The IIASA Health Care Resource Allocation Submodel: DRAM Calibration for Data from the South West Health Region, UK

In many developed countries the problem of allocating resources within the Health Care System (HCS) is perennial. Health care administrators are continually asking what are the consequences of changing the mix of resources. The disaggregated resource allocation model (DRAM) has been developed to assist health care administrators with this problem. The model simulates how the HCS in aggregate allocates limited supplies of resources between competing demands. The principal outputs of the model are the numbers of patients treated in different categories, and the modes and quotas of treatment they receive. Health care planners in the South West Health Region of England are concerned about the consequences for hospital inpatient care of increasing the number of hospital doctors and decreasing the number of hospital beds. This paper indicates how DRAM could be used to assist in the solution of this problem. Parameters were estimated for a model of hospital inpatient care for the region. This model consisted of seven patient categories (general surgery, general medicine, obstetrics and gynaecology, traumatic and orthopaedic surgery, otorhinolarygology, paediatrics and ophthalmology) and two resource types (hospital beds and hospital doctors). The ability with which this model was able to reproduce actual allocations or resources had similarities with a model (of identical structure) of Czechoslovakian hospital in-patient care. It was considered appropriate to reduce the number of patient categories to three (general surgery, general medicine, and obstetrics and gynaecology). Parameters for this three-patient-category model were re-evaluated. Within the assumed predictive accuracy of this model, it successfully predicted health care resource allocation across time and space. The three-patient-category/two-resource type model was then used to explore the consequences of changing the mix of resources in the South West Health Region. Firstly, the consequences of changes from the existing resource levels which involved no estimated increase in running costs were considered. More general changes where this constraint no longer held, were then examined