Physicians' and pharmacists' acceptance of online, real-time "drug combinations to avoid" messages

Background: Computerized physician order entry (CPOE) systems combined with a clinical decision-support system (CDSS) can prevent potential drug-therapy problems; however, medication-related decision- support alerts with low specificity often generate alert fatigues and are ignored by health-care providers. The computerized prospective drug utilization review (DUR) program supported by the Korean government has been providing "drug combinations to avoid (DCA)" alerts to physicians and pharmacists since December 2010. Objectives: The aim of this study is to explore providers' responses to DCA alerts and identify the characteristics of institutions, providers, and drug components associated with providers' acceptance of online, real-time DCA messages. Methods: A retrospective analysis was conducted on the DUR prescription-transfer data from April 1, 2011 to March 31, 2012 in South Korea. The frequency of DCA alerts and provider acceptance was observed in two ways: "within-prescription review" and "between-prescriptions review". Multivariate regression was used to determine the predictors of acceptance. Results: The overall acceptance rate was 37.9%, which was relatively high compared with the results from the previous studies with regular drug-drug interaction (DDI) alerts. The acceptance rates varied depending on the origin and types of DCA. The types and location of institutions and severity of DCA were positive predictors of alert acceptance. Conclusions: To improve providers' adaptation to DCA alerts, various strategies, including tiering, applying human factors, and eliminating inappropriate alerts should be considered.