{ "background": "Urban primary care networks in sub-Saharan Africa face significant challenges in resource allocation and clinical management due to volatile patient demand and fragmented health data. Existing forecasting tools are often ill-suited to the high-frequency, non-stationary time-series data characteristic of these settings.", "purpose and objectives": "This study aimed to methodologically evaluate a novel hybrid forecasting model for predicting key clinical outcomes, specifically monthly outpatient attendance and antiretroviral therapy (ART) clinic default rates, within Ugandan urban primary care networks.", "methodology": "We conducted an intervention study applying a Seasonal AutoRegressive Integrated Moving Average with eXogenous factors (SARIMAX) model, defined as $\\phi(B)\\Phi(B^s)\\nabla^d\\nablas^D yt = \\theta(B)\\Theta(B^s)\\epsilont + \\beta Xt$, where $X_t$ included rainfall and public holiday indicators. The model was trained on historical administrative data from multiple urban clinics. Forecast accuracy was assessed against a hold-out sample using mean absolute percentage error (MAPE) and evaluated with 95% confidence intervals.", "findings": "The SARIMAX model produced statistically robust forecasts, reducing the MAPE for ART default rate predictions by 18.7 percentage points compared to a naive seasonal benchmark. The 95% confidence intervals for one-month-ahead outpatient attendance forecasts contained the true observed value 92% of the time in the validation period.", "conclusion": "The evaluated time-series model provides a technically sound and context-adapted methodological tool for forecasting clinical outcomes in resource-constrained urban primary care systems.", "recommendations": "Health network managers should integrate such forecasting models into routine monthly planning cycles. Further research should focus on operationalising these models within existing health management information system dashboards.", "key words": "forecasting, health systems, primary health care, time-series analysis, resource-limited settings, clinical operations", "contribution statement": "This paper provides the first application and validation of a SAR