약물부작용 감시정보시스템의 도입과 평가: pilot study

Abstract

A computerized surveillance system is the most recent tool to monitor, identify and prevent ADEs, based on alerts generated by matching signals such as abnormal lab data with electronic medical records (EMRs). Since this system has been proven to be simple and cost-effective in detecting ADEs, there is a growing trend towards routine application. The objectives of this study were to implement the system in Samsung Medical Center (SMC), a 1300-bed tertiary hospital, and evaluate its applicability and validity in terms of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The retrospective analysis was performed using 1-month EMRs in 598 discharged patients from 2 ICUs and 5 general wards. Alerts as potential indicators of ADEs were generated by this system using our modified signals. Patient charts were carefully reviewed by clinical pharmacists to detect ADEs and to verify whether these ADEs were also recognized by the signals. A total of 718 alerts were generated by the system and 187 ADEs were detected by chart review. One hundred forty eight alerts were found to be associated with ADEs. The sensitivity of the modified signals for identifying ADEs was estimated to be 79.1 (95% confidence interval [CI], 73.3~85.0) percent, and the estimated specificity was 29.2 (95% CI, 26.1~32.3) percent. The PPV and the NPV were 20.6 (95% CI, 17.7~23.6) percent and 85.8 (95% CI, 81.6~89.9) percent, respec tively. Our computerized surveillance system showed that the sensitivity value was higher (79.1% vs. ca 50%) and the specificity was lower (29.2% vs. ca 70~80%) than those reported in the literature. It would be necessary to further modify the signals to improve the specificity and rate of ADE detection. More studies will be needed to confirm the usefulness and cost-effectiveness of this system.