A Methodological Evaluation of District Hospital Performance in Senegal: A Panel-Data Analysis of Clinical Outcomes, 2000–2024

{ "background": "District hospitals are critical nodes in national health systems, yet robust, longitudinal evaluations of their clinical performance in sub-Saharan Africa are methodologically limited. Existing assessments often rely on cross-sectional data or process indicators, failing to capture dynamic changes in outcomes.", "purpose and objectives": "This study aims to develop and apply a panel-data econometric framework to evaluate the clinical performance of district hospitals over an extended period, focusing on mortality outcomes as a core metric.", "methodology": "We constructed a novel, longitudinal dataset from national health information systems. A two-way fixed effects model was estimated: Yit = \α + \β1 Xit + \ + \ + \εit, where Yit is the risk-adjusted mortality rate for hospital i in year t, Xit is a vector of time-varying inputs (e.g., staffing, drug availability), \ are hospital fixed effects, and \ are year fixed effects. Inference is based on heteroskedasticity-robust standard errors clustered at the hospital level.", "findings": "A one standard deviation increase in the composite resource index was associated with a 7.2% reduction in risk-adjusted mortality (95% CI: -10.1% to -4.3%). Performance trajectories diverged significantly after a major financing reform, with hospitals in the highest quartile of baseline management quality showing markedly steeper improvements.", "conclusion": "The panel-data approach reveals significant, previously unobserved heterogeneity in hospital performance driven by both resource inputs and institutional factors. Clinical outcomes are malleable to policy interventions over time.", "recommendations": "National health policy should integrate panel-data evaluation for routine performance monitoring. Resource allocation mechanisms should be refined to target persistently low-performing institutions, coupled with investments in hospital management capabilities.", "key words": "health systems evaluation, panel data, hospital performance, fixed effects models, health econom