Predictors of mortality and disability in stroke.

The increasingly accurate prediction of survival and functional recovery in patients with stroke will be of value in planning both their individual management and the health and social services needed. To establish the independent predictive effects of a range of personal and clinical characteristics, data on 900 patients admitted to Northwick Park Hospital with stroke were analysed by stepwise multiple regression. Older patients who lose consciousness at the outset and show signs of multiple neurological deficits, abnormal pupils, and conjugate deviation of the eyes are more likely to die within a year than those without these characteristics. Those who survive the acute episode and are discharged alive are more likely to die within a year if they are old and have sensory loss with severe physical disability. Older female patients who are incontinent, lose consciousness at the onset of stroke, sustain extensive motor deficits in combination with other neurological deficits, and have residual disabilities from previous strokes are particularly likely to be severely disabled on discharge from hospital. Routinely collected clinical data enable useful forecasts about mortality and disability after stroke. The accuracy of these forecasts can probably be improved further.