Direct cost analysis of rapid MRI in the emergency department evaluation of patients suspected of having acute ischemic stroke*.

Abstract

Abbreviated "rapid MRI" protocols have become more common for the evaluation of acute ischemic stroke (AIS). Prior research has not evaluated the effect of rapid MRIs on cost or hospital length of stay in AIS patients. We retrospectively identified AIS patients who presented within 6h of acute neurologic symptom onset to an emergency department (ED) and activated a "brain attack" code. We included sequential patients from January 2012 to September 2015, before rapid MRI was available, who had CT perfusion (CTP) and compared them to patients from October 2015 to May 2018 who had a rapid MRI. We used inverse-probability-weighting (IPW) to balance the cohorts. The primary outcomes were direct cost to our healthcare system and total hospital length of stay (LOS). We included 408 brain attack activations (mean ± SD age 62.1 ± 17.6years, 47.8% male): 257 in the CTP cohort and 151 in the MRI cohort. Discharge diagnosis was ischemic stroke in 193/408 (47.3%). After patient matching, we found significant reductions for the MRI cohort in total cost (-18.7%, 95% CI -35.0, -2.4, p = 0.02) and hospital LOS (-17.0%, 95% CI -31.2, -2.8, p = 0.02), with no difference in ED LOS (p = 0.74) as compared to the CTP cohort. Although these results are preliminary and hypothesis-generating, we found that the use of a rapid MRI protocol in emergency department brain attacks was associated with a 18.7% reduction in total direct cost and 17% reduction in hospital length of stay.