Criteria to Request Pharmacoeconomic Data and Data Sources for Hospital Formulary Decisions

A national mail survey was designed to study how pharmacoeconomic (PE) data were applied to formulary decisions in teaching hospitals in the United States. The study objectives were to determine the criteria for using PE data for formulary decisions, to describe PE data sources that were used, and to assess the opinions of pharmacy directors regarding the quality of available PE data.

One hundred and sixty-six (39.6%) pharmacy directors responded to the survey from April to July 1999. There were no significant differences between respondents and non-respondents with regard to hospital demographics. One hundred and thirty eight hospitals (83.1%) reported using PE data. The two most important criteria leading to a PE analysis were the impact that a new medication may have on institutional cost, and the anticipated annual pharmacy budget for a new medication (the threshold of potential cost was concentrated between $25,000 and $100,000 per medication). Classes of drugs that frequently “ranked first” in requiring PE data were antiplatelets, systemic anti-infective drugs, and antineoplastics. Drugs that required PE data were those which were recently approved by the Food and Drug Administration (FDA) and that were relatively expensive (eg, Glycoprotein IIb/IIIa inhibitor; Cox-II inhibitor; Low molecular weight heparin). The most important sources of PE data were published literature, pharmaceutical companies, and in-house PE assessments. Hospitals that could identify greater cost-savings used more in-house PE data. Seventy-six percent of pharmacy directors had a moderate to high level of skepticism of PE data derived from PE modeling and 79% of them felt that the use of PE data might expand if the FDA review process incorporated approval of such data.