Establishment Of An Individualized Nomogram Model For Predicting The Risk Of SIRS After PCNL

Objective: To investigate the risk factors for the occurrence of systemic inflammatory response syndrome(SIRS)after percutaneous nephrolithotomy(PCNL),and to further establish a reliable and accurate clinical prediction model to provide a basis for early identification of patients at high risk of SIRS after PCNL.Methods: Based on inclusion and exclusion criteria,detailed clinical data were retrospectively collected from 249 patients who underwent PCNL in the Department of Urology,Gansu Provincial People’s Hospital from June 1,2018 to May 31,2022.The group was divided into non-SIRS and SIRS groups based on whether SIRS occurred after surgery.Independent risk factors for the occurs of SIRS after PCNL were screened separately using both single factor analysis-multifactor logistic regression analysis with least absolute value convergence and selection operator(LASSO)regression-multifactor logistic regression analysis,and nomogram model were constructed separately using the corresponding independent risk factors,and the predictive performance of the constructed nomogram model was tested using the discrimination and calibration degree.Distinctness was evaluated using the area under(AUC)the subject operating curve(ROC),and calibration was evaluated using the Hosmer-Lemeshow test and calibration plots.The predictive efficacy of the nomogram model constructed in both ways was compared based on the discrimination and calibration combined,and the model with the better predictive performance was selected as the outcome of this study.The clinical decision curve was used to determine the net clinical benefit of the nomogram prediction model with better predictive performance.Results: A total of 249 patients were included according to the inclusion and exclusion criteria,of which a total of 85 patients developing SIRS after surgery,with an incidence rate of 34.1%.Independent risk factors screened based on univariate analysis-multifactor logistic regression analysis were Guy’s classification,preoperative urine culture,and preoperative SII,and a prediction model was constructed with an AUC=0.741(95% CI: 0.675-0.808),which was tested by the Hosmer-Lemeshow test(χ2= 1.88,P=0.984)and calibration plots showed that the model had a good fit.The independent influences screened based on LASSO regression-multifactor logistics regression analysis were Guy’s classification,preoperative urine culture,HDL cholesterol,and preoperative SII,and a prediction model with AUC=0.749(95% CI: 0.682-0.816)was developed with the HosmerLemeshow test(χ2=7.56,P=0.477)and calibration plots revealed that the model had a better fit than the predictive model established by single factor analysis-multifactor logistic regression analysis.In summary,the prediction model constructed by LASSO regression-multifactor logistic regression analysis had better calibration and discrimination,and the clinical decision curves were plotted,and the results showed that the clinical net benefit curves of the prediction model were far from the two nullity curves,and the net benefit of the prediction model was higher when the threshold probability was 15%-90%.Conclusions: Systemic inflammatory response syndrome is a common complication after percutaneous nephrolithotomy with an incidence of 34.1%.Stone Guy’s classification,preoperative urine culture and preoperative SII were independent risk factors for the development of SIRS after PCNL,while preoperative HDL cholesterol was a protective factor for the development of SIRS after PCNL.The prediction model built based on LASSO regression-multifactor logistic regression analysis had better predictive efficacy than the prediction model constructed based on single factor analysis-multifactor logistic regression analysis with AUC=0.749(95%CI: 0.682-0.816).