Determinants of ureteral obstruction after percutaneous nephrolithotomy

Background Ureteral obstruction after percutaneous nephrolithotomy (PCNL) may require prolonged drainage with a nephrostomy tube (NT) or ureteral stent, but it is not well understood how and why this occurs. The goal of this study was to identify risk factors associated with postoperative ureteral obstruction to help guide drainage tube selection. Methods Prospective data from adult patients enrolled in the Registry for Stones of the Kidney and Ureter (ReSKU) who underwent PCNL from 2016 to 2020 were used. Patients who had postoperative NTs with antegrade imaging-based flow assessment on postoperative day one (POD1) were included. Patients with transplanted kidneys or those without appropriate preoperative imaging were excluded. We assessed the association between patient demographics, stone characteristics, and intraoperative factors using POD1 antegrade flow, a proxy for ureteral patency, as the primary outcome. Stepwise selection was used to develop a multivariate logistic regression model controlling for BMI, stone location, stone burden, ipsilateral ureteroscopy (URS), access location, estimated blood loss, and operative time. Results We analyzed 241 cases for this study; 204 (84.6%) had a visual clearance of stone. Antegrade flow on POD1 was absent in 76 cases (31.5%). A multivariate logistic regression model found that stones located anywhere other than in the renal pelvis (OR 2.63, 95% CI 1.29–5.53; p = 0.01), non-lower pole access (OR 2.81, 95% CI 1.42–5.74; p