Construction of an alignment diagram model for predicting calculous obstructive pyonephrosis before PNL

Abstract

Background: To develop a model for the accurate prediction of calculous obstructive pyonephrosis prior to percutaneous nephrolithotomy (PNL), leading to early local anaesthesia microchannel nephrostomy for drainage of pyonephrosis. Methods: By comparing the differences in baseline clinical indicators between the pyonephrosis group and nonpyonephrosis groups, independent risk factors were screened out, and a diagnostic alignment diagram model for predicting calculus obstructive pyonephrosis before PNL was established. Results: Multivariate regression analysis showed that preoperative blood neutrophil count (Neu), serum creatinine level (Scr), serum albumin level (Alb), urine nitrite (UN), hydronephrosis density (HD) and fever history within one month (HFWOM) were independent risk factors for calculous obstructive pyonephrosis. The AUC value of the receiver operating characteristic (ROC) curve was 0.929. The calibration curves showed that the predictive model was well corrected and that the predictive model had strong consistency. Decision analysis curves showed good clinical efficacy of the model. Conclusion: The alignment diagram model accurately predicts patients with preoperative calculous obstructive pyonephrosis in the PNL and provides an evidence-based basis for early renal microchannel nephrostomy.

Keywords

• Pyonephrosis
• Nephrostomy
• Percutaneous nephrolithotomy
• Alignment diagram
• Prediction