Using a clinical data repository to estimate the frequency and costs of adverse drug events

As a result of increased attention to medical errors, many institutions are contemplating increased use of information technology and clinical decision support. We conducted a retrospective analysis to estimate the frequency and cost of adverse drug events (ADEs) for inpatients at the University of Virginia. Applying published criteria for the detection of potential adverse events, we used a clinical data warehouse to identify patients and cases with potential ADEs. Again using published criteria, we then estimated the actual number of adverse drug events and preventable adverse drug events, as well as their attributable costs and excess length of stay. Our results showed a higher estimate (10.4-11.5 events per 100 admissions) for ADEs than seen in the ADE Prevention Study, highlighting the importance of considering the generalizability of published ADE studies to other settings. Our analysis demonstrates that retrospective analysis can be an efficient and powerful technique to evaluate rules and criteria used to detect ADEs and to assess their impact.