BackgroundThe purpose of this study was to identify predictors of 3-month mortality in critically ill older persons under medical care and to assess the clinical impact of an ICU stay on physical and cognitive dependence and subjective health status in survivors.MethodsWe conducted a prospective observational cohort study including all older persons 75 years and older consecutively admitted into ICU during a one-year period, except those admitted after cardiac arrest, All patients were followed for 3 months or until death. Comorbidities were assessed using the Charlson index and physical dependence was evaluated using the Katz index of Activity of Daily Living (ADL). Cognitive dependence was determined by a score based on the individual components of the Lawton index of Daily Living and subjective health status was evaluated using the Nottingham Health Profile (NHP) score.ResultsOne hundred patients were included in the analysis. The mean age was 79.3 ± 3.4 years. The median Charlson index was 6 [IQR, 4 to 7] and the mean ADL and cognitive scores were 5.4 ± 1.1 and 1.2 ± 1.4, respectively, corresponding to a population with a high level of comorbidities but low physical and cognitive dependence. Mortality was 61/100 (61%) at 3 months. In multivariate analysis only comorbidities assessed by the Charlson index [Adjusted Odds Ratio, 1.6; 95% CI, 1.2-2.2; p < 0.003] and the number of organ failures assessed by the SOFA score [Adjusted Odds Ratio, 2.5; 95% CI, 1.1-5.2; p < 0.02] were independently associated with 3-month mortality. All 22 patients needing renal support after Day 3 died. Compared with pre-admission, physical (p = 0.04), and cognitive (p = 0.62) dependence in survivors had changed very little at 3 months. In addition, the mean NHP score was 213.1 ± 132.8 at 3 months, suggesting an acceptable perception of their quality of life.ConclusionsIn a selected population of non surgical patients 75 years and older, admission into the ICU is associated with a 3-month survival rate of 38% with little impact on physical and cognitive dependence and subjective health status. Nevertheless, a high comorbidity level (ie, Charlson index), multi-organ failure, and the need for extra-renal support at the early phase of intensive care could be considered as predictors of death.