Color Doppler ultrasound-guided tract creation reduces bleeding in percutaneous nephrolithotomy: an overlap-weighted cohort study

Abstract To compare the effects of Color Doppler flow imaging (CDFI) versus conventional two-dimensional ultrasound (2D-US) on bleeding control and surgical outcomes during percutaneous nephrolithotomy (PCNL). Between January 2023 and December 2024, 268 consecutive patients were enrolled in this retrospective-prospective hybrid cohort study. 134 patients received conventional two-dimensional ultrasound-guided puncture (2023 historical control cohort), and another 134 patients received CDFI-guided puncture that avoided renal vessels ≥ 2 mm (2024 observational cohort). This trial was retrospectively registered in ChiCTR (ChiCTR2500114024) on December 5, 2025. The primary endpoint was intraoperative estimated blood loss; postoperative hemoglobin decline was set as a supporting collinear derived measure. Secondary endpoints included transfusion, interventional embolization, stone-free rate, and perirenal hematoma. Overlap-weighting was used to balance baseline stone burden, and Benjamini-Hochberg FDR correction was applied for multiple secondary comparisons. After weighting, CDFI-guided PCNL significantly reduced the primary endpoint of intraoperative blood loss (304.8 ± 17.3 mL vs. 373.3 ± 23.8 mL; effect − 68.6 mL; 95% CI − 126.1, − 11.0; P = 0.020). The collinear supporting measure of postoperative hemoglobin decline showed a consistent reduction (9.51 ± 0.53 g L⁻¹ vs. 11.73 ± 0.75 g L⁻¹; effect − 2.22 g L⁻¹; 95% CI − 4.02, − 0.42; P = 0.016), but was not treated as an independent confirmatory outcome. Stone-free rates, transfusion, embolization, and hematoma rates were comparable after FDR correction (all P > 0.05). CDFI guidance during 18-Fr single-tract PCNL is associated with a statistically significant reduction in intraoperative estimated blood loss, with a consistent supporting trend in hemoglobin decline, without compromising stone clearance efficacy. These findings support CDFI as a useful adjunctive optimization tool for renal access, but confirmatory multicenter randomized trials are still required.