Safety Culture Data.2024
Description
Background: The aim of this study is to analyze the temporal trend of patient safety culture within a network of private hospitals in Brazil, while also evaluating the relationship between safety culture and the hospital accreditation process. Method: A retrospective study was carried out through the application of the Hospital Survey on Patient Safety Culture (HSOPSC) questionnaire and distributed to 68 hospitals from 2014 to 2022 with a total of 259,268 responders. The analyzed hospitals were classified as accredited (AH) or non-accredited (NAH). Findings: Eleven dimensions of HSOPSC significantly improved over an 8-year period, except for the frequency of reported events, which remained unchanged over time. Two dimensions had less than 50% of positivity “communication openness” (47.13% [38.19 - 58.73]) and “nonpunitive response to errors” (41.24% [34.13 - 51.98]). AH demonstrated an increase in the safety culture across 11 dimensions, except for the frequency of reported events, which remained unchanged. In NAH, the frequency of reported events decreased over time and the remaining dimensions showed no significant change. Interpretation: These results suggest that patient safety culture improved in this network of private hospitals from 2013 to 2022. Accreditation has been a tool that has fostered a culture of safety over time. Non-accredited hospitals tended to decrease the number of reported adverse events, which suggests underreporting and missed opportunities for healthcare system improvement through adverse event analysis.
Files
Steps to reproduce
Statistical analysis The data were entered into Excel spreadsheets using a double-entry method, and a comprehensive review was conducted to detect and rectify any discrepancies. Descriptive statistics were used to present the frequencies of sample characteristics and patient safety climate factors. Negatively worded items were inverted to ensure that positive responses indicated a higher score. Consequently, a higher score signified a more positive attitude among healthcare workers towards patient safety culture. A linear mixed-effects model regression was employed to fit the trend in dimension score over time. Differences between groups (accredited hospitals vs. non-accredited hospitals) in dimension scores are presented as point estimates along with 95% confidence intervals. The magnitude of the effects was defined by the Hodges-Lehmann method to estimate the median difference (MD) and its 95% confidence intervals (CI) between the groups. Scores for accredited hospitals and non-hospitals were compared using the Mann-Whitney test. Dimension scores for hospitals that transitioned from non-accredited to accredited status during the historical series were analyzed in a paired manner using the Wilcoxon rank-sum test. Categories variables were compared using the χ2 test. All analyses were conducted using R software, version 4.1.2 (R Foundation for Statistical Computing). Statistically significant results were indicated by P values less than 0.05.