Objectives: To estimate the effects of age on 6-month survival of critically ill patients with cancer.
Design: Prospective cohort study analyzed using Cox proportional hazard models.
Setting: Ten-bed oncologic medical-surgical intensive care unit.
Patients: Eight hundred sixty-two patients with cancer, excluding bone marrow transplant patients.
Interventions: None.
Measurements and main results: The mean age was 57.8+/-16.2 yrs. The hospital and 6-month mortality rates were 48% and 58%, respectively. Age was independently associated with increased mortality (hazard ratio, 1.015; 95% confidence interval, 1.009-1.021). Martingale residual analysis, however, suggested an inflection point in the effect of age, with an upward trend for patients aged>60 yrs. Therefore, patients were stratified in two groups: young (<or=60 yrs, n=431, 50%) and elderly (>60 yrs, n=431, 50%). In young patients, uncontrolled cancer, mechanical ventilation, and number of organ failures were associated with poor outcome, whereas surgery before intensive care unit admission was protective. The variables associated with increased mortality for elderly patients were performance status 3-4, uncontrolled cancer, number of organ failures, and the presence of a severe comorbidity. In this group, age was associated with a lower survival rate. In general, the effect of covariates on the outcome was higher in the elderly group.
Conclusions: Aging was associated with increased mortality, especially for patients>60 yrs. The severity of organ failures and the presence of uncontrolled cancer were the main predictive factors, but there were important differences among the outcome predictors for young and elderly patients. Our results suggest that selected older patients with cancer can benefit from intensive care.