Aim: The charts of 388 patients admitted for a recent stroke at the University of Kinshasa Hospital, Department of Medicine, were analysed to assess the case fatality rate and its determinants.
Results: During the hospitalization period (average 31 days), 169 deaths (44%) were recorded. Mortality was higher (P<0.05) in women (54%) than in men (42%), in haemorrhagic (61%) than in ischaemic (26%) strokes, and in patients with depressed consciousness, motor deficits, tachycardia or atrial fibrillation on admission. Mortality increased with advancing age, the depth of coma, the severity of motor deficit and the plasma glucose level on admission. Plasma glucose, urea and the leucocyte count were higher in patients who died than in survivors. Patients with and without a history of hypertension, diabetes or stroke recurrence had similar mortality rates. The latter tended to be higher in smokers and alcohol consumers. Using logistic regression analysis, depressed consciousness (P<0.001), tachycardia (P<0.001), plasma urea on admission (P<0.05) and a history of alcohol consumption (P<0.05) emerged as significant predictors of mortality. Admission systolic/diastolic blood pressure was similar in those who died (177 +/- 41/104 +/- 26 mmHg) and in survivors (171 +/- 41/101 +/- 23 mmHg). However, in the whole study population and in subgroups of patients who were comatose or conscious on admission, mortality showed a J-shaped relationship to admission systolic blood pressure with low rates in the range 160-199 mmHg before and after adjustment for age and sex, higher rates being observed for higher or lower blood pressure ranges.
Conclusions: Depressed consciousness, ECG abnormalities, delayed admission and elevated plasma urea predict a high case fatality rate whereas systolic blood pressure within the range 160-199 mmHg appears to be optimal for survival in the patient group studied.