Health professionals’ perceptions of patient safety competencies: psychometric properties of the French version of the H-PEPSS in France and Switzerland

Abstract

BackgroundSeveral initiatives have been implemented to develop, manage, and assess patient safety (PS) competencies, which are considered as a serious public health issue across the world. The Health Professional Education in Patient Safety Survey (H-PEPSS) is widely used as a psychometric scale for evaluating perceived PS competencies but has not been validated in French. The purpose of the study was to investigate the main psychometric properties of the French version of the H-PEPSS.MethodsA total of 449 students enrolled in nursing and physiotherapy schools in France and French-speaking Switzerland completed a self-administered questionnaire. The 38 items of the H-PEPSS were translated into French following a committee approach. The scale’s construct validity was assessed using confirmatory factor analysis. Reliability of the six factors of the H-PEPSS was evaluated using Cronbach α and McDonald’s ω. Measurement invariance across countries and academic majors as well as discriminant validity were also investigated.ResultsAfter we removed one item, the H-PEPSS 6-factor model demonstrated adequate goodness-of-fit statistics (χ2[194] = 316.633, χ2/df = 1.632, p < 0.001, CFI = 0.934, TLI = 0.922, RMSEA = 0.041 [0.033, 0.049], SRMR = 0.044). The total score can be also used as an overall measure of PS competence (χ2[203] = 342.251, χ2/df = 1.686, p < 0.001, CFI = 0.925, TLI = 0.915, RMSEA = 0.043 [0.035, 0.051], SRMR = 0.047). One item was removed because of its high multicollinearity with other items. The reliability was deemed satisfactory (Cronbach α ≥ 0.60), except for the “Understanding human and environmental factors” subscale. Consistently, this subscale was often reported with the lowest reliability in previous studies. We confirmed scalar invariance between countries and partial scalar invariance between majors (ΔCFI ≤ 0.01). The heterotrait-monotrait ratio of correlations ranged from 0.63 to 0.91. In our results, country, academic year, and academic satisfaction were frequently the main predictors of self-reported PS competencies.ConclusionPerceived PS competencies can be assessed and fairly compared across France and Switzerland and across nursing and physiotherapy students. We discuss the relevance of the introduction of the H-PEPSS in the training pathway of health professions degree courses and the fallout in clinical contexts.