Abstract
Adverse events can cause huge damages to both patients and hospitals financially and socially. Countries are seeking ways to resolve the problems. To do so, the first step of any institutional actions should be the development of procedures and tools to efficiently detect and monitor adverse events.ObjectivesTo study the reliability and validity of the methods for screening and detecting adverse events used by the Harvard Medical Practice Study when applied in ThailandTo estimate incidence, and distributions of adverse events among hospitalized patients in Thailand MethodsThis was a descriptive study. 400 cases were randomly selected from computerized medical records kept in CD-ROM of the Office of Social SecurityA three - step medical record (chart) review by 4 trained nurses and 3 physicians was employed. The procedures were adapted from the methods used by the Harvard Medical Practice Study.ResultsThe agreement of screening process among the nurses and a physician was low (Kappa = 0.34, 95% C.I. of Kappa = 0.24 to 0.43, Sensitivity = 70.23 %, Specificity = 66.27 %, Positive predictive value = 51.69 %, Negative predictive value = 81.25 %).The agreement of the 2 physicians in detecting the occurrence of adverse events was also low (Kappa = 0.393, 95% C.I. of Kappa = 0.25 - 0.53). The incidence of adverse events was 9.1 per cent (95 % C.I. = 6.19 ถึง 11.94 ). 71.4 % of adverse events occurred were considered preventableConclusions The reliability of the procedures to screen and detect adverse events in hospitalized patients applied in the study was low. Therefore, studies to improve the reliability of the procedure are needed.
