Abstract
Background：COVID-19 is a rapidly spreading disease with high rates of infectivity, morbidity, and fatality, Nurses face heightened risks of infection since China published full liberalization policy .
Aim：To pinpoint the specific risk factors associated with depression among Chinese nurses during the comprehensive liberalization phase of the COVID-19 pandemic in 2022 and to formulate a predictive model for risk assessment.
Methods:a cross-sectional study from December 9, 2022, to March 26, 2023, recruiting 293 nurses from a tertiary hospital in Anhui Province. Participants were categorized into depression and without depression. The data of the two groups were analyzed using SPSS 23.0. Four predictive machine learning models—Logistic Regression, Support Vector Machine, Extreme Gradient Boosting Machine, and Adaptive Boosting —were developed.
Results：The AUC for the Logistic Regression, SVM, XGBoost, and AdaBoost models were 0.86, 0.88, 0.95, and 0.93 respectively, while their F1 scores were 0.79, 0.83, 0.90, and 0.89. The XGBoost model demonstrated the highest predictive accuracy. The Extreme Gradient Boosting Machine model, tailored to risk factors prevalent among Chinese nurses, offers a potent tool for predicting depression risks.
Conclusions：This model can aid clinical managers in accurately identifying and addressing potential risk factors during and post the comprehensive liberalization phase of the COVID-19 pandemic.