Analysis of adverse drug reactions collected by an electronic reporting system in a single hospital

Kim, Min Gang; Kang, Hye Ryun; Kim, Joo Hee; Ju, Young Soo; Park, Sung Hoon; Hwang, Yong Il; Jang, Seung Hun; Kim, Dong Gyu; Jung, Ki Suck

Abstract

BACKGROUND/AIMS
The recent introduction of computerized surveillance systems has promoted the monitoring of adverse drug reactions (ADRs), a feature that facilitates voluntary reports and enables prompt feedback. To investigate the causative agents and severity of ADRs that occurred in a single hospital, we analyzed the features of 980 ADRs that occurred over 14 months after developing a computerized ADR reporting system in Hallym Sacred Heart Hospital.
METHODS
ADR data collected prospectively from September 2007 to October 2008 by a computerized reporting system were analyzed. The World Health Organization-Uppsala Monitoring Center (WHO-UMC) criteria were used to determinate causality for each ADR.
RESULTS
The number of ADR cases reported voluntarily increased rapidly since the introduction of the computerized ADR reporting system. Of the 980 cases, antibiotics (34.5%) were the most common causative drugs, followed by analgesics such as tramadol and its compound (15.2%), radiocontrast media (7.0%), narcotics (5.9%), and nonsteroidal anti-inflammatory drugs (NSAIDs) (5.5%). Fifty-nine (6.0%) and 206 (21.0%) cases were classified as severe and moderate reactions, respectively. The mean age was older in patients with severe ADRs than in patients with non-severe ADRs. The most common clinical features were skin manifestations, such as pruritus, skin eruptions, and urticaria. Gastrointestinal symptoms including nausea, vomiting, and diarrhea were the second most frequently reported ADRs. Among antibiotics, first-generation cephalosporins were the most frequently reported causative drugs, followed by second-generation cephalosporins, penicillin/beta-lactamase inhibitors, and third-generation cephalosporins. While 11.6% of ADRs related to penicillin/beta-lactamase inhibitors were classified as severe, there was only one severe ADR (1.1%) for first-generation cephalosporins. Most ADRs were reported equally in men and women, although female cases constituted about two thirds of ADRs associated with tramadol and NSAIDs.
CONCLUSIONS
We believe that a computerized reporting and replying system promoted the monitoring of ADRs. Antibiotics were reported most frequently as the causative agent of ADRs. Elderly patients seemed to be more susceptible to severe ADRs. With the voluntary reporting system, skin manifestations and gastrointestinal symptoms were detected successfully, while laboratory abnormalities without prominent symptoms seemed to be overlooked. Further efforts to screen for automated ADR signals are required.