Detection of adverse drug events and other treatment outcomes using an electronic prescribing system.

BACKGROUND: Current pharmacosurveillance methods do not provide timely information on drug safety and effectiveness. Real-time surveillance using electronic prescribing systems could address this problem; however, the data collected using these systems has not been validated. We investigated the accuracy of using orders for drug discontinuation and dose change in an electronic prescribing system as a potential source of information for drug safety and effectiveness. OBJECTIVES: To determine the accuracy of an electronic prescribing and drug management system in documenting orders for discontinuation and dose changes of prescription drug treatment, and in identifying the reasons for the drug discontinuation and dose change. STUDY DESIGN AND SETTING: We prospectively assessed the accuracy of electronic prescription orders for drug discontinuation and dose change by comparing them with treatment changes documented by physician-facilitated medical chart review (gold standard). Validity was evaluated in 620 patients of 22 community-based primary care physicians in addition to the reasons for these treatment changes. RESULTS: A total of 141 (41.7%) drug discontinuation orders and 197 (58.3%) changes in drug doses were identified by chart review, the majority of which were for cardiovascular and CNS drugs. Ineffective treatment (30.8%), adjusting dose to optimize treatment (25.1%) and adverse drug reactions (21.9%) were the most common reasons for treatment change. The sensitivity of the electronic prescribing system in identifying physician-initiated drug discontinuations and dose changes was 67.0% (95% CI 54.1, 77.7) and the specificity was 99.7% (95% CI 99.5, 99.9). The positive and negative predictive values of electronic treatment discontinuation and change orders were 97.3% (95% CI 95.6, 98.7) and 95.8% (95% CI 92.9, 97.7), respectively. CONCLUSION: An electronic prescribing and drug management system documents drug discontinuation and dose-change orders with high specificity and moderate sensitivity. Ineffective treatment, dose optimization and adverse drug reactions were the most common reasons for drug discontinuation or dose changes. The electronic prescribing system offers a new method for augmenting pharmacosurveillance.