| Retrospective Drug Utilization Review (RDUR) programs use computerized administrative data to intervene on apparently problematic prescribing. Interventions occur after the prescription has been dispensed. Such programs are required of all US Medicaid programs and are commonly used by private insurers. We studied the effectiveness of Medicaid RDUR programs in reducing the frequency of apparent drug therapy problems (exceptions) and in improving clinical outcomes in patients with exceptions.; Using pre- and post-implementation data, we looked for an effect of RDUR on the frequency of exceptions. We were unable to detect an effect, even in high dose subgroups.; We also examined whether drug therapy problems occurring after RDUR implementation resulted in hospital admission less often than those occurring pre-implementation. Again, we were unable to discern an effect of RDUR, including in high dose and high risk subgroups.; Study programs issued alerts (usually a letter to the prescriber) for 1% to 25% of drug therapy problems.; Our results suggest that RDUR programs, as currently implemented, are ineffective in affecting future prescribing or reducing the risk of hospitalization among those with an exception. A number of factors may be responsible, including the low alert rate, the questionable validity of the criteria, and the inefficacy of retrospectively mailed alert letters to improve prescribing. Given the lack of evidence that RDUR programs, as implemented, are effective in improving clinical outcomes, and the potential for unintended consequences of such programs, policymakers should consider withdrawing the mandate for such programs. Improved versions of RDUR should be shown to be effective before being implemented. An agenda for future research on improving medication use is presented. |
