Cross-cultural adaptation and validation of hospital survey on patient safety culture in Bengali version (B-HSOPSC 2.0) among nurses in Bangladesh: a cross-sectional study

BACKGROUND: The Hospital Survey on Patient Safety Culture (HSOPSC 2.0) was developed and updated by the Agency for Healthcare Research and Quality in 2019. It has now been widely adopted and translated into different languages worldwide. However, the validity and reliability of the Bengali version of HSOPSC 2.0 (B-HSOPSC 2.0) have not been tested among healthcare professionals. This study aimed to determine the validity and reliability of the B-HSOPSC 2.0 with cross-cultural adaptations, among hospital nurses in Bangladesh. METHODS: The study was conducted among nurses in eight tertiary-level government medical college hospitals in Bangladesh. A two-step study design was employed, encompassing the translation, cultural adaptation, and psychometric evaluation of B-HSOPSC 2.0. The translation process included forward and backward translation by panel, expert consensus, review, and pretesting. Content validity, reliability, and test-retest reliability were assessed. Construct validity was evaluated through confirmatory factor analysis (CFA), and convergent validity was examined using average variance extraction and Spearman's correlation coefficient. RESULTS: Out of 7,170 eligible nurses, 4,982 were analyzed (valid response rate: 69.5%), and a subset (n = 424) provided retest responses. The words "manager" and "clinical leader" were removed, and the word "units" was replaced with "wards," as they were deemed inappropriate for the Bangladesh healthcare system. The content validity index provides strong evidence of adequacy for the instrument measurements (I-CVI = 0.83-1.00, S-CVI = 0.98). B-HSOPSC 2.0 demonstrated a good internal consistency (Cronbach's α = 0.70-0.76). The test-retest reliability was acceptable at the group level (ICC = 0.65-0.76) and low in the single measure (0.071, 0.069). In the CFA model, the indices for the 10 and 9 dimensions were CFI = 0.79, 0.83; NFI = 0.78, 0.82; TLI = 0.74, 0.79; GFI = 0.91, 0.92; RMSEA = 0.06, 0.06; SRMR = 0.07, 0.06, respectively. CONCLUSIONS: The psychometric properties of the Bengali version were inconsistent and not sufficiently robust. However, content validity and test-retest reliability were deemed acceptable. Since there is no validated tool to evaluate the patient safety culture in the Bengali version, these study findings may help further to evaluate future nurses' perceived patient safety culture interventions in hospital settings in Bangladesh. Further studies are required, as the psychometric properties of B-HSOPSC 2.0 among other healthcare professionals in Bangladesh remain to be confirmed. CLINICAL TRIAL NUMBER: Not applicable.