Model Development and Validation

Firstly, the individual cases with incomplete data will be excluded from the database, and the data of the remaining cases were divided into training set and validation set according to the ratio of 7:3. The training set data were screened by applying univariate analysis, then the risk factors with P-value less than 0.05 were included in the multivariate logistic regression analysis to establish a risk prediction model and nomogram, and finally, the validation set data were applied to evaluate the prediction effect of this model. The performance of the nomogram was evaluated by the discrimination and calibration. Discrimination was measured by the concordance index (C-index), obtained by the area under ROC curve (AUC), and calibration was demonstrated by the calibration curve and evaluated by applying the Hosmer-Lemeshow test. A decision curve analysis (DCA) was also performed to determine the net benefit threshold of prediction.