Abstract Background and aims In-hospital stroke (IHS) imposes a great challenge on clinical care as comorbidities and recent surgery often preclude intravenous thrombolysis (IVT), while endovascular treatment (EVT) remains an option for large-vessel-occlusion (LVO). This study compares baseline, treatment and outcome parameters between IHS and out-of-hospital stroke (OHS) patients. Methods We analyzed EVT patients enrolled in the prospective, multicenter German Stroke Registry (GSR-ET) between June 2015 and December 2023 grouped by IHS and OHS. Primary endpoint was the modified Rankin Scale (mRS) shift at 90 days after propensity score matching (PSM) in a multivariable ordinal regression analysis. Secondary parameters of interest were frequency of IVT, treatment times, length of hospitalization, and adverse events (AEs). Results 18,069 patients were included, 17,459 (96.6%) as OHS and 610 (3.4%) as IHS. IVT was administered less frequently in IHS (18.9% vs. 46.9%; P0.001). IHS was associated with shorter time from symptom-onset to flow-restoration (198min IQR 135-333 vs. 250min IQR 190-337; P0.001) and shorter time from last-known-well to flow-restoration (436min IQR 244-935 vs. 575min 355-895; P=0.029), but longer hospital stays (10d IQR 6-18 vs. 8d IQR 5-13; P0.001). More AEs were noted for IHS (12.3% vs 7.7%; P0.001). PSM yielded two groups of n=409 patients. Compared to OHS, patients with IHS had worse 90d-mRS (cOR 1.512; 95%-CI 1.271-1.800; P0.001). Conclusions EVT in IHS with LVO is associated with worse functional outcomes even though EVT could be performed more rapidly. In addition to lower IVT rates, pre-existing conditions, complications, and longer recovery times are likely explanations. Conflict of interest Dr. Schaefer reports travel reimbursements by Bayer AG outside of the submitted work. Dr. Bohmann reports speaker honoraria from Laerdal, AstraZeneca, Bristol-Myers-Squibb, Pfizer, Medtronic and grants and personal fees from Alexion, AstraZeneca, Stryker Neurovascular, Boehringer Ingelheim. The other authors report no conflicts. Figure 1 - belongs to Results Figure 2 - belongs to Conclusions
Schäefer et al. (Fri,) studied this question.