Effective Analysis of Inpatient Satisfaction: The Random Forest Algorithm

Purpose: To identify the factors influencing inpatient satisfaction by fitting the optimal discriminant model. Patients and Methods: A cross-sectional survey of inpatient satisfaction was conducted with 3888 patients in 16 large public hospitals in Zhejiang Province. Independent variables were screened by single-factor analysis, and the importance of all variables was comprehensively evaluated. The relationship between patients' overall satisfaction and influencing factors was established, the relative risk was evaluated by marginal benefit, and the optimal model was fitted using the receiver operating characteristic curve. Results: Patients' overall satisfaction was 79.73%. The five most influential factors on inpatient satisfaction, in this order, were: patients' right to know, timely nursing response, satisfaction with medical staff service, integrity of medical staff, and accuracy of diagnosis. The prediction accuracy of the random forest model was higher than that of the multiple logistic regression and naive Bayesian models. Conclusion: Inpatient satisfaction is related to healthcare quality, diagnosis, and treatment process. Rapid identification and active improvement of the factors affecting patient satisfaction can reduce public hospital operating costs and improve patient experiences and the efficiency of health resource allocation. Public hospitals should strengthen the exchange of medical information between doctors and patients, shorten waiting time, and improve the level of medical technology, service attitude, and transparency of information disclosure.

Automated summary

This study aimed to identify the factors influencing inpatient satisfaction by fitting the optimal discriminant model. A survey of inpatient satisfaction was conducted with 3888 patients in 16 large public hospitals in Zhejiang Province, and the five most influential factors on inpatient satisfaction were identified as patients' right to know, timely nursing response, satisfaction with medical staff service, integrity of medical staff, and accuracy of diagnosis. The prediction accuracy of the random forest model was found to be higher than that of the multiple logistic regression and naive Bayesian models.