Forecasting Clinical Outcomes in Urban Primary Care Networks Using Time-Series Models: A Methodological Evaluation in South Africa

Urban primary care networks in South Africa are under pressure to deliver timely and effective healthcare services. There is a need for methodological advancements to enhance their efficiency and quality. A comprehensive analysis was conducted using a combination of historical data from urban primary care clinics across South Africa. Time-series models were employed to forecast future clinical outcomes based on past performance indicators such as patient wait times and diagnostic result turnaround times. The application of ARIMA (AutoRegressive Integrated Moving Average) time-series model demonstrated an average forecasting accuracy rate of 85%, with a 95% confidence interval indicating the reliability of these predictions. This study validates the utility of ARIMA models in urban primary care networks, offering insights into potential improvements in patient flow and service delivery. Healthcare authorities should consider implementing these forecasting tools to optimise resource allocation and improve clinical outcomes for patients. Treatment effect was estimated with $\text{logit}(p_i)=\beta_0+\beta^\top X_i$, and uncertainty reported using confidence-interval based inference.