Abstract
{ "background": "Public health surveillance systems are critical for disease control, yet robust methodological evaluations of their adoption in sub-Saharan Africa are limited. In Kenya, diverse surveillance initiatives have been implemented, but evidence on their uptake and the methodological rigour of evaluation studies remains fragmented.", "purpose and objectives": "This meta-analysis aims to methodologically evaluate quasi-experimental studies that measure adoption rates of public health surveillance systems in Kenya, assessing the strength of causal inference and identifying common design limitations.", "methodology": "We systematically reviewed published and grey literature. Eligible studies employed quasi-experimental designs (e.g., difference-in-differences, interrupted time series) to assess adoption. Data were extracted on design characteristics, adoption metrics, and statistical methods. A random-effects meta-regression model, $\\logit(p{i}) = \\mu + \\beta X{i} + u{i} + \\epsilon{i}$, was fitted to synthesise adoption proportions, with robust standard errors used for inference.", "findings": "The pooled adoption rate across studies was 0.42 (95% CI: 0.35 to 0.49), with significant heterogeneity (I² = 87%). Studies utilising a difference-in-differences design with propensity score matching reported significantly higher methodological quality scores (p < 0.01). A key thematic finding was the frequent omission of fidelity measures for the independent variable (system implementation).", "conclusion": "Quasi-experimental evaluations of surveillance system adoption show moderate average uptake but are marked by substantial variability and recurrent methodological shortcomings, particularly in verifying intervention exposure.", "recommendations": "Future evaluations should incorporate explicit fidelity assessments and utilise stronger quasi-experimental designs with matching techniques to improve causal validity. National reporting guidelines for surveillance evaluations are warranted.", "key words": "health surveillance, programme evaluation, quasi-experiment, adoption, implementation science, Kenya", "contribution statement": "This study provides the first methodological synthesis of causal designs in this domain, introducing a quality