Inter‐hospital transfer for thrombectomy: transfer time is brain

Abstract Background and purpose Patients with acute ischaemic stroke and a large vessel occlusion who present to a non‐endovascular‐capable centre often require inter‐hospital transfer for thrombectomy. Whether the inter‐hospital transfer time is associated with 3‐month functional outcome is poorly known. Methods Acute stroke patients enrolled between January 2015 and December 2022 in the prospective French multicentre Endovascular Treatment of Ischaemic Stroke registry were retrospectively analysed. Patients with an anterior circulation large vessel occlusion transferred from a non‐endovascular to a comprehensive stroke centre for thrombectomy were eligible. Inter‐hospital transfer time was defined as the time between imaging in the referring hospital and groin puncture for thrombectomy. The relationship between transfer time and favourable 3‐month functional outcome (modified Rankin Scale 0–2) was assessed through a mixed logistic regression model adjusting for centre and symptom‐onset‐to‐referring‐hospital imaging time, age, sex, diabetes, referring hospital National Institutes of Health Stroke Scale score, Alberta Stroke Programme Early Computed Tomography Score, occlusion site and intravenous thrombolysis use. Results Overall, 3769 patients were included (median inter‐hospital transfer time 161 min, interquartile range 128–195; 46% with favourable outcome). A longer transfer time was independently associated with lower rates of favourable outcome (p < 0.001). Compared to patients with transfer time below 120 min, there was a 15% reduction in the odds of achieving favourable outcome for transfer times between 120 and 180 min (adjusted odds ratio 0.85; 95% confidence interval 0.67–1.07), and a 36% reduction for transfer times beyond 180 min (adjusted odds ratio 0.64; 95% confidence interval 0.50–0.81). Conclusions A shorter inter‐hospital transfer time is strongly associated with favourable 3‐month functional outcome. A speedier inter‐hospital transfer is of critical importance to improve outcome.

in the prospective French multicentre Endovascular Treatment of Ischaemic Stroke registry were retrospectively analysed.Patients with an anterior circulation large vessel occlusion transferred from a non-endovascular to a comprehensive stroke centre for thrombectomy were eligible.Inter-hospital transfer time was defined as the time between imaging in the referring hospital and groin puncture for thrombectomy.
The relationship between transfer time and favourable 3-month functional outcome (modified Rankin Scale 0-2) was assessed through a mixed logistic regression model adjusting for centre and symptom-onset-to-referring-hospital imaging time, age, sex, diabetes, referring hospital National Institutes of Health Stroke Scale score, Alberta Stroke Programme Early Computed Tomography Score, occlusion site and intravenous thrombolysis use.
Results: Overall, 3769 patients were included (median inter-hospital transfer time 161 min, interquartile range 128-195; 46% with favourable outcome).A longer transfer time was independently associated with lower rates of favourable outcome (p < 0.001).Compared to patients with transfer time below 120 min, there was a 15% reduction in the odds of achieving favourable outcome for transfer times between 120 and 180 min (adjusted odds ratio 0.85; 95% confidence interval 0.67-1.07),and a 36% reduction for transfer times beyond 180 min (adjusted odds ratio 0.64; 95% confidence interval 0.50-0.81).

Conclusions:
A shorter inter-hospital transfer time is strongly associated with favourable 3-month functional outcome.A speedier inter-hospital transfer is of critical importance to improve outcome.

K E Y W O R D S
inter-hospital transfer, ischaemic stroke, thrombectomy

INTRODUC TI ON
In patients with acute ischaemic stroke harbouring a large vessel occlusion (LVO), the efficacy of endovascular therapy (EVT) is well established in a large proportion of patients up to 24 h from symptom onset.Most LVO-related stroke patients are first evaluated at primary stroke centres or community hospitals before being transferred to a centre that performs EVT [1].
The inter-hospital transfer is a complicated process, involving numerous stakeholders and requiring a high degree of coordination between the referring hospital, transportation and receiving hospital teams.Hence, long delays are commonly observed, with median transfer times consistently exceeding 2-3 h in most settings [1][2][3].The effectiveness of EVT is well known to be time-sensitive, due to the progression of the irreversibly injured ischaemic brain tissue over time [4,5].However, this association has mainly been reported and quantified for elapsing time from symptom onset to EVT reperfusion, not specifically for inter-hospital transfer time.Demonstrating such an association is of major importance, as it may underscore the urgent need for strategies aimed at expediting transfers, as well as indicate that inter-hospital transfer is a promising time-window for neuroprotection trials.
Here, the aim was to study whether inter-hospital transfer time is associated with 3-month functional outcome in a large multicentre registry.

ME THODS
Our analysis was conducted according to the STROBE criteria for observational studies.Informed consent was obtained from each patient or their relatives.The data supporting the study findings are available from the authors upon reasonable request.

Data sources
The data were extracted from the Endovascular Treatment of Ischaemic Stroke (ETIS) registry between January 2015 and December 2022, for which detailed methods have been published previously.Briefly, ETIS is an ongoing French multicentre prospective observational study collecting data of consecutive patients with LVO undergoing EVT (NCT03776877).Data from acute stroke patients who fulfilled the following criteria were extracted for the current study: (1) initial admission at a non-EVT-capable centre where a brain imaging showed an anterior circulation LVO, (2) subsequent transfer to a comprehensive centre for consideration of EVT and (3) pre-stroke modified Rankin Scale (mRS) score 0 or 1.
All patients receiving groin puncture for EVT were included, regardless of whether EVT was eventually attempted and whether successful reperfusion was achieved.

Clinical and radiological data
Clinical and imaging data routinely recorded in the acute stroke setting were collected.Imaging data were locally assessed by senior neuroradiologists.Time of inter-hospital transfer was operationally defined as the time between baseline imaging in the referring hospital (i.e., when the LVO was first identified) and the groin puncture for EVT in the comprehensive centre.The primary outcome was favourable 3-month outcome, defined as an mRS ≤2.Secondary outcome was the overall distribution of 3-month mRS.Three-month mRS scores were collected by certified investigators during routinely scheduled visits or by trained research nurses during a standardized telephone interview.

Statistical analysis
The association of inter-hospital transfer time with favourable outcome and overall mRS distribution (shift analysis) was assessed using mixed logistic regression models (binary for favourable outcome and ordinal for shift analysis) considering centre as a random

RE SULTS
During the study period, 3769 patients met inclusion criteria.Mean Longer inter-hospital transfer time was independently associated with lower rates of favourable outcome (p < 0.001, Table 1 and Figure 1).Compared to patients with transfer time below 120 min, there was a 15% reduction in the odds of achieving favourable outcome for transfer times between 120 and 180 min (adjusted odds ratio 0.85; 95% confidence interval 0.67-1.07)and a 36% reduction for transfer times beyond 180 min (adjusted odds ratio 0.64; 95% confidence interval 0.50-0.81)(Table 1).Similar results were observed in the shift analysis across the entire mRS score (Table 1).
No significant heterogeneity in the association between transfer time and favourable outcome was observed across the pre-specified onset-to-referring-hospital imaging time, admission (referring hospital) NIHSS score and ASPECTS, occlusion site, intravenous thrombolysis and time-of-day of referring hospital imaging subgroups (Figure S1).

DISCUSS ION
In this study including 3769 LVO-related stroke patients transferred from a non-endovascular centre for EVT, longer inter-hospital transfer time was independently associated with poorer 3-month functional outcome.To our knowledge, such association has never been shown.One study has reported a non-significant trend between TA B L E 1 Association between inter-hospital transfer time and 3-month functional outcome.door-in door-out time referring hospital and 3-month outcome yet was limited by a small sample size [7].
Our data expand the known association between symptom onset to reperfusion time and functional outcome [4,5] and highlight the urgent need for developing strategies to expedite inter-hospital transfer to improve functional outcome.
Our results show that transfer time is long in day-to-day practice, with three-quarters of patients with transfer time longer than 2 h, in line with other reports [2,3].There are several potential ways to speed up inter-hospital transfer workflow, the first of which is reducing door-in door-out time in the referring hospital, with the establishment of in-house protocols for early arterial imaging, cloudbased image sharing with the EVT-capable centre, and early mobilization of transport resources [8].The use of positive feedback strategies has also shown benefit to reduce in-hospital workflow [9,10].Quality improvement programmes focused on door-in door-out time should be used in non-EVT-capable centres to provide feedback on performance and improve workflow.The second way is shortening the transportation time, for instance facilitating the use of air transportation rather than ground ambulances in appropriate settings [11].The third way is reducing door-to-reperfusion time upon comprehensive centre arrival, for example through direct admission in the angiosuite [12].Last, transferring a flying intervention team instead of transferring the patient has been shown to significantly reduce time to EVT [13,14].
In addition to transfer time reduction, which is of critical importance yet may reach a limit given the intricate nature of the process, innovative neuroprotective therapies aimed at slowing down infarct growth emerge as particularly relevant during interhospital transfer [1,2,15].Indeed, the potential benefits of neuroprotective therapies are more likely to be observed in transferred patients, where a substantial decrease in infarct growth could be achieved due to the expected long treatment exposure.This contrasts with patients directly admitted to endovascular-capable centres, where shorter imaging-to-recanalization times are typically observed.
No significant heterogeneity in the association between transfer time and favourable outcome across our pre-specified subgroups was observed, suggesting that fast transfer should be targeted regardless of baseline clinical or imaging characteristics.Poor leptomeningeal collaterals on computed tomography angiography or perfusion imaging have been shown as the primary factor associated with fast infarct growth during inter-hospital transfer for EVT and may therefore help to identify a subgroup of patients for whom ultrafast inter-hospital transfer should be prioritized [1][2][3].However, collateral assessment in the referring hospital was not available in our registry.This warrants further research.
This study has limitations.The transferred patients who did not undergo groin puncture were not enrolled in our registry, which might have biased the results.However, a direct-toangiosuite protocol is performed for transferred patients in most comprehensive stroke centres in France.Also, collateral or core volume assessment in the referring hospital was not recorded in our registry, impairing the study of heterogeneity according to these key variables.

CON CLUS ION
In LVO-related acute stroke patients transferred from a nonendovascular-capable centre for EVT, longer inter-hospital transfer time is strongly associated with poorer 3-month functional outcome.
Expediting inter-hospital transfer is of critical importance to improve outcome.
effect and adjusted on the following pre-specified confounding factors: symptom-onset-to-referring-hospital imaging time, age, sex, diabetes, admission (referring hospital) National Institutes of Health Stroke Scale (NIHSS) score, Alberta Stroke Programme Early Computed Tomography Score (ASPECTS), occlusion site and intravenous thrombolysis use.Finally, heterogeneity in associations between transfer time and favourable outcome according to symptom-onset-to-referring-hospital imaging time, NIHSS score, ASPECTS, occlusion site, intravenous thrombolysis use and timeof-day of referring hospital imaging was assessed by including the corresponding interaction term in the multivariable mixed logistic regression models [1, 6].More details regarding statistical analysis are provided in Appendix S1.

(
standard deviation) age was 70 (15) years and 1892 (50%) patients were male.Median (interquartile range, IQR) NIHSS score on admission in the referring hospital was 16 (IQR 10-20), and median time from symptom onset to referring hospital imaging was 127 min (IQR 97-174 min).On the referring hospital imaging, median ASPECTS was 8 (IQR 6-9) and occlusion site was the intracranial internal carotid in 596 (16%) patients and the first and second segments of the middle cerebral artery in 2597 (69%) and 573 (15%) patients, respectively.Intravenous thrombolysis was administered in the referring hospital in 2227 (59%) patients.Median inter-hospital transfer time was 161 min (IQR 128-195 min).Transfer time was shorter than 120 min in 738 (20%) patients, between 120 and 180 min in 1705 (45%) patients and longer than 180 min in 1326 (35%) patients.Upon comprehensive centre arrival, the EVT procedure was attempted in 89.6% of patients, and successful reperfusion (modified Thrombolysis in Cerebral Infarction score 2b-3) was achieved in 89.1% of them.Median time from groin puncture to successful recanalization was 35 min (IQR 24-54 min) amongst patients successfully treated by EVT.Favourable outcome was observed in 1722 (46%) patients.

Fully
Abbreviations: CI, confidence interval; cOR, common OR; mRS, modified Rankin Scale; OR, odds ratio.a Odds ratio per 1 standard deviation log increase in transfer time and p values obtained using a mixed logistic regression model including logtransformed transfer time as continuous variable.b Adjusted on centre, symptom-onset-to-referring-hospital imaging, age, sex, diabetes, admission (referring hospital) NIHSS score and ASPECTS, occlusion site and intravenous thrombolysis.c p values obtained using a mixed logistic regression model including transfer time as categorical variable.