"background": "Evaluating the performance of community health centres is critical for improving health systems in sub-Saharan Africa. Panel-data econometric methods offer powerful tools for analysing longitudinal data on facility performance, yet their application and methodological rigour in this specific context require systematic assessment. ", "purpose and objectives": "This review critically evaluates the methodological approaches used in panel-data estimations to measure health systems performance and agricultural yield improvements within Tanzanian community health centres. It aims to assess model specification, variable selection, and the handling of common panel-data challenges. ", "methodology": "A systematic search identified relevant studies employing panel-data methods. Methodological quality was appraised against established econometric principles, focusing on model choice (e. g. , fixed vs. random effects), control for unobserved heterogeneity, and treatment of endogeneity. The core model evaluated is Y{it = \ + \ Xit + \ +, where \ᵢ denotes entity-specific effects. ", "findings": "A key finding is that studies employing fixed-effects models with cluster-robust standard errors produced more reliable inferences, with one major theme being the underreporting of diagnostic tests for serial correlation. A concrete methodological result is that nearly 40% of reviewed studies failed to account for potential endogeneity from time-varying confounders, threatening causal claims. ", "conclusion": "While panel-data methods are increasingly applied, significant methodological shortcomings persist, limiting the robustness and policy relevance of findings on health centre performance and its linkages to local food yields. ", "recommendations": "Future research must prioritise rigorous model diagnostics, explicitly test for and address endogeneity using instrumental variable or generalised method of moments techniques, and improve transparency in reporting estimation procedures and uncertainty metrics. ", "key words": "health systems research, panel data, econometrics, fixed effects, community health, agricultural yield, Tanzania", "contribution statement": "This review provides the first dedicated methodological critique of panel-data applications in this interdisciplinary field,
Mwinyi et al. (Tue,) studied this question.