This research focuses on the allocation of available hospital resources to different disease types to better serve the needs of patients. The proposed method examines a network of hospitals in a geographical area and determines the proportion of resources that each hospital should devote to different disease classes. The objective is to maximize the customer's utility (specifically, speed of access to medical care) rather than private profit. The approach accounts for the spatio-temporal pattern of disease incidence within the geographic area, and it incorporates the driving distance to a hospital as well as the types of roads in determining travel time. The application focuses on the Greater Los Angeles area, consisting of 525 ZIP codes. Results show a significant shortage of available bed-days for diseases related to socio-economic status, especially in the center-city area.