We evaluated utilization of a limited-sequence MRI protocol for rapid assessment of acute stroke in patients with high concern for mimic. We hypothesized this would reduce delay to appropriate management.

Stroke-related costs in the US were close to 53 billion dollars as of 2018, including those of health care services, medicine, and missed days of work. We often admit stroke mimics to the stroke or observation units for up to several days awaiting MRI to determine the etiology of their symptoms. 

We performed a retrospective analysis on patients seen through the ED before and after initiation of our limited-sequence MRI protocol (DWI, ADC, GRE, and T2 FLAIR). Each patient was evaluated by a neurology resident and appropriate CT imaging obtained. Stroke fellow or attending, after discussion with resident, determined the appropriateness of intervention. Patients qualified if their case was highly suggestive of mimic. Information regarding MRI and disposition were collected. Data were analyzed with Student's T-test with statistical significance defined as p<0.05.

The number of cases were 31 (control) and 27 (intervention). Time from order to scan (in hours) was significantly decreased (p <0.0001) with the average time being 2.7 in the intervention group vs. 16.9 in control. Discharge from ED increased (control 4 to intervention 15). Admissions to the Observation Unit decreased (control 8 to intervention 0), as did those to the Stroke Unit (control 16 to intervention 8). The average length of stay (in days) decreased (2.10 control to 1.22 intervention).

Appropriate and timely medical decision making is of ultimate importance for acute stroke patients. This requirement can be optimized for effective utilization of time and resources. With our protocol of clinical evaluation by neurology staff and rapid protocol MRI, we streamlined our process and minimized unnecessary further evaluation in cases of stroke mimics.