Two methods for allocating pharmacy cost per patient case.

The consequences of two different methods of allocating pharmacy costs per patient case were studied. The study was conducted using cost data from Sunnybrook Health Science Centre (SHSC), a 1205-bed teaching hospital in Toronto, Canada. A sample of the 1991-1992 cases for the three case mix groups (CMGs) with the highest total pharmacy cost and the three CMGs with the highest pharmacy cost per case were examined. Information was obtained from patient records and used to produce two sets of data: pharmacy prescription unit costs and pharmacy costs per case using the relative value unit (RVU) method, and pharmacy prescription unit costs and pharmacy costs per case using the workload measurement system (WMS) method. For each case, the difference between the RVU and WMS pharmacy costs was determined. The RVU method consistently produced higher pharmacy costs per case for the CMGs with the highest pharmacy cost per case. If these CMGs are typical of other CMGs with high pharmacy costs per case, then case reimbursement based on the WMS method of cost allocation would result in underfunding of hospitals whose case mix has a high proportion of CMGs with high pharmacy costs per case and overfunding of hospitals whose case mix has a high proportion of CMGs with low pharmacy costs per case. However, the RVU method of cost allocation, although it appears to be more accurate, places a greater data collection burden on pharmacy managers. The RVU and WMS methods of pharmacy cost allocation gave significantly different pharmacy costs per case for the six CMGs studied.