Flexible and navigable suction ureteral access sheath–assisted versus conventional ureteral access sheath–based strategies for retrograde intrarenal surgery of lower pole renal stones: a retrospective cohort study

Background Lower pole renal stones remain challenging during retrograde intrarenal surgery (RIRS) because dependent anatomy may restrict endoscopic maneuverability and fragment clearance. This study compared a flexible and navigable suction ureteral access sheath (FANS-UAS)–assisted strategy with a conventional ureteral access sheath (c-UAS)–based strategy in routine practice. Methods Adults undergoing RIRS from September 2022 through December 2025 were reviewed retrospectively. The non-contrast CT (NCCT) cohort comprised 96/142 eligible patients (41 c-UAS, 55 FANS-UAS); 32 with kidney–ureter–bladder radiography (KUB) and 14 without follow-up were retained for sensitivity analyses. The primary endpoint was stone-free status (< 2-mm residual fragments on NCCT); complete clearance (0 mm) was secondary. Because lithotripsy and fragment management differed, findings were interpreted at the strategy level. Calendar period was examined because the strategies were largely non-contemporaneous. Results The unadjusted NCCT-based primary endpoint did not differ significantly between the FANS-UAS and c-UAS strategies (63.6% vs. 48.8%; p = 0.146), whereas the secondary complete-clearance endpoint was more frequent with FANS-UAS (45.5% vs. 14.6%; p = 0.001). Among KUB-only patients, no visible residual stone > 2 mm was recorded in 6/10 FANS-UAS and 15/22 c-UAS patients. Incorporating these proxy outcomes reduced the between-group difference (63.1% vs. 55.6%; p = 0.386), while extreme-case assignment of the 14 missing outcomes yielded FANS-minus-c-UAS differences ranging from − 1.5 to 17.5% points. Dusting was more frequent and basket use less frequent with FANS-UAS (87.3% vs. 4.9% and 10.9% vs. 70.7%, respectively). Exploratory safety comparisons were inconclusive. The adjusted OR from the primary model was imprecise (4.27, 95% CI 1.37–13.37); the exploratory Firth model including calendar period was also highly imprecise (OR 2.89, 95% profile CI 0.51–17.02) and did not resolve temporal confounding. Conclusions The unadjusted NCCT-based stone-free endpoint (< 2 mm) did not differ significantly between strategies, whereas complete clearance (0 mm) was more frequent with FANS-assisted RIRS. Because the bundled strategies were largely non-contemporaneous, temporal confounding remained unresolved and an independent strategy or device effect could not be established. Differential NCCT ascertainment may have enlarged the apparent FANS-favoring difference. Prospective contemporaneous comparisons with standardized lithotripsy, fragment management, and imaging are needed.