Relationship Between ICU Length of Stay and Long-Term Mortality for Elderly ICU Survivors

OBJECTIVES:To evaluate the association between length of ICU stay and 1-year mortality for elderly patients who survived to hospital discharge in the United States. DESIGN:Retrospective cohort study of a random sample of Medicare beneficiaries who survived to hospital discharge, with 1- and 3-year follow-up, stratified by the number of days of intensive care and with additional stratification based on receipt of mechanical ventilation. INTERVENTIONS:None. PATIENTS:The cohort included 34,696 Medicare beneficiaries older than 65 years who received intensive care and survived to hospital discharge in 2005. MEASUREMENTS AND MAIN RESULTS:Among 34,696 patients who survived to hospital discharge, the mean ICU length of stay was 3.4 days (± 4.5 d). Patients (88.9%) were in the ICU for 1–6 days, representing 58.6% of ICU bed-days. Patients (1.3%) were in the ICU for 21 or more days, but these patients used 11.6% of bed-days. The percentage of mechanically ventilated patients increased with increasing length of stay (6.3% for 1–6 d in the ICU and 71.3% for ≥ 21 d). One-year mortality was 26.6%, ranging from 19.4% for patients in the ICU for 1 day, up to 57.8% for patients in the ICU for 21 or more days. For each day beyond 7 days in the ICU, there was an increased odds of death by 1 year of 1.04 (95% CI, 1.03–1.05) irrespective of the need for mechanical ventilation. CONCLUSIONS:Increasing ICU length of stay is associated with higher 1-year mortality for both mechanically ventilated and non–mechanically ventilated patients. No specific cutoff was associated with a clear plateau or sharp increase in long-term risk.