"background": "District hospital systems in Ghana face persistent challenges in resource allocation and resilience planning. A systematic assessment of methodological approaches for evaluating these systems and forecasting future performance is required to inform evidence-based health policy. ", "purpose and objectives": "This review critically evaluates methodological frameworks used to assess district hospital system performance. Its primary objective is to propose and validate a time-series forecasting model designed to quantify future risk reduction in service delivery metrics. ", "methodology": "A systematic literature review identified and appraised methodological approaches. A novel autoregressive integrated moving average with exogenous variables (ARIMAX) model, yt = \ + =1^{p\ yt-i + =1^q\ -i + =1^r\ Xt-i + \ₜ, was developed and calibrated using historical administrative data. Model robustness was assessed via rolling-origin forecast evaluations. ", "findings": "The review found a predominant reliance on cross-sectional designs, limiting longitudinal insight. The proposed ARIMAX model, incorporating climatic and economic covariates, forecasts a 12–18% reduction in critical drug stock-out incidence over a five-year horizon, with prediction intervals narrowing significantly when spatial dependencies are accounted for. ", "conclusion": "Time-series forecasting provides a superior methodological framework for proactive health system management compared to static evaluations. The integration of environmental and socioeconomic covariates is critical for accurate risk projection. ", "recommendations": "Health planners should adopt integrated time-series models for strategic resource forecasting. Investment in longitudinal data infrastructure is essential to support such analytical approaches and enable pre-emptive interventions. ", "key words": "health systems resilience, forecasting model, ARIMAX, resource allocation, predictive analytics, West Africa", "contribution statement": "This review provides the first validated time-series model specifically configured for forecasting district-level hospital system risks in a resource-constrained setting, demonstrating a move from descriptive evaluation to predictive planning
Mensah et al. (Sun,) studied this question.