Psychometric properties of the Hospital Survey on Patient Safety Culture for hospital management (HSOPS_M)

Abstract

Background From a management perspective, it is necessary to examine how a hospital's top management assess the patient safety culture in their organisation. This study examines whether the Hospital Survey on Patient Safety Culture for hospital management (HSOPS_M) has the same psychometric properties as the HSOPS for hospital employees does. Methods In 2008, a questionnaire survey including the HSOPS_M was conducted with 1,224 medical directors from German hospitals. When assessing the psychometric properties, we performed a confirmatory factor analysis (CFA). Additionally, we proved construct validity and internal consistency. Results A total of 551 medical directors returned the questionnaire. The results of the CFA suggested a satisfactory global data fit. The indices of local fit indicated a good, but not satisfactory convergent validity. Analyses of construct validity indicated that not all safety culture dimensions were readily distinguishable. However, Cronbach's alpha indicated that the dimensions had an acceptable level of reliability. Conclusion The analyses of the psychometric properties of the HSOPS_M resulted in reasonably good levels of property values. Although the set of dimensions within the HSOPS_M needs further scale refinement, the questionnaire covers a broad range of sub-dimensions and supplies important information on safety culture. The HSOPS_M, therefore, is eligible to measure safety culture from the hospital management's points of view and could be used in nationwide hospital surveys to make inter-organisational comparisons.