Role of a stroke data bank in evaluating cerebral infarction subtypes: Patterns and outcome of 1,776 consecutive patients from the Besancon Stroke Registry

The purpose of this study was to estimate the frequency of various risk fac tors, courses and outcome of infarct subtypes in a large hospital-based str oke registry. Methods: From 1987 to 1994, 1,776 stroke patients with a firs t-ever infarction were included in the Besancon Stroke Registry. All patien ts were evaluated by a standard protocol (risk factors, stroke onset, strok e courses, clinical characteristics, neuroimaging, Doppler ultrasonography and cardiac investigations). Outcome was evaluated at 30 days using the Ran kin scale. Results: There were 1,012 men (mean age 67.2 +/- 13.7 years) and 764 women (mean age 71.4 +/- 15.6 years). At least two neuroimaging examin ations were performed in 81.4% (n = 1,446) of the patients and an infarct w as visible in 80.9% (n = 1,436). The second neuroimaging examination (CT or MRI) was performed after 8.2 +/- 1.6 days. 85.4% of patients were admitted on the first day of the stroke: 28.3% within 3 h and 48.4% within 6 h. In addition, stroke severity was well correlated with the short time interval between stroke onset and admission. Past medical history of hypertension wa s the major risk factor occurring in 57.5% of a II types of infarction. Whi le diabetes was more frequently found in small deep infarct, atrial fibrill ation and history of heart failure were found in anterior circulation infar cts. The distribution of clinical presentations was conventional. Hemorrhag ic transformation was found in 14.9% of the patients, especially in MCA and PCA infarcts. In all patients, logistic regression analysis determined ind ependent predictive factors for death: clinical deterioration at the 48th h our (OR 7.5, 95% CI 4.9-11.3), initial loss of consciousness (OR 3.3, 95% C I 2.1-4.9), age (OR 1.05, 95% CI 1.03-1.06), complete motor deficit (OR 2.6 , 95% CI 1.7-3.8), history of heart failure (OR 1.9, 95% CI 1.3-3.0), lacun ar syndrome (OR 0.25, 95% CI 0.10-0.60) and regressive stroke onset (OR 0.2 4, 95% CI 0.10-0.52). However, the outcome was clearly correlated with the infarct location. The in-hospital mortality rate was lowest in patients wit h small deep infarct (2.9%) or border zone infarcts (3.4%) and the highest in patients with total middle cerebral artery infarct (47.4%) or multiple i nfarcts (27.6%). Conclusion: Our registry appears to be a useful tool to un derstand the course and outcome of a large group of nonselected patients wi th subtypes of infarction. It can also help to analyze the influence of spe cific stroke management in the different categories of stroke types. Copyright (C) 2000 S. Karger AG, Basel.