Long-term survival from intensive care: a review

Objective: To determine whether the long-term benefit of an ICU requires prolonged patient follow-up we reviewed long-term survival of patients from general ICUs. Method: We carried out a computerised search of online databases Medline (1966-2004), Embase (1966-2004) and Cochrane Library (1966-2004) for studies reporting patients' long-term survival for greater than 12 months from general ICUs. Selected studies: We identified 19 studies that met the selection criteria. The casemix and severity of illness varied. Differences included the services provided, investigator inclusion/exclusion criteria and proportion of medical patients (range 13-79%). Results: Mean reported ICU length of stay was 5.3 days. The study initiation time for follow-up varied (mostly from time of ICU admission), as did the duration of follow-up (16 months-13 years). ICU and hospital mortality rates ranged from 8% to 33% and 11% to 64%, respectively. The reported 5-year mortality ranged from 40% to 58%. Conclusions: Well designed studies on long-term outcomes are needed to demonstrate the value of intensive care. Deficiencies in design, methodology, and reporting make interpretation and comparison difficult. Recommendations are made for the reporting of outcome from the ICU. Optimum duration of follow-up has not been determined.