Data from the STRATIS registry presented at the Society of NeuroInterventional Surgery annual meeting (SNIS; 25–28 July, Boston, USA) show that the transfer of stroke patients from one hospital to another is responsible for delays in revascularisation and suggest that there may be a place for bypassing closer hospitals for one that is capable of administering mechanical thrombectomy. The data were presented by Michael Froehler (Vanderbilt University Medical Center, Nashville, USA).
The STRATIS registry is a prospective, multicentre study of patients with large vessel occlusion treated with the Solitaire stent retriever within eight hours of onset. To assess the systems of care efficiency and interhospital transfer delays, patients were separated by direct presentation (direct group) versus interhospital transfer to the enrolling hospital (transfer group), and further characterised by the use of IV tPA or mechanical thrombectomy alone. Total alarm-to-revascularisation time measured overall performance of the system of care. Investigators also calculated a hypothetical “bypass” scenario by comparing the door-to-tPA times for the two groups and adding the transfer time to the direct group for an extremely conservative estimate of additional travel time.
A total of 688 patients were analysed. Froehler reported that median times from stroke onset to revascularisation for direct versus transfer patients in the mechanical thrombectomy alone sub-group were 238.5 and 325 minutes, respectively, and in the IV tPA plus mechanical thrombectomy group the median times were 192 and 305.5 minutes, respectively.
Further, median alarm-to-revascularisation times for direct versus transfer patients in the mechanical thrombectomy only subgroup were 189 and 264 minutes, respectively (75 minute difference, p=0.0001), and in the IV tPA subgroup were 169 and 268, respectively (99 minute difference; p˂0.0001).
These differences were accounted for by imaging-to-transfer time, which was 93 minutes for mechanical thrombectomy alone and 87 minutes for IV tPA. Median door-to-tPA times were 56.5 minutes at regional hospitals and 38 minutes at enrolling sites (p˂0.0001). Transfer time was 32 minutes, making the hypothetical bypass time-to-tPA 70 minutes.
“Patient transfer for endovascular treatment for large vessel occlusion results in a significant delay of approximately an hour and a half. That delay is entirely accounted for by the delay between intracranial imaging and departure of the first hospital which I think reflects the time that treatment decisions are being made and transfer arrangements are being made. Certainly that is a target for improvement for systems of care in the future, but it also really raises the question of whether or not we should be considering bypass. For that we are now doing some more specific patient-level analyses to look at individual routing decisions,” Froehler explained.
In their conservative model of hospital bypass, it is suggested that successful revascularisation would be achieved 99 minutes earlier by bypass. He added that this may have significant implications for regional stroke systems of care.