Imaging essential to distinguish hemorrhagic, ischemic stroke
By Eleanor McDermid
15 June 2010
JAMA 2010; 303: 2280–2286

MedWire News: Several clinical factors can help to gauge whether patients have ischemic or hemorrhagic stroke, but no combination is definitively diagnostic in all patients, say the authors of a systematic review.

Shauna Runchey and Steven McGee (University of Washington, Seattle, USA) cover the subject in a Rational Clinical Examination article in the Journal of the American Medical Association.

Stroke patients presenting within the thrombolysis time window must undergo imaging to exclude hemorrhage. But the team says that clinical impression is helpful in later presenting patients and essential if imaging is not available.

It is even important in patients undergoing thrombolysis, say Runchey and McGee. “According to treatment guidelines, the infusion should be immediately discontinued (and imaging repeated) if severe headache, acute hypertension, nausea, or vomiting develop, a statement implying that these symptoms and signs suggest hemorrhage.”

The systematic review included 19 prospective studies, involving 6438 patients, 24% of whom had hemorrhagic stroke.

Overall, the likelihood of having hemorrhagic stroke increased 6.2 fold if patients were in a coma, 5.0 fold if they had neck stiffness, 4.7 fold if they had seizures, 4.3 fold if their diastolic blood pressure exceeded 110 mmHg, 3.0 fold if they vomited, 2.9 fold if they had a headache, and 2.6 fold if they lost consciousness.

Patients with cervical bruit or prior transient ischemic attack were, respectively, 88% and 66% less likely to have hemorrhage. The presence of peripheral artery disease or a history of atrial fibrillation decreased the likelihood by 59% and 56%, respectively.

The Siriraj score was useful for distinguishing patients with hemorrhagic and ischemic stroke. The score encompasses factors including coma, vomiting, headache, diabetes, and intermittent claudication. Patients were 5.7-fold more likely to have hemorrhagic than ischemic stroke if they had a score higher than 1, and were 71% less likely if they had a score lower than -1.

However, the team notes that some 20% of patients have scores between -1 and 1, which is “diagnostically unhelpful.”

Finally, one study found that clinicians having an overall clinical impression of hemorrhage increased this likelihood in their patients 6.2 fold, whereas if their overall impression was of ischemic stroke the likelihood of hemorrhage was reduced by 72%. The clinicians’ impressions were as accurate as the Siriraj score, even though they did not use specific rules, and were not allowed to classify patients as “uncertain.”

“Neither the clinical impression of experienced clinicians nor the most accurate stroke score can improve the post-test probability of hemorrhage to greater than 50%,” conclude Runchey and McGee.

“While combinations of findings are more predictive than individual findings, diagnostic certainty requires neuroimaging.”

MedWire (www.medwire-news.md) is an independent clinical news service provided by Current Medicine Group, a trading division of Springer Healthcare Limited. © Springer Healthcare Ltd; 2010

Free abstract

Friendly links


Follow me on Twitter