Abstract:
Hospital incidence reporting system has limited sensitivity in detecting the adverse events. The previous retrospective double-staged medical record review could overcome underreport problem but has disadvantages of high work load and inconsistency of the tool. Trigger tool might be more practical for detecting adverse events. In order to determine adverse events, internal validity, inter-rater reliability and cost-effectiveness compared with direct physician medical record review, the cross-sectional medical record review to identify adverse events in hospitalized patients by the Global Trigger Tool (GTT) in a tertiary-care university hospital in Bangkok was conducted. A sample of 576 medical records of in-patients discharged during 1-31 Jan. 2008 was selected using a systematic random sampling, and reviewed by a team of in GTT nurses and a team of physicians. The 576 medical records yielded 4.460 patient-days. Totally, 236 adverse events were detected with a mean rate of 41.0 events per 100 patients or 50.4 events per 1,000 patient-days. Sensitivity and specificity of GTT were 57.3 percent and 87.8 percent respectively. Positive and negative predictive values were 62.3 percent and 85.4 percent respectively. Inter-rater reliability between two nurse-reviewer was very good (K = 0.858). Costs of GTT review and the Direc': Physician Review were 593 and 418 Baht/ adverse event consecutively. Specificity and negative predictive value of GTT were high and reliability was very good. The GTT was suitable for using as a diagnostic test for adverse events. However, the GTI sensitivity could be improved and the GTT costs could be reduced by focusing on high risk patients and by revising some triggers to better reflect the context of hospitals in Thailand.