Enhancing door-in-door-out time in a rural primary stroke centre using a large vessel occlusion alert: a quality improvement project.

Background: Thrombectomy is the primary treatment in the management of acute large vessel occlusions (LVO) strokes; however, many rural hospitals are not able to perform this procedure and must transfer patients to thrombectomy-capable centres expeditiously. This quality improvement initiative was aimed at optimising the door-in-door-out (DIDO) time for patients with stroke at a rural primary stroke centre in the USA, with a specific goal of DIDO time of <90 min, 50% of the time and <120 min, 75% of the time.

Methods: System inefficiencies in serial event processing, radiology interpretation, teleneurology application and activation of transferring emergency medical services were identified. We implemented several system changes using multiple plan-do-study-act (PDSA) cycles. The final implementation was an LVO-Alert process, which incorporated prehospital alert based on an LVO scale, parallel event processing of emergency medicine, teleneurology and radiology activities, as well as early dispatch of Emergency Medical Services transfer resources. Two years premetrics and 1 year postmetrics were compared.

Results: We saw change on the fifth PDSA cycle, in which we implemented the LVO-Alert. For all LVO transfers, the group for whom the LVO-Alert was used (n=21) demonstrated decreased mean DIDO times compared with the no LVO-Alert group (n=20): 103.7 min, SD=49.6 vs 167.9 min, SD=64. Pre-LVO-Alert implementation, 20% of LVO patients had DIDO times <120 min and 5% had times <90 min. Post-LVO-Alert, 76% of LVO patients had DIDO times <120 min and 52% had times <90 min.

Conclusion: The application of a parallel process bundle of care model, with early activation from a prehospital positive LVO scale, improved DIDO time in this patient population.