PRIME: A Randomized Controlled Trial to Test Portable MRI for Neurological Emergencies in the Emergency Department

Despite evidence supporting the use of MRI for diagnosing neurologic emergencies, limited access in the emergency department (ED) leads to delays, missed diagnoses, and potentially avoidable transfers. Portable MRI (pMRI) has emerged as an accessible tool for rapid bedside imaging. We evaluated whether pMRI improves operational efficiency while maintaining diagnostic performance compared with conventional MRI (cMRI).

We conducted a single-center, prospective, randomized, parallel-group trial at a tertiary academic ED. Adult patients were prospectively randomized 2:1 to an intervention arm where they received pMRI versus a control arm where they received cMRI alone. Participants in the intervention arm also received a standard-of-care cMRI. The primary outcome was time from order to scan start comparing the intervention to control arms. Secondary outcomes included diagnostic quality, detection of pathology, and volumetric agreement. We excluded 4 cMRI studies that were completed after discharge from time comparisons. Differences were assessed using the Mann-Whitney U test.

We enrolled 100 patients with 68 randomized to intervention and 32 to control (mean age 54.0±18.7, 44% male, 40% non-White). Patients in the intervention group received pMRI faster than patients in the control arm received cMRI (median [IQR], 1.28 hours [0.96-1.60] vs 7.76 hours [4.51-11.02], p < 0.001). Of the participants who received both pMRI and cMRI, pMRI as the primary diagnostic tool provided the opportunity to reduce ED length of stay by a median [IQR] of 6.35 [3.53-9.58] hours. Neuroradiologists determined that 94% of pMRI scans were diagnostically adequate. No serious adverse events occurred. We found that 17.8% of pMRI studies had critical findings, including acute ischemic stroke (n=4), mass lesion (n=5), mass effect (n=2), and hydrocephalus (n=1), with 4 strokes and 2 instances of mass effect identified on pMRI before cMRI or CT. Finally, pMRI exhibited high volumetric correlation to cMRI across all 15 measured brain structures (n=55, all r > 0.82).

In this randomized trial, subjects randomized to the portable MRI intervention received MRI more than six hours before those in the conventional MRI control arm. Portable MRI may serve as an efficient and reliable adjunct to emergency neuroimaging and has the potential to substantially reduce ED length of stay and diagnostic delays.