Predictive values of trigger tools for identifying adverse events in hospitalised patients using a medical record review: A systematic review

Abstract Background Efforts to identify the prevalence rate of adverse events have been implemented in hospital settings using different methods. The trigger tool method constitutes one option, and involves a retrospective review of paper-based, electronic or hybrid medical records. The aim of the systematic review was to provide a comprehensive description of the predictive value of trigger tools used to identify adverse events in hospitalised patients. Methods A systematic search of MEDLINE, EMBASE, CINAHL, and the Cochrane Library was conducted for studies published between 2000 and October 2024. Eligible studies were peer-reviewed, published in English or Spanish, and reported a trigger tool methodology used to identify the prevalence of adverse events. Two independent reviewers extracted and synthesized the data on study characteristics, methodologies and outcomes. When reported, tool predictive values were pooled by calculating the arithmetic mean across studies. The risk of bias was assessed using the Joanna Briggs Institute critical appraisal checklist for prevalence studies. Results In total, 100 studies from 37 countries were included, 21 high-, 7 upper-middle, 7 lower-middle, and 2 low-income countries. Thirty-four studies reported a predictive value that involved individual triggers (n = 20) and the original tool (n = 14). The Institute for Healthcare Improvement Global Trigger Tool was the most frequent trigger tool used to identify adverse events in hospitalised patients, with an average positive predictive value of 54.5%, negative predictive value of 80.9%, sensitivity of 86.6%, and specificity of 68.2%. An average positive predictive value of 43.9%, negative predictive value of 37.8%, sensitivity of 84.5% and specificity of 11.5% was reported for the Harvard Medical Practice Study. Conclusion Based on the available evidence, the Institute for Healthcare Improvement Global Trigger Tool demonstrates relatively strong predictive values in identifying adverse events in hospitalised patients, with its flexibility and feasibility further supporting its selection as a suitable tool.