Interhospital variations in hospital pharmacy markups.

The relationship between pharmacy pricing policies and overall hospital objectives was analyzed for 64 South Carolina hospitals in 1980. Using published hospital data, a linear regression model was developed to test the relationship between a pricing variable (the charge-to-cost ratio for each pharmacy) and five independent variables: demand (defined as hospital occupancy rate), industry structure (defined as affiliation or nonaffiliation with a multihospital system), level of hospital use by Medicare and Medicaid patients (defined as the proportion of hospital admissions for which reimbursement was based on costs), the contribution of nonoperating revenue, or subsidization (defined as the proportion of hospital nonoperating revenue to total expenses), and the contribution of operating revenues of departments other than pharmacy (defined as the proportion of non-pharmacy charges to total charges). In these predominantly nonprofit hospitals, the mean (+/- standard deviation) pharmacy markup was 1.99 +/- 0.49, with considerable inter-hospital variation. The level of hospital use by Medicare and Medicaid patients had the greatest influence on variation in markup, indicating that hospitals were responding to cost-based payer reimbursement practices by raising charges in areas with a high cost base, such as pharmacy. Nonoperating revenue and operating revenue of departments other than pharmacy also were significantly related to the charge-to-cost ratio. The relationship between the charge-to-cost ratio and either occupancy rate or affiliation with a multihospital system was not significant. The hospitals studied set pharmacy revenues to contribute to overall target income, but pharmacy prices were not set to achieve maximum profits by responding to changes in demand.