Abstract
ABSTRACT: Patient violence toward others, including doctors and nurses, is a serious concern worldwide. A wealth of literature supports the assertion that violent behavior can be prevented with proper screening and management policies. This project aimed to evaluate the Crisis Triage Rating Scale (CTRS) within a 12-hospital integrated healthcare delivery system located in the southeastern United States. An initial sample of 112,708 unique patient visits between January 2019 and December 2020 was included in this retrospective review of electronic health records. We found that the CTRS harm triage question and risk levels were significant predictors of harm to others. Consistent with previous literature, positive predictive values ranged between 0.025 and 0.070 and negative predictive values ranged between 0.991 and 0.995. Our results support the assertion that clinicians should make balanced judgments about using a positive risk score to allocate safety measures. Variations in practice were evident across our healthcare systems. Improving appropriate assessment procedures may improve the diagnostic tools and risk stratification. When documented correctly, the CTRS performed as expected in an environment where harm to others occurred infrequently.