Methodological Evaluation and Time-Series Forecasting of Clinical Outcomes in Senegalese District Hospital Systems: A Systematic Review

{ "background": "District hospital systems are critical for healthcare delivery in Senegal, yet there is a lack of consolidated evidence on the methodological rigour of studies evaluating their clinical outcomes and the application of forecasting models to inform system management.", "purpose and objectives": "This systematic review aims to critically appraise the methodologies used in studies of clinical outcomes within Senegalese district hospitals and to synthesise evidence on the application and performance of time-series forecasting models in this context.", "methodology": "A systematic search of multiple electronic databases was conducted following PRISMA guidelines. Studies were screened against pre-defined eligibility criteria. Data were extracted on study design, clinical outcomes, forecasting methodologies, and model performance metrics. Quality assessment was performed using appropriate tools for observational and modelling studies.", "findings": "Of the studies utilising forecasting models, a predominant theme was the reliance on autoregressive integrated moving average (ARIMA) frameworks, typically specified as Xt = \μ + \φ1 Xt-1 + \θ1 \εt-1 + \. However, model performance was frequently suboptimal; a key finding was that forecast uncertainty, as indicated by wide prediction intervals, was substantially underestimated in over 60% of applicable studies, compromising their utility for operational planning.", "conclusion": "The methodological quality of existing evaluations is heterogeneous. While forecasting is recognised as a valuable tool, current applications often lack the sophistication and validation required for robust health system decision-making in this setting.", "recommendations": "Future research should prioritise the development and validation of hybrid models that incorporate both clinical and operational covariates. Mandatory reporting of prediction interval coverage probabilities and external validation on held-out data are essential to improve forecasting utility.", "key words": "systematic review, health systems research, forecasting models, clinical outcomes, district hospitals, Senegal, ARIMA, methodological evaluation", "contribution statement": "This review provides the first methodological synthesis and critique of time-series forecasting applications for clinical outcomes in