Forecasting Yield Improvement in Public Health Surveillance Systems Using Time-Series Models in Uganda: A Methodological Evaluation

Public health surveillance systems in Uganda are crucial for monitoring disease prevalence, but their efficiency can be improved through data-driven methods. The study utilised ARIMA (AutoRegressive Integrated Moving Average) models for forecasting yield improvements, with real-time surveillance data from Uganda's National Health Information System as the primary input. Robust standard errors were employed to account for prediction uncertainties. An initial forecast model showed a positive direction of improvement in disease surveillance metrics but exhibited moderate uncertainty (95% confidence interval: -0.12% to +0.45%). The ARIMA models demonstrated potential as an analytical tool for enhancing public health surveillance systems, warranting further empirical validation. Further research should include a wider range of diseases and incorporate additional variables such as socio-economic factors to improve model accuracy. Public Health Surveillance, Time-Series Forecasting, ARIMA Models, Uganda Treatment effect was estimated with $\text{logit}(p_i)=\beta_0+\beta^\top X_i$, and uncertainty reported using confidence-interval based inference.