Post-procedural access blood flow and repeat endovascular intervention in hemodialysis vascular access: a recurrent-event cohort study

Abstract

The prognostic value of access blood flow (Qa) measured after endovascular intervention is uncertain. We performed an episode-based recurrent-event cohort study of maintenance hemodialysis patients undergoing endovascular intervention for arteriovenous access dysfunction. The primary exposure was post-procedural Qa modeled continuously per 100 mL/min higher Qa; post-procedural Qa <1000 mL/min was examined as a secondary clinically interpretable threshold. The outcome was time to the next access intervention, analyzed using an Andersen–Gill counting-process Cox model with patient-level cluster-robust variance. A total of 131 patients contributed 382 intervention episodes, and post-procedural Qa was available for 315 episodes. Median time to repeat intervention was shorter for Qa <1000 mL/min than for Qa ≥1000 mL/min. In the primary adjusted model, each 100 mL/min higher post-procedural Qa was associated with a lower hazard of repeat intervention (aHR 0.941, 95% CI 0.896–0.990, p = 0.018). In the secondary threshold-based analysis, post-procedural Qa <1000 mL/min was associated with a higher hazard of repeat intervention (aHR 2.005, 95% CI 1.351–2.974, p < 0.001). Findings were generally consistent across sensitivity analyses, including IPW for missing post-procedural Qa, exclusion of occlusion episodes, shared-frailty modeling, adjustment for available pre-procedural Qa, and cause-specific terminal-event analysis. Lower post-procedural Qa may serve as a practical hemodynamic marker for post-procedural risk stratification after endovascular intervention.