Quantitative and qualitative estimation of patient safety culture in private and public hospitals using machine learning methods

Abstract

Abstract Bacground: Patient safety is a critical component of health care systems and a worldwide concern. So that, many patients are at risk of harm as a result of medical errors. The necessity of maintaining patient safety is to create and promote a patient safety culture, and in order to do so, the patient safety culture must be evaluated. Therefore, in this study, the patient safety culture in the country's hospitals (both public and private) has been evaluated. Materials and methods: In this study, patient safety culture was evaluated using the HSOPSC questionnaire. To estimate the sample size, cluster and then stratified sampling methods were used, and the sample size was 767 people. Then the neural network algorithm and decision tree algorithm were used to evaluate qualitative data, and random forest algorithm and linear regression were used for quantitative data evaluation. SPSS software was used for statistical analysis and, Orange Data Mining version 3 was used for quantitative and qualitative data analysis. Results: The score of safety culture in public and private hospitals was 41.99% and 40.96%, respectively, and they did not differ significantly from each other in terms of the effective dimensions of safety culture. In general, the average safety culture in public hospitals was higher than in private hospitals. Also, patient safety culture has a direct relationship with education level, work experience, gender, income, and organizational position. Data analysis with data mining algorithms such as decision tree and neural network showed that the dimensions that have the most impact on patient safety culture, in order of priority, include: dimension of feedback and communication about errors, organizational learning and continuous improvement, management support for patient safety. Among the demographic characteristics, age, work experience, income, education level, and organizational position are among the factors influencing patient safety culture. Conclusion: According to the results, the examined hospitals have a weak patient safety culture. Therefore, it is suggested to improve patient safety culture in hospitals through training and theallocation of resources, considering the significant differences observed between the variables of work experience, education, and organizational position. Considering the importance of patient safety culture in preventing accidents and reducing injuries, the results of the present study and the presented models can be used to predict patient safety culture in hospitals. Considering the limitations of the present study, such as the study area being limited to the city of Tabriz and the time of the Corona epidemic and the effect of these factors on the collected data, more research is recommended in the field of patient safety culture. Keywords: Patient safety, public, private, HSOPSC (Hospital Survey on Patient Safety Culture), Orange Data Mining, Neural network algorithm, decision tree algorithm, feature regression