Modeling and predictionof hypertension in Komfo Anokye Teaching Hospital (KATH), Ghana

Abstract

To explore the temporal trends of hypertension in a Ghana population and to predict future values, which will, in turn, help control and reduce the risk of hypertension-related health events. We enrolled 108,100 cases with essential hypertension from January 2015 to December 2019 at the Komfo Anokye Teaching Hospital (KATH), Ghana. The Box-Jenkins Autoregressive Integrated Moving Average model (ARIMA) was used to identify trends and forecast data from a specified time series. The root mean square error (RMSE), Q-statistic, residual variance (RV), and Akaike’s information criteria (AIC) were used to assess the performance of the model. The most optimal model was ARIMA(1, 1, 0) with RV(7061), RMSE(82.6155), AIC(693.48), Q-value(19.187), parameter(-0.4034) and constant(188.6501). The best fitting model was Yt = (1-0.4034)Yt-1 -0.4034Yt-2 +1801.6670. The model estimated an increase in hypertension cases for the next period, which was a critical input in managerial and administrative decision making. The forecast was accurate enough to allow for better planning and control than could be accomplished without the forecast. Keywords:Hypertension, Forecast, ARIMA, RMSE, RV, AIC