Predicting bed requirement for a hospital using regression models

High hospital bed occupancy levels have resulted into a shortage of beds to meet increasing demand. This paper describes a bed prediction model in aiding hospital planners to anticipate bed demand so as to manage resources efficiently. Through the regression models, it was found that the number of weekly mean occupied beds is related to both the rainfall and the data on Dengue cases as provided by the Ministry of Health. The regression models performed well for predicting average class B2 and class C occupied beds in the following week. Previous week¿s mean occupied beds, emergency admissions numbers, A&E attendances and special events on the week were found to be predictors of bed occupancy in class B2 and class C wards.