Background and Purpose—Although the pathophysiological heterogeneity of stroke may be highly relevant to the development of acute-phase therapies, discriminating between ischemic stroke subtypes soon after onset remains a challenge. We conducted a study of the accuracy of a clinical diagnosis of lacunar stroke in the first 6 hours after symptom onset.

Methods—We analyzed data from 1367 patients in the Glycine Antagonist In Neuroprotection (GAIN) Americas trial. The Trial of ORG10172 in Acute Stroke Treatment (TOAST) category "small vessel (lacunar)" disease at day 7 or at hospital discharge was used as the reference standard to determine the accuracy of a diagnosis of a lacunar stroke made within 6 hours of symptom onset using the Oxfordshire Community Stroke Project (OCSP) classification "LACS." Outcome was analyzed by comparing the proportions of patients classified as "LACS" at baseline or "small vessel (lacunar)" at 7 days who were dead or dependent at 3 months.

Results—The positive predictive value of an OCSP diagnosis of a lacunar stroke was 76% (95% CI: 69% to 81%; sensitivity 64% [95% CI: 58% to 70%]; specificity 96% [95% CI: 95% to 97%]; negative predictive value 93% [95% CI: 92% to 94%]; accuracy 91% [95% CI: 89% to 92%]). The 3-month outcomes of patients classified as either OCSP "LACS" within 6 hours of onset or TOAST "small vessel (lacunar)" at 7 days were not significantly different.

Conclusions—An OCSP LACS diagnosis made within 6 hours of stroke onset is reasonably predictive of a final diagnosis of "small vessel (lacunar)" disease made using TOAST criteria and has a similar relationship to outcome at 3 months.