Development of an Effective Approach to Model the Non-Linear Causal Relationship Between Key Control Factors and Patient Arrival at an Emergency Department of Hospital

Description

There is a strong and increasing demand for enhancement of service level in healthcare industry. In an Emergency Department (ED) of a hospital, the hospital authority has laid down performance indicators to measure the waiting time of patients. In order to determine the optimum staffing level to meet those indicators, an accurate forecast on patient arrival is critical. In addition to temporal factors (e.g. hour, day, etc), non-temporal factors (e.g. temperature, rainfall, etc) must be considered in predicting the patient arrival pattern. Use of linear methods to forecast patient arrival is common but there is a lack of non-linear approaches in the current forecasting. This project aims at filling this gap by development of an effective approach to model the non-linear causal relationship between key factors and patient arrival at an ED. A new modeling technique, belief rule-based methodology, is investigated to model such non-linear causal relationship in ED.