Cardiac failure, severe paresis, urinary incontinence, coma, and atrial fibrillation predicted 1-year mortality after stroke
ACP J Club. 1995 Mar-April;122:46. doi:10.7326/ACPJC-1995-122-2-046
Anderson CS, Jamrozik KD, Broadhurst RJ, Stewart-Wynne EG. Predicting survival for 1 year among different subtypes of stroke: results from the Perth Community Stroke Study. Stroke. 1994 Oct;25: 1935-44.
To ascertain and validate risk factors that predict 1-year mortality in patients with acute stroke of different pathologic subtypes.
Inception cohort assembled in 1989 and 1990 and followed for 1 year (Perth Community Stroke Study).
A community in Western Australia.
Patients in the community who had possible acute cerebrovascular events were assessed (n = 883). Stroke was defined using World Health Organization definitions. 492 patients (52% men, mean age 73 y) with acute stroke (353 had first-ever strokes) were included. Strokes were divided into pathologically distinct subtypes (large-artery occlusive, embolic, lacunar, and boundary zone infarctions; primary intracerebral hemorrhage; and subarachnoid hemorrhage). Data from 321 patients were used to develop a predictive model and data from 171 patients were used to validate it. Follow-up was 100%.
Assessment of prognostic factors
Physical examinations and histories were used to assess patterns of disability, social activity, smoking status, and clinical features at onset.
Main outcome measures
The similar results for first-ever and subsequent strokes were combined. The 28-day mortality rate was 24% and the 1-year mortality rate was 38% (range 0% for boundary zone infarctions [ n = 12] to 100% for subarachnoid hemorrhage [ n = 1]). Multivariate analysis for all strokes (n = 321) showed that baseline 1-year mortality predictors were cardiac failure (relative risk [RR] 6.5, 95% CI 2.8 to 15.1), severe paresis (RR 4.9, CI 1.6 to 15.5), urinary incontinence (RR 3.9, CI 1.4 to 10.6), coma (relative risk [RR] 3.0, CI 1.1 to 8.4), and atrial fibrillation (RR 2.0, CI 1.1 to 3.5). For cerebral infarction (n = 224), the predictors for 1-year mortality were cardiac failure (RR 6.7, CI 2.4 to 18.8), urinary incontinence (RR, 7.3, CI 2.5 to 21.3), atrial fibrillation (RR 2.3, CI 1.0 to 5.1), unpartnered marital status (RR 2.4, CI 1.1 to 5.4), and history of claudication (RR 2.2, CI 1.0 to 4.7).
Cardiac failure, severe paresis, urinary incontinence, coma, and atrial fibrillation predicted 1-year mortality in patients with acute stroke.
Sources of funding: National Health and Medical Research Council of Australia, the Australian Brain Foundation, and the Medical Research Foundation of Royal Perth Hospital.
For article reprint: Dr. C.S. Anderson, Department of Medicine, Clinical Trials Research Unit, University Auckland, Private Bag 92019, Auckland, New Zealand. FAX 64-9-373-1710.
The article by Anderson and colleagues highlights several important points about the outcome of patients with stroke. The first is that severe neurologic deficits are associated with increased mortality. This has been shown previously for both stroke and severe paresis (1-3) and remains a practical variable to be used at the bedside, particularly in relation to persons with occlusion of large arteries and more extensive damage of neural tissue. A second finding of this study, however, is that urinary incontinence and cardiac failure increase mortality. This is reasonable because the former is a sign of bilateral extensive neurologic dysfunction, and the latter is a marker of poor cardiac function. On the other hand, it is of interest that age did not independently influence mortality, a fact that disputes previous findings (1, 4) and that clearly constitutes a topic of practical clinical importance.
A matter of concern is the way this information will be used at the bedside. Clinicians should not give in to the temptation to withhold therapy solely on the basis of predictors of outcome such as those described. A more reasonable approach is to consider each patient's premorbid quality of life as an additional determining factor in starting aggressive therapies. Perhaps the usefulness of the information provided by this study is in determining how long to continue aggressive treatment efforts when patients do not respond favorably.
Camilo R. Gomez, MD
Saint Louis University Medical CenterSt. Louis, Missouri, USA