Original Article
A Model to Prevent Fibrinolysis in Patients with Stroke Mimics

https://doi.org/10.1016/j.jstrokecerebrovasdis.2011.04.018Get rights and content

Background

Many patients with stroke-mimicking conditions receive treatment with intravenous fibrinolysis (IVF), a treatment associated with potentially serious complications. We sought to determine if any clinical or radiographic characteristics can help predict stroke mimics among IVF candidates.

Methods

This retrospective study was carried out at a single institution. Patients treated with intravenous recombinant tissue plasminogen activator (rt-PA; n = 193) were divided into 3 categories: acute ischemic stroke (n = 142), aborted stroke (n = 21), and stroke mimics (n = 30). Analysis of variance and the chi-square test were used to assess differences, while logistic regression models were computed to predict groups.

Results

Mimics treated with rt-PA did not experience complications (intracranial bleeding, systemic hemorrhage, or angioedema), and had better neurologic and functional outcomes than stroke patients (P < .05). Several variables helped differentiate strokes from mimics, including atherosclerosis on computed tomographic angiography (odds ratio [OR] 23.6; 95% confidence interval [CI] 8.4-66.2), atrial fibrillation (OR 11.4; 95% CI 1.5-86.3), age >50 years (OR 7.2; 95% CI 2.8-18.5), and focal weakness (OR 4.15; 95% CI 1.75-9.8). Other variables decreased chances of stroke: migraine history (OR 0.05; 95% CI 0.01-0.4), epilepsy (OR 0.13; 95% CI 0.02-0.8), paresthesia (OR 0.1; 95% CI 0.04-0.3), and precordialgia (OR 0.045; 95% CI 0.002-0.9). A regression model using focal weakness, computed tomographic angiography findings, and precordialgia had a 90.2% predictive accuracy.

Conclusions

IVF has low complication rates in stroke mimics. Certain clinical characteristics appear predictive of stroke mimics, particularly normal computed tomographic angiography. If confirmed, this may help prevent giving IVF to patients without stroke.

Section snippets

Study Design

Records from all patients treated with rt-PA from 2007 to 2008 at a tertiary referral center were retrospectively reviewed. The local institutional review board approved this study. Patients were divided into 3 groups: AIS, stroke mimics, and aborted AIS (AbS). AIS was defined as sudden focal neurologic dysfunction with evidence of acute brain ischemia on diffusion-weighted images (DWIs) or serial head computed tomographic (CT) scans. AbS was defined as instances of sudden focal neurologic

Characteristics of Study Subjects

Of 193 patients receiving IV rt-PA, 142 had AIS, 21 had AbS, and 30 had mimics. MRI scans were not obtained in 30 AIS patients. Of these, 27 had large artery occlusion on CTA or MRA, 1 had evolving hypodensities on serial head CT scan, and no reasons were stated for not pursuing MRI in the remaining cases. Five stroke mimics did not receive MRI: 1 left against medical advice, 1 had a pacemaker, and no reasons were stated for the remainder. Two AbS patients did not receive MRI, both because of

Discussion

These data suggest that it is possible to identify stroke mimics using several clinical variables that are usually available on initial evaluation in the emergency room. Focal weakness, history of AF, and abnormal CTA increased the odds of actual stroke, whereas a history of migraine or epilepsy and complaints of paresthesia increased chances of a stroke mimic. Three variables—atherosclerotic changes on CTA, focal weakness, and chest pain—had high accuracy at forecasting AIS. The described

Acknowledgments

We are indebted to Dr. Jeffrey Saver for his thoughtful comments.

References (15)

There are more references available in the full text version of this article.

Cited by (0)

View full text