, Kenya: a randomized controlled study”: reply

Dear Editor, We thank Ampofo and colleagues for their thoughtful and collegial commentary on our randomized controlled study evaluating home-based follow-up care for hypertension management at Thika Level Five Hospital, Kiambu County, Kenya.1,2 We welcome the opportunity to clarify aspects of the study design, analytical approach and intervention delivery that may assist readers in interpreting the findings within their appropriate methodological and policy context. Our study was conceived as a pragmatic, effectiveness-oriented trial aimed at assessing whether an integrated, community-supported, follow-up model could improve hypertension control compared with routine hospital-based care in a public-sector setting. The primary outcome was hypertension control at 6 mo, chosen as a clinically and programmatically meaningful endpoint aligned with routine chronic disease management and policy decision-making. Sample size estimation was guided by feasibility considerations and anticipated clinically meaningful effect sizes informed by prior community-based hypertension interventions in similar settings. Given operational constraints typical of low- and middle-income health systems, the study was powered to detect differences of practical relevance rather than small incremental effects across multiple outcomes. Repeated blood pressure measurements were obtained within individuals over time. Our primary analyses focused on pre- and post-intervention comparisons within and between randomized groups using established statistical tests commonly applied in pragmatic trials. Randomization resulted in comparable baseline characteristics, and all participants included in follow-up analyses had uncontrolled hypertension at baseline. While we acknowledge that alternative longitudinal modelling approaches may provide additional statistical efficiency, we prioritized analytical methods that were transparent and readily interpretable for clinicians, health managers and policymakers, who represent the intended audience for this work. Analyses were conducted among participants with documented follow-up assessment. Attrition was low, and missing outcome data were minimal. In this context, formal imputation procedures were not undertaken. We agree that future studies with larger samples, longer follow-up and broader analytical capacity would benefit from prospectively specified intention-to-treat frameworks and longitudinal modelling approaches. The intervention evaluated was intentionally designed as a multicomponent, integrated care package incorporating home visits, counselling, text message reminders and structured follow-up. Our objective was not to isolate the effects of individual components, but to assess the effectiveness of the intervention as it would realistically be delivered within existing Kenyan health system structures. Operational monitoring of intervention delivery was undertaken to support implementation fidelity, although a formal process evaluation was beyond the scope of the present report. We agree that future work would benefit from explicit integration of process measures to inform scalability and adaptation. We appreciate the authors’ emphasis on clear reporting and rigorous analytical framing, particularly for trials intended to inform policy and scale-up decisions in resource-constrained settings. We view this exchange as constructive and believe that these clarifications further strengthen the contribution of our study to the growing evidence base supporting community-oriented strategies for hypertension control in sub-Saharan Africa. Yours sincerely, Veronica Njaramba, David Kamau, Jesse Gitaka Veronica Njaramba (Conceptualization, Writing—original draft, Writing—review & editing), David Kamau (Investigation, Writing—review & editing), and Jesse Gitaka (Conceptualization, Methodology, Writing—review & editing) None declared. None declared. Not required for this correspondence. All related data have been presented within the article.