We acessed the potential impact of diabetes mellitus and hyperglycemia on early prognosis in a study of 100 cases of CT proved brain infarction. Main end-points of this study were 30-days case-fatality rates and the probability of early discharge or prolonged hospitalization. Proportions of patients exhibiting diverse evolutions were compared by Fisher's test and, when the number of observations permitted, by chi-square test. Diabetes mellitus was diagnosed by history in 15.6% (17 patients). In 86 cases analysed within 72 hours of admission, hyperglycemia (glucose level higher than 110 mg/dL.) was detected in 67.4% (including all 14 diabetic patients evaluated). A nonsignificant increase in case fatality rates was detected in diabetic patients (35.3% vs 21.7%; p=0.18), but these patients probably were discharged early less frequently (23.5% vs 51.1%; p==0.06). Compared to the normoglycemic patients, hyperglycemic patients also fared worse, when considering altogether the early discharge and case fatality rates and the need for prolonged hospitalization (p=0.06). They also exhibited a nonsignificant increase in early case fatality rate (31% vs 15.4%; p=0.21) and a markedly reduced chance of obtaining early hospital discharge (37.9% vs 65.4%; p=0.03). Case fatality rates were maximal in diabetic hyperglycemic patients (42.9%) and seemed intermediate in non-diabetic hyperglycemic patients (27.3%). Considering the three end-points evaluated, the apparent differences between non-diabetic hyper and normoglycemic patients did not reach statistical significance (p=0.19). These results suggest that both diabetes mellitus and hyperglycemia «per se» have deleterious effects on the clinical evolution of patients with cerebral infarction. It is now time to consider a prospective study on the potential benefits of early correction of hyperglycemia in these patients.
brain infarction; diabetes mellitus; hyperglycemia; prognosis