ABSTRACT
BACKGROUND: Medication-related harm can be detected using the adverse drug event (ADE) trigger tool and the medication module of the Global Trigger Tool (GTT) developed by the Institute for Healthcare Improvement (IHI). In recent years, there has been some controversy on the performance of this method. In addition, there are limited data on the performance of the medication module of the GTT as compared with the ADE trigger tool. OBJECTIVES: To evaluate the performance of the ADE trigger tool and of the medication module of the GTT for identifying ADEs. METHODS: The methodology of the IHI was used. A random sample of 20 adult admissions per month was selected over a 12-month period in a teaching hospital in Belgium. The ADE trigger tool was adapted to the Belgian setting and included 20 triggers. The positive predictive value (PPV) of each trigger was calculated, as well as the proportion of ADEs that would have been identified with the medication module of the GTT as compared with the ADE trigger tool. RESULTS: A total of 200 triggers and 62 ADEs were found, representing 26 ADEs/100 admissions. Nineteen ADEs (31%) were found spontaneously without the presence of a trigger. Three triggers never occurred. The PPVs of other triggers varied from 0 to 0.67, with half of them having PPVs less than 0.20. If we had used the medication triggers included in the GTT (n = 11), we would have identified 77% of total ADEs and 67% of preventable ADEs. CONCLUSIONS: Applying the trigger tool method proposed by the IHI to a Belgian hospital led to the identification of one ADE out of 4 admissions. To increase performance, refining the list of triggers in the ADE trigger tool and in the medication module of the GTT would be needed. Recording nontriggered events should be encouraged.
