Determinants of long-term survival after intensive care

Abstract

Objective: To identify prognostic determinants of long-term survival for patients treated in intensive care units (ICUs) who survived to hospital discharge. Design: An ICU clinical cohort linked to state-wide hospital records and death registers. Setting and Patients: Adult patients admitted to a 22-bed ICU at a major teaching hospital in Perth, Western Australia, between 1987 and 2002 who survived to hospital discharge (n = 19,921) were followed-up until December 31, 2003. Measurements: The main outcome measures are crude and adjusted survival. Main Results: The risk of death in the first year after hospital discharge was high for patients who survived the ICU compared with the general population (standardized mortality rate [SMR] at 1 yr = 2.90, 95% confidence interval [CI] 2.73-3.08) and remained higher than the general population for every year during 15 yrs of follow up (SMR at 15 yrs = 2.01, 95% CI 1.64-2.46). Factors that were independently associated with survival during the first year were older age (hazard ratio [HR] = 4.09; 95% CI 3.20-5.23), severe comorbidity (HR = 5.23; 95% CI 4.25-6.43), ICU diagnostic group (HR range 2.20 to 8.95), new malignancy (HR = 4.60; 95% CI 3.68-5.76), high acute physiology score on admission (HR = 1.55; 95% CI 1.23-1.96), and peak number of organ failures (HR = 1.51; 95% CI 1.11-2.04). All of these factors were independently associated with subsequent survival for those patients who were alive 1 yr after discharge from the hospital with the addition of male gender (HR = 1.17; 95% CI 1.10-1.25) and prolonged length of stay in ICU (HR = 1.42; 95% CI 1.29-1.55).Conclusions: Patients who survived an admission to the ICU have worse survival than the general population for at least 15 yrs. The factors that determine long-term survival include age, comorbidity, and primary diagnosis. Severity of illness was also associated with long-term survival and this suggests that an episode of critical illness, or its treatment, may shorten life-expectancy.