Measuring Safety Culture in Belgian Psychiatric Hospitals: Validation of the Dutch and French Translations of the Hospital Survey on Patient Safety Culture

Abstract

Objectives. To measure safety culture in Belgian psychiatric hospitals on 12 dimensions and to examine the psychometric properties of the Dutch and French translations of the Hospital Survey on Patient Safety Culture (HSPSC) for use in psychiatric hospitals. Methods. The authors analyzed 6,658 completed questionnaires (70.5% response rate) from a baseline measurement (2007-2009) in 44 psychiatric hospitals and 8,353 questionnaires (71.5% response rate) from a follow-up measurement (2011) in 46 psychiatric hospitals. Psychometric properties of the questionnaire were evaluated using item analysis, exploratory factor analysis (EFA), confirmatory factor analysis (CFA), reliability analysis (Cronbach's alpha), and analysis of composite scores and inter-correlations. Results. For both translations, CFA showed an acceptable fit with the original 12-dimensional model. For the Dutch and French translations, EFA showed a 10-factor and a 9-factor optimal measurement model, respectively. Cronbach's alpha indicated an acceptable level of reliability (>=0.70) for 7 of 12 dimensions. Most pair-wise correlations were significant and <0.5, implying good construct validity. Conclusion. The Dutch and French translations of the HSPSC were found tobe valid and reliable for measuring patient safety culture in psychiatric hospitals. Our results also suggest the use of combinations of specific dimensions as recommended in previous research. (Journal of Psychiatric Practice 2015;21:124-139)