Measuring alarm fatigue in intensive care units: translation and validation of the Charité Alarm Fatigue Questionnaire: a multi-centre study

Résumé fourni par la source

BACKGROUND: Alarm fatigue can negatively impact patient safety and quality of health care, particularly in Intensive Care Units (ICUs). The Charité Alarm Fatigue Questionnaire (CAFQa) is a German questionnaire measuring alarm fatigue in two scales (alarm stress and alarm coping) in nurses and physicians. Its hypothesized dimensionality has not been investigated in non-German-speaking countries. This multi-centre study aimed to translate and validate CAFQa in Dutch-speaking ICUs. METHODS: Forward and backward translation of CAFQa by two bilingual staff members, evaluation by an expert panel, cognitive interviews and a pilot test were performed to obtain the final Dutch questionnaire. Confirmatory factor analysis was performed, Cronbach's alpha (internal homogeneity) and correlation between self-reported alarm fatigue and item scores (convergent validity) were calculated. RESULTS: CAFQa was translated into Dutch, approved by the expert panel and cognitive interviews showed good face validity. Responses from 418 healthcare professionals from 11 centres (10 Dutch, 1 Belgian) were used. Cronbach's alpha was 0.69 (overall), 0.74 (alarm stress scale) and 0.46 (alarm coping scale). Strong correlations of item scores with self-reported alarm fatigue were found (overall: r = 0.51, alarm stress: r = 0.48, alarm coping: r = 0.30), supporting convergent validity. Confirmatory factor analysis indicated an adequate model fit (CFI = 0.961, TLI = 0.946, RMSEA = 0.052), and all factor loadings (0.20-0.74) proved statistically significant, albeit being higher for alarm stress (0.60-0.74) than alarm coping (0.21-0.70). CONCLUSIONS: CAFQa was translated and validated in a Dutch speaking population for the first time. This instrument can benchmark, measure and compare alarm fatigue of nurses and physicians.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.