Interhospital Transfer Dependency and Delays to Arterial Puncture in UK Thrombectomy Centres: A National Observational Study

Introduction Mechanical thrombectomy (MT) is a highly time-dependent intervention, with functional outcomes deteriorating as the interval from stroke onset to arterial puncture lengthens. Interhospital transfer pathways broaden regional access to MT but introduce additional steps that may prolong this interval. A widely held assumption in health service planning is that hubs treating a greater proportion of transferred patients therefore experience greater average delay; however, this assumption is yet to be examined at a national level. This study evaluates the association between transfer dependency and delay to arterial puncture at a national level and identifies the operational contributors to delay within the transferred pathway. Materials and methods This was a two-year retrospective observational analysis of aggregated, centre-level data from the Sentinel Stroke National Audit Programme (SSNAP) for all MT procedures nationally between April 2023 and March 2025. Direct and transferred pathways were compared for each independent hyperacute stroke unit (HASU), and transfer-associated delay (TAD) was derived in this study to examine the association between transfer dependency and delay. Transferred pathways were decomposed into door-in-door-out (DIDO), conveyance, and receiving-centre door-to-puncture (DTP) intervals to identify operational contributors to delay. Pearson and Spearman correlations were calculated, and sensitivity analyses (small-caseload exclusion, volume weighting, absolute transferred volume and a negative control) were performed to assess robustness. Results Among 7,203 MT procedures (3,421 in 2023/24 and 3,782 in 2024/25), approximately 60% followed interhospital transfer. The dependency-TAD association was inconsistent between years (r = +0.50, 95% CI 0.10 to 0.76, p = 0.017 in 2023/24 versus r = -0.06, 95% CI -0.46 to 0.36, p = 0.78 in 2024/25) and supports a cautious interpretation of the relationship between transfer dependency and delay. Structural decomposition of transferred pathways demonstrated meaningful associations between DIDO, DTP and TAD: median DIDO was 131 and 129 minutes respectively in the two years and accounted for 55-57% of the median total transferred interval. Centre-level transfer dependency, DIDO and DTP values were highly stable year-on-year (r = 0.86-0.89), suggesting that these delays are consistent and reproducible features of pathway organisation. Both DTP and DIDO correlated with TAD substantially in both years. Conclusion Whilst transferred pathways demonstrated substantial additional delay to arterial puncture, higher transfer dependency was not consistently associated with delay. DIDO was uniformly prolonged across the national network and, together with receiving-centre DTP, correlated with TAD in both years, suggesting that delay within the transferred pathway is more closely related to organisational factors. Both represent potentially valuable targets for regional quality improvement initiatives.