Impacts of in-hospital workflow on functional outcome in stroke patients treated with endovascular thrombectomy.

High-performance in-hospital workflow may save time and improve the efficacy of thrombectomy in patients with acute ischemic stroke. However, the optimal in-hospital workflow is far from being formulated, and the current models varied distinctly among centers. This study aimed to evaluate the impacts of in-hospital workflow on functional outcomes after thrombectomy. Patients were enrolled from a multi-center registry program in China. Based on in-hospital managing procedure and personnel involved, two workflow models, neurologist-dominant and non-neurologist-dominant, were identified in the participating centers. Favorable outcome was defined as a mRS score of ≤ 2 at 90 days of stroke onset. After patients being matched with propensity score matching (PSM) method, ratios of favorable outcomes and symptomatic intracerebral hemorrhage (sICH) were compared between patients with different workflow models. Of the 632 enrolled patients, 543 (85.9%) were treated with neurologist-dominant and 89 (14.1%) with non-neurologist-dominant model. 88 patients with neurologist-dominant model and 88 patients with non-neurologist-dominant model were matched with PSM. For the matched patients, no significant differences concerning the ratios of successful recanalization (92.0% vs 87.5%, P = 0.45), sICH (17.0% vs 14.8%, P = 0.85), favorable outcome (42.0% vs 42.0%, P = 1.00) were detected between patients with neurologist-dominant model and those with non-neurologist-dominant model. Patients with neurologist-dominant model had shorter door to puncture time (124 (86-172) vs 156 (120-215), P = 0.005), fewer passes of retriever (2 (1-3) vs 2 (1-4), P = 0.04), lower rate of > 3 passes (11.4% vs 28.4%, P = 0.004), and lower incidence of asymptomatic intracerebral hemorrhage rate (27.3% vs 43.2%, P = 0.045). Although the neurologist-dominant model may decrease in-hospital delay and risk of asymptomatic intracerebral hemorrhage, workflow models may not influence the functional outcome significantly after thrombectomy in patients with acute ischemic stroke.