Poster 32

[P-32-06] A Standardization Study on the Development and Validation of a Shortened Korean Version of the Psychiatric Crisis Instrument for Psychiatric intervention

*HEEWOO LEE¹, Jee hoon Sohn², Kihoon You³, Seung-Hee Ahn², Jeung Suk Lim⁴, Hye-Young Min⁵ (1.Department of Psychiatry, Kangwon National University Hospital (Korea), 2.Public healthcare center Seoul national univ hospita(Korea), 3.School of Medicine, Kyungpook National University(Korea), 4.Happy Care Together, Seoul(Korea), 5.Department of Social Welfare, Sogang University(Korea))

Keywords:

Psychiatric Crisis,community,Scale

BackgroundThe Crisis Rating Instrument for Psychiatric intervention (CRI) is used nationwide in Korea for psychiatric crisis assessment but is limited by its 23-item length and narrow focus on risk of harm to self or others. These constraints hinder rapid decision-making in emergency settings. ObjectiveThis study aimed to develop and validate a shortened version (CRI-SF) to facilitate swift and consistent crisis intervention. MethodsData from 2,774 CRI assessments collected between 2020 and 2023 were reviewed. Inclusion criteria required documented field intervention and classification as treatment linkage or emergency response. Through three sequential evaluation phases—including confirmatory factor analysis (CFA), content validity assessment, and classification performance analysis—the scale was reduced. Items were retained based on factor loadings, content validity indices, clinical relevance, and discriminative performance (sensitivity, specificity, AUC, Youden’s index). ResultsInitially, six items were identified as most relevant, later refined to five: current risk of harm to self or others, aggression associated with hallucinations or substance use, psychiatric symptoms, cognitive impairment, and presence of a support system. Items were weighted to reflect their criticality, with risk-related items assigned higher scores. The final CRI-SF showed excellent model fit (CFI = 0.970, TLI = 0.939, RMSEA = 0.037). The risk of harm item demonstrated the strongest classification accuracy (AUC = 0.934, sensitivity = 100%, specificity = 86.9%). ConclusionsAlthough the study lacked external validation measures and retained the dichotomous scoring format, the CRI-SF enables efficient identification of psychiatric emergencies while maintaining compatibility with existing data. This abbreviated scale supports rapid assessment and decision-making in community mental health crisis response.