Venous Access Alone Versus Arterial and Venous Access for Patent Arterial Duct Device Closure in Childhood

ABSTRACT Background The persistently patent arterial duct accounts for ~12% of congenital heart lesions. Untreated, it may result in heart failure due to volume loading of the left heart, pulmonary hypertension, and infective endarteritis. Percutaneous device closure is the preferred occlusion technique, with the standard approach consisting of femoral artery access for angiography and venous access for device delivery (AA). A venous‐only strategy (VA) for angiography and device delivery can also be employed. Hypothesis We hypothesized that VA would eliminate the need of arterial entry, reduce procedure times and radiation exposure compared to standard AA. Methods This is a retrospective cohort study of isolated arterial duct device closure at the Hospital for Sick Children from January 1, 2011, through December 31, 2022. Exclusions included premature neonates, children requiring arterial access for monitoring, and those who underwent other procedures. Children were categorized based upon initial access determined by operator preference into VA or AA groups. Results The cohort consisted of 405 children, 252 (62.2%) females, with a median age of 3.1 years (IQR 1.30–5.84), median weight 13.2 kg (IQR 9.0–19.5), and duct diameter of 2.9 mm (IQR 2.0–3.5) with no significant differences between the groups. Type A ducts were more frequent in the AA group (90% vs. 72%). The VA group included 106 children, of which 14 (13.2%) required AA conversion for angiography due to complex ductal anatomy, to assess device position before release, but remained in the VA group for analysis. Children in the VA group had lower dose area product (DAP) ( p < 0.001), fluoroscopy times ( p = 0.025), contrast volumes ( p < 0.001), procedure times ( p < 0.001), and recovery room lengths of stay (LOS) ( p < 0.001). Six (5.7%) VA children required admission compared to 44 (14.7%) in the AA group ( p = 0.015) with no difference in reintervention rates. Weighted regression analysis showed VA was associated with reduced admission likelihood (OR: 0.354 [0.131, 0.822], p = 0.024), DAP (coef −126.4 [−213.3, −39.4], p = 0.004), and contrast volumes (coef 31.2 [−36.6, −25.9], p < 0.001) compared to AA. Conclusions Venous‐only access was associated with lower DAP and recovery room LOS. Additionally, VA was associated with a lower likelihood of admission with no difference in reintervention rates, suggesting procedural safety. These findings support the consideration of VA as a preferred approach for appropriate cases.