The impact of Magnetic Resonance Imaging (MRI) on ischemic stroke detection and incidence: minimal impact within a population-based study

Dawn Kleindorfer, Jane Khoury, Kathleen Alwell, Charles J. Moomaw, Daniel Woo, Matthew L. Flaherty, Opeolu Adeoye, Simona Ferioli, Pooja Khatri, Brett M. Kissela

There are several situations in which magnetic resonance imaging (MRI) might impact whether an cerebrovascular event is considered a new stroke. These include clinically non-focal events with positive imaging for acute cerebral infarction, and worsening of older symptoms without evidence of new infarction on MRI. We sought to investigate the impact of MRI on stroke detection and stroke incidence, by describing agreement between a strictly clinical definition of stroke and a definition based on physician opinion, including MRI imaging findings. All hospitalized strokes that occurred in five Ohio and Northern Kentucky counties (population 1.3 million) in the calendar year of 2005 were identified using ICD-9 discharge codes 430-436. The two definitions used were: "clinical case definition" which included sudden onset focal neurologic symptoms referable to a vascular territory for >24 h, compared to the "best clinical judgment of the physician definition", which considers all relevant information, including neuroimaging findings. The 95 % confidence intervals (CI) for the incidence rates were calculated assuming a Poisson distribution. Rates were standardized to the 2000 U.S. population, adjusting for age, race, and sex, and included all age groups. There were 2403 ischemic stroke events in 2269 patients; 1556 (64 %) had MRI performed. Of the events, 2049 (83 %) were cases by both definitions, 185 (7.7 %) met the clinical case definition but were non-cases in the physician's opinion and 169 (7.0 %) were non-cases by clinical definition but were cases in the physician's opinion. There was no significant difference in the incidence rates of first-ever or total ischemic strokes generated by the two different definitions, or when only those with MRI imaging were included. We found that MRI findings do not appear to substantially change stroke incidence estimates, as the strictly clinical definition of stroke did not significantly differ from a definition that included imaging findings. Including MRI in the case definition "rules out" almost the same number of strokes as it "rules in".