Abstract Body: Introduction
Acute treatment for stroke often requires emergent interhospital transfer for access to advanced therapies not available at the initial hospital. Prolonged transfer times have been associated with worse outcomes. Door-in-door-out time (DIDO: the amount of time a patient spends in the transferring emergency department [ED]) is an important quality metric in acute stroke care, with current recommendations for DIDO times ≤ 120 minutes. We sought to characterize trends and predictors of DIDO times for interhospital stroke transfers using the Greater Cincinnati Northern Kentucky Stroke Study (GCNKSS).

Methods
We utilized data from the GCNKSS, a population-based epidemiologic stroke study, from the following time points: 1999, 2005, 2010, 2015, and 2020. Patients ≥18 years with acute ischemic stroke (AIS) or hemorrhagic stroke (HS) who presented to an initial ED and were not admitted but were transferred to another hospital were included. The primary outcome was DIDO time. Temporal trends in DIDO time were tested using the Mann-Kendall trend test. Generalized linear mixed effects models with hospital-specific random intercepts were constructed to evaluate the associations between patient- and hospital-level covariates and DIDO time.

Results
Of 13,678 stroke cases over the time periods studied, 1574 patients met inclusion criteria for the overall stroke group (mean age 64.7 [SD: 15.6], 51.6% female), with 851 (54.1%) having AIS and 723 (45.9%) HS. Over the time periods examined, the median DIDO time for the overall stroke group was 213 minutes (IQR 142-305), and DIDO times significantly increased over time (Figure 1; P<.0001). In the overall stroke group, hospital-level factors explained 4.0% of the variation in DIDO time. In the multiple regression model, factors associated with increased DIDO time included: history of prior stroke (+33.6 minutes; P<.0001) and receipt of MRI prior to transfer (+165.7 minutes; P<.0001); whereas EMS transport (-30.0 minutes; P<.001) and increasing NIHSS score (P<.0001) were associated with decreased DIDO time. Results were similar for AIS and HS subgroups (Table 2).

Conclusions
In this population-based epidemiologic study, DIDO times exceeded recommended time targets and increased over time. Hospital-level factors accounted for only a minor proportion of the overall variation in DIDO times, suggesting that future quality improvement efforts should target modifiable clinical and systems factors to improve DIDO times.