The Stroke Data Bank

We studied 1,273 patients with ischemic cerebral infarction who were entered into the Stroke Data Bank, a prospective, observational study involving four university hospitals and the Biometry and Field Studies Branch of the National Institute of Neurological Disorders and Stroke. Forty patients had noniatrogenic recurrent stroke within 30 days after the index cerebral infarction. Using life tables, the 30-day cumulative±SE risk of early recurrence for all infarctions was 3.3±0.4%. The risk of early recurrence was greatest for atherothrombotic infarction (7.9±2.2%, eight of 113 patients) and least for lacunar infarction (2.2±1.2%, eight of 337 patients). Both cardioembolic infarction (4.3±0.9%, 10 of 246 patients) and infarction of undetermined cause (3.0±0.5%, 14 of 508 patients) had intermediate risks. History of hypertension and diabetes mellitus, as well as diastolic hypertension and elevated blood sugar concentration at admission, were associated with early recurrence. Logistic regression analysis estimated the risk of early recurrence to be 8.56% in those with coexisting hypertension and a glucose concentration of 300 mg/dl versus 0.77% in the absence of these two abnormalities. Early recurrence was associated with longer median duration of initial hospital stay (27 vs. 14 days) and a higher 30-day case-fatality rate (20% vs. 7.4%). Increased weakness scores were associated with early recurrent stroke. Identification of the determinants of early recurrent stroke may lead to better secondary prevention and may help select high-risk patients for further study. (Stroke 1989;20:983-989)