Methodological Evaluation and Yield Optimisation of Public Health Surveillance Systems in Tanzania: A Bayesian Hierarchical Meta-Analysis, 2000–2026

{ "background": "Public health surveillance is critical for disease control, yet the methodological rigour and yield of such systems in sub-Saharan Africa require systematic evaluation. In Tanzania, diverse surveillance modalities have been implemented, but a consolidated quantitative assessment of their performance and determinants of yield is lacking.", "purpose and objectives": "This meta-analysis aims to methodologically evaluate the performance of public health surveillance systems in Tanzania and to identify key factors optimising case detection yield, using a Bayesian hierarchical framework.", "methodology": "We conducted a systematic review and meta-analysis of studies reporting surveillance system outputs. A Bayesian hierarchical model was fitted to estimate pooled yield and study heterogeneity. The core model was yi \~ (\, \σ²i), \ \~ (\μ, \τ²), where $yi$ are log-transformed yield measures, \ are study-specific means, \μ is the overall mean log-yield, and \τ² represents study variance. Prior distributions were weakly informative. Inference was based on posterior distributions with 95% credible intervals (CrI).", "findings": "The pooled analysis indicated that integrated community-based surveillance significantly outperformed facility-based reporting, with a median yield ratio of 1.82 (95% CrI: 1.45 to 2.24). System yield was positively associated with the use of digital reporting tools and regular supervisory feedback, accounting for approximately 40% of the study variance.", "conclusion": "Surveillance yield in Tanzania is substantially enhanced by community integration and technological support. Methodological heterogeneity across studies remains a challenge for comparative evaluation.", "recommendations": "National policy should prioritise investment in integrated, community-centred surveillance architectures with robust digital infrastructure. Future research must adopt standardised reporting metrics to facilitate cross-system comparison.", "key words": "disease surveillance, health systems, Bayesian meta-analysis, yield optimisation