drugs or drug classes, and volume or quantity limits. Basic formulary and formulary system limitations of drugs in health care systems are identified. New data on restrictions in private health care formularies and formulary systems were collected for the IOM by the Academy of Managed Care Pharmacy. The VA National Formulary is not as inclusive as many Medicaid formularies, but it rarely designates drugs or drug classes as absolutely excluded or requires prior authorization of drugs as Medicaid and managed care formulary systems frequently do, nor does it impose absolute limits on numbers of prescriptions as some Medicaid formularies have done, or tiered copayments as is often the practice in managed care. The IOM committee concluded that the VA National Formulary is not overly restrictive. In some respects it is more, but in many respects less, restrictive than other public or private formularies. The committee has identified deficiencies in the implementation and management of the National Formulary and recommended changes.

The IOM examined VA aggregate outpatient drag use data by VISN by month for FY 1994 through FY 1999 for six drug classes that had been closed or preferred at some point by the National Formulary (omitting the luteinizing hormone-releasing hormone [LHRH] class for which data were lacking) and for eight other classes that had remained open. Person-level data were not available to support comparisons of spending for inpatients, outpatients, and pharmacy, or analyses of cost shifting and spending in various budgets associated with implementation of the National Formulary. With some gaps, average VA drug purchase price data were available by VISN by month. To arrive at expenditures for each drug, these prices were multiplied by units used. Using other data on total system users, age and other demographic factors, as well as various statistical and analytical techniques, attempts were made, within the limits of the data provided, to control for variables that might distort estimates of VHA pharmacy expenditures. For 2 years before and after implementation of the National Formulary, VA inpatient discharges for diagnoses of conditions likely to be affected by changes in treatment with drugs in closed or preferred classes were examined. The IOM assessment of cost savings associated with the VA National Formulary was limited to six drug classes and was very conservative. A recent, higher VA estimate was based on a longer time and inclusion of a wider scope and number of activities and policies. The IOM committee found that the VA National Formulary was associated with substantial changes in utilization, prices, and market share of drugs in closed and preferred classes compared to drugs in open classes. Savings in pharmacy expenditures approaching $100 million over the approximately 2-year time span since formulary implementation have probably been realized. This figure is about 15% of the six analyzed drug class expenditures over those 2 years. An exploratory analysis of the distribution of inpatient discharges for illnesses treatable by two closed drug classes did not