Construction and validation of a multifactorial risk prediction model for postoperative incision infection by retro-peritoneal laparoscopy
WANG Xian
WANG Xin-peng
GUI Hong-wei
LI Xiang
Abstract：Objective To identify risk factors associated with postoperative incision infection following retroper-itoneal laparoscopic partial nephrectomy(RLPN)and to develop an individualized risk prediction model.Methods Clinical data of 200 patients with clear cell renal cell carcinoma who underwent RLPN between February 2019 and August 2024 were retrospectively analyzed.Patients were divided into an infection group(n=40)and a non-infection group(n=160).Univariate and multivariate logistic regression analyses were conducted to identify independent predictors of in-cision infection.A nomogram was constructed using R software,and its discrimination,calibration,and clinical utility were evaluated using receiver operating characteristic(ROC)curves,calibration plots,and decision curve analysis(DCA).An external validation cohort of 142 RLPN patients from another institution was used to verify the model's per-formance.Results Compared with the non-infection group,patients in the infection group had higher tumor complexity scores,elevated leukocyte counts,lower hemoglobin levels,and higher levels of urinary leukocytes,erythrocytes,and bacteria(all P＜0.05).Serum creatinine and urea nitrogen were increased,whereas eGFR was reduced(P＜0.05).CRP and PCT levels were markedly elevated in the infection group(P＜0.05).Operative indicators-including tumor in-vasion,operative time,blood loss,perioperative antibiotic use,and transfusion rates-were also significantly higher,while R0 resection rates were lower(P＜0.05).Multivariate analysis identified hemoglobin,urinary leukocytes,urinary bacteria,urinary erythrocytes,creatinine,PCT,CRP,and operative time as independent predictors.A nomogram incor-porating these factors demonstrated excellent discrimination(C-index=1.000).In the external validation cohort(infec-tion incidence 10.56％),the model achieved a C-statistic of 0.940(95％CI:0.85-1.000),sensitivity of 90.00％,and specificity of 88.00％.Calibration analysis revealed good agreement between predicted and observed risks(Brier score=0.065;calibration slope=0.956).DCA showed that use of the model provided substantial net clinical benefit across a broad range of threshold probabilities.Conclusion A reliable and clinically applicable risk prediction model for postoperative incision infection following RLPN was established and externally validated.The model exhibits strong predic-tive performance and offers valuable support for individualized preoperative risk assessment.
Keywords：complex renal tumorsnephron-sparing surgeryinfection complicationsmultivariate risk factorspredictive models
Publication Date：2025-12-15
Online Publishing Date：2026-01-12(First online date of this platform, not the publication date of the document)
Pages：7( 1871-1877 )