Supplementary Material for: Characterizing In-Hospital Acute Ischemic Strokes: Clinical Profiles and Predictors of Acute Treatment

Background and Objectives
Treatment of in-hospital acute ischemic strokes (AIS) is challenging. We aimed to characterize in-hospital AIS and identify predictors of receiving thrombolysis and thrombectomy.

Methods
We conducted an analysis of a retrospective cohort study using administrative data and chart review as part of the INPATIENTS study (Comparing In-Hospital and Community-Onset Strokes in Alberta). All in-hospital AIS patients admitted in the province between January 1, 2018 and December 31, 2022 were included. Clinical characteristics and quality of care measures were compared between treated and non-treated patients. We used multivariable logistic regression to identify predictors of acute treatment and assessed model performance using ROC curves and calibration plots.

Results
Only 7.3% (158 of 2,159) in-hospital AIS patients received thrombolysis or thrombectomy between 2018-2022. Treated patients had higher NIHSS scores (median 12 vs. 8), fewer recent invasive procedures (42% vs. 53%), and were less likely to have altered consciousness (12.0% vs. 52.1%). Common reasons for not receiving thrombolysis included delayed recognition and recent procedures. Treated patients more often received standard stroke evaluation. The final logistic regression model included age, sex, NIHSS, altered consciousness, admitting service, and comorbidities as predictors of treatment. It showed good discrimination (AUC = 0.8371), though calibration issues may affect its generalizability.

Discussion
In-hospital AIS patients treated with thrombolysis and thrombectomy had more severe strokes, were less likely to have altered consciousness, and more often received standard stroke evaluations than non-treated patients. These differences emphasize the need to better understand barriers and develop new approaches to in-hospital stroke management.