Predictors of survival after cerebral infarction: importance of cardiac factors

Abstract

Stroke is the third leading cause of death in much of the developed world. There are two approaches to lowering mortality from stroke: reducing the incidence of stroke and reducing case fatality rate. To determine factors, identified at presentation, that were predictive of mortality in elderly persons with acute cerebral infarction. A consecutive series of 215 elderly persons admitted to hospital with acute cerebral infarction, confirmed by computerised tomography, were followed for an average of one year. A proportional hazards model was used to identify predictors of mortality. The following variables were identified on univariate analysis as predictors of death: raised serum creatinine, interstitial oedema on chest radiograph, low score on the Mini-Mental State Examination, atrial fibrillation, advanced age, cardiomegaly, raised leucocyte count, pulmonary venous congestion and homonymous hemianopia. Interstitial oedema was the most powerful predictor of death among the four measures of cardiac status. Two successful multivariate models included: 1) interstitial oedema, serum creatinine, age and homonymous hemianopia; 2) serum creatinine and Mini-Mental State Score. Further studies are required to assess the relationship between elevated serum creatinine and poor survival following stroke.