USING ARTIFICIAL NEURAL NETWORKS FOR PREDICTING MONTHLY SURGICAL CASE VOLUMES AT GWERU PROVINCIAL HOSPITAL IN ZIMBABWE

Abstract

Modeling and forecasting surgical case volumes can potentially support robust and reliable staff schedules, especially in health facilities such as the Gweru Provincial Hospital (GPH) where patient volumes vary on a monthly basis. This piece of work uses monthly time series data on surgical caseloads at Gweru Provincial Hospital (GPH) from Janaury 2010 to December 2019, to predict surgical cases over the period January 2020 to December 2021. The study applied the ANN (12, 12, 1) model. Residual analysis of this model indicates that the model is stable and therefore suitable for predicting monthly surgical cases at GPH over the out-of-sample period. The results of the study reveal that surgical caseloads for GPH will gradually increase over the out-of-sample period. The study calls for the need for proper monthly staff schedules, especially with regards to improving healthcare service delivery while at the same reducing stress levels among both staff and patients.