Understanding Real-World MRI Utilization for ED Dizziness: Ava L. Liberman, MD

Amid ongoing questions about how best to evaluate dizziness in the emergency department (ED), new research is taking a closer look at when advanced imaging may matter most. In this interview from the International Stroke Conference 2026, Ava L. Liberman, MD, assistant professor of emergency medicine at Weill Cornell Medicine, discusses a large, population-based analysis examining brain magnetic resonance imaging (MRI) use among patients presenting to the ED with dizziness or vertigo who were ultimately discharged home.

Drawing on data from nearly 1 million ED visits across 11 states, the study tracks trends in MRI utilization, associated ED length of stay, and the risk of stroke hospitalization within 30 days. Liberman emphasizes that the work is descriptive rather than prescriptive, aimed at characterizing real-world practice patterns and their downstream associations, while underscoring the continued importance of careful bedside examination and guideline-informed decision-making in this common but diagnostically challenging patient population.

Transcript

Your analysis suggests that MRI use in the ED is associated with substantially longer visit times. From a population-level perspective, how should we think about that tradeoff when evaluating diagnostic strategies for patients presenting with dizziness?

Yeah, well, thank you for asking. So it's important to clarify that the work I did was really only on emergency medicine. Patients were being discharged from the ED, so the increased length of time is among all ED patients who were discharged from the emergency setting, not patients who were subsequently admitted. So I think when thinking about the time differential between patients who got an MRI in the ED and those who did not, the contrast could be to patients who are hospitalized and admitted and wait overnight for an MRI. So I think the time metric very much depends on the population that's being studied. And so I think to your question of, like, is that a good idea? I mean, this is a purely descriptive study. I'm just kind of describing what I'm seeing.

Advanced imaging in the emergency setting is often framed as a cost driver. Based on your findings, how might earlier MRI use influence downstream utilization, missed diagnoses, or long-term costs that aren’t captured in a single ED encounter?

This was not a cost-effectiveness analysis. This is really just a descriptive study of utilization of [MRI] among patients with dizziness or vertigo complaints in the ED who are discharged to home. So it doesn't really address any cost-effectiveness-like issues. And I think when you know we're right to do a cost-effectiveness analysis, the things I would look at are the cost of admission, as opposed to an MRI in the ED and being discharged to home, and the cost of a missed stroke, which may be nontrivial. A patient can come in with a minor stroke and then subsequently be hospitalized with a major stroke. And so I didn't address cost-effectiveness at all. I think there are important questions in imaging utilization in the ED, but this study was really about people who come in and then are sent home who don't have a lot of medical comorbidities, broadly speaking.

In your view, what is the biggest barrier preventing hospitals from implementing the more aggressive MRI screening your study suggests?

Yeah, I'm not advocating increased MRI screening in dizziness patients at all. And actually, the GRACE-3 guidelines are worth looking at, which came out a couple years ago, which is an emergency medicine-based way to sort of think about patients with vertiginous symptoms, so within the subcategory of acute vestibular syndrome, which are patients who may have stroke as the cause of their acute vestibular syndrome—physical exam maneuvers are really key to diagnosis. And I think MRI in settings should always be informed by the physical exam findings you have on exam. So this was really not about getting MRIs on everybody, or subpopulation, or whatever. I was just kind of describing practice and the increased use of MRIs in this population in the US.

Based on your work here, what is the next big question we should be asking about the intersection of emergency imaging and long-term stroke recovery?

Well, I don't see how imaging has much to do with long-term stroke recovery, aside from preventing stroke misdiagnosis, which could at least impact disability in patients in whom a stroke or minor stroke occurred but was not initially detected. And so I think imaging utilization can help identify those few patients who have infarcts on imaging that are not kind of recognized as such and who aren't on the appropriate sort of secondary stroke prevention strategy. So I think from that standpoint, that's an area of opportunity; patients in whom no one thinks they're having a stroke, but in fact, they are, imaging can sometimes be useful in that small subgroup of patients.

Reference

Liberman AL, Navi BB, Ch’ang J, et al. Use of MRI in emergency department patients with dizziness and association with subsequent stroke. Presented at: International Stroke Conference 2026. February 3-6, 2025; New Orleans, LA. Abstract DP214.