Examining the psychometric properties of a crisis recovery in psychosis (CRISP) measure

Background Psychotic crises are complex idiosyncratic experiences and there are no validated measures which examine the recovery from such experiences. The aim of this study was to develop and psychometrically validate the crisis recovery in psychosis (CRISP) measure.Methods The 23-item CRISP was developed with people with lived experience, family/carers and clinicians. A longitudinal design was adopted with 80 participants completing the CRISP along with other relevant measures at three time points. The CRISP was examined for internal consistency, test-retest reliability, criterion validity, construct validity, sensitivity to change and floor/ceiling effects.Results Initial examination of item performance resulted in nine items being removed due to low inter-item correlations (< 0.2). A Principal Axis factor analysis with Direct Oblimin rotation produced a 2-factor solution, including 14 items. The measure was demonstrated to be internally reliable (Cronbach alpha = 0.912 (CI: 0.880–0.938, p < 0.001)) and reproducible (Cronbach alpha = 0.842 (CI: 0.742 to 0.904, p < 0.001)). The measure was deemed to be valid and demonstrated criterion validity, construct validity, and sensitivity to change. No floor or ceiling effects were identified.Discussion The CRISP is a simple, reliable, and valid measure of crisis recovery in people experiencing psychosis. Future research needs to be undertaken to examine its interpretability.