Preoperative prediction of infectious-inflammatory complications after endoscopic ureterolithotripsy in adults: development and internal validation of a discriminant model

Infectious-inflammatory complications (IIC) remain among the most clinically significant adverse events after endoscopic treatment of ureteral stones and, in their severe forms, may progress to urosepsis; reliable preoperative risk stratification could allow prophylaxis and monitoring to be individualized. This study aimed to identify independent predictors of IIC in adults after endoscopic ureterolithotripsy and to develop and internally validate a discriminant prognostic model. We analyzed 385 adult patients who underwent endoscopic treatment of ureteral stones at a single tertiary urology center; demographic, laboratory, stone-related, and operative variables were compared between patients with and without IIC using univariate tests, independent predictors were selected by stepwise discriminant analysis, and the resulting canonical discriminant function was internally validated on an independent prospectively assembled sample of 43 patients. IIC occurred in 76 of 385 patients (19.7%) and were predominantly mild: postoperative fever without an infectious focus (50.0%) and urinary tract infection (28.9%) prevailed, whereas acute pyelonephritis (18.4%) and urosepsis (2.6%) were uncommon; female sex (p = 0.031), preoperative bacteriuria and pyuria, combined bacteriuria plus pyuria, preoperative upper-tract drainage, longer operative time, and greater blood loss were significantly associated with IIC, whereas age, body-mass index, and stone location, size, and density were not. The final four-variable model — pyuria, peripheral-blood leukocyte count, bacteriuria, and emergency upper-tract drainage — showed strong discrimination (eigenvalue 0.754; canonical correlation 0.656; Wilks' λ = 0.570; χ² = 214.1; p < 0.000001), with sensitivity 76.3%, specificity 81.5%, and overall accuracy 80.4%; on internal validation, overall accuracy was 81.4%, sensitivity 81.8%, and specificity 81.3%. A simple four-predictor discriminant model built from routinely available preoperative parameters thus reliably stratifies adults by the risk of IIC after endoscopic ureterolithotripsy, and its high negative predictive value identifies low-risk patients and supports individualized antimicrobial prophylaxis and postoperative monitoring.