Predictors of short- and long-term mortality in critically ill, older adults admitted to the emergency department: an observational study.

ABSTRACT

BACKGROUND: In the future, we can expect an increase in older patients in emergency departments (ED) and acute wards. The main purpose of this study was to identify predictors of short- and long-term mortality in the ED and at hospital discharge. METHODS: This is a retrospective, observational, single-center, cohort study, involving critically ill older adults, recruited consecutively in an ED. The primary outcome was mortality. All patients were followed for 6.5-7.5 years. The Cox proportional hazards model was used. RESULTS: Regarding all critically ill patients aged ≥ 70 years and identified in the ED (n = 402), there was a significant association between mortality at 30 days after ED admission and unconsciousness on admission (HR 3.14, 95% CI 2.09-4.74), hypoxia on admission (HR 2.51, 95% CI 1.69-3.74) and age (HR 1.06 per year, 95% CI 1.03-1.09), (all p < 0.001). Of 402 critically ill patients aged ≥ 70 years and identified in the ED, 303 were discharged alive from hospital. There was a significant association between long-term mortality and the Charlson Comorbidity Index (CCI) > 2 (HR 1.90, 95% CI 1.46-2.48), length of stay (LOS) > 7 days (HR 1.72, 95% CI 1.32-2.23), discharge diagnosis of pneumonia (HR 1.65, 95% CI 1.24-2.21) and age (HR 1.08 per year, 95% CI 1.05-1.10), (all p < 0.001). The only symptom or vital sign associated with long-term mortality was hypoxia on admission (HR 1.70, 05% CI 1.30-2.22). CONCLUSIONS: Among critically ill older adults admitted to an ED and discharged alive the following factors were predictive of long-term mortality CCI > 2, LOS > 7 days, hypoxia on admission, discharge diagnosis of pneumonia and age. The following factors were predictive of mortality at 30 days after ED admission unconsciousness on admission, hypoxia and age. These data might be clinically relevant when it comes to individualized care planning, which should take account of risk prediction and estimated prognosis.