MODELING AND FORECASTING PATIENTS WITH CARDIOVASCULAR DISEASE AT ESUT TEACHING HOSPITAL, PARK LANE, ENUGU (2020-2023)

ABSTRACT This study focuses on the time series analysis and forecasting of cardiovascular disease cases at ESUT Teaching Hospital, Park Lane, Enugu, from January 2020 to August 2023. The primary objectives are to evaluate the stationarity of the data, select an appropriate time series model, and forecast future cardiovascular disease cases. A comprehensive analysis was carried out using the Augmented Dickey-Fuller (ADF) test, Autoregressive Moving Average (ARMA) model selection, and diagnostic checks. The data presented a clear spike in cases in March 2022, suggesting a possible anomaly or event driving a significant increase in patient numbers. The ADF test confirmed that the series was stationary in its original form, eliminating the need for differencing. Autocorrelation and partial autocorrelation functions indicated non-randomness in the data, revealing potential autoregressive and moving average components. Two candidate models, ARMA (1,1) and ARMA (1,2) were considered, with the ARMA (1,2) model proving to be the best fit based on diagnostic checks and model significance. The ARMA (1,2) model explained approximately 57.6% of the variability in cardiovascular disease cases. Forecasts from September 2023 to August 2024 suggest periodic fluctuations without a significant upward or downward trend, indicating the hospital can expect continued variability in patient numbers. This study provides valuable insights into the hospital's prevalence of cardiovascular disease cases, enabling better resource allocation and planning. The model's ability to capture short-term patterns highlights its utility for hospital management and forecasting future healthcare demands.