OBJECTIVE

To investigate the epidemiology of and possible factors associated with end-of-life decisions in a surgical intensive care unit.

DESIGN

Analysis of prospectively collected data.

SETTING

University hospital surgical intensive care unit.

PATIENTS

All patients admitted to the surgical intensive care unit between September 2002 and July 2006.

MEASUREMENTS AND MAIN RESULTS

During the study period, 14,720 patients were admitted to the surgical intensive care unit (61.8 male; mean age, 62 yrs). The prevalence of end-of-life decisions was 2.7% (n = 398); 230 patients (1.6%) had a do-not-resuscitate order, 90 (0.6%) had a decision to withhold therapy, and 78 (0.5%) had a decision to withdraw life-supportive therapy. Patients with end-of-life decisions had higher severity scores on the day of intensive care unit admission, were mostly unplanned admissions, were older, and were more commonly referred from the emergency room or other hospitals compared to those who did not have an end-of-life decision. The prevalence of end-of-life decisions increased significantly with the severity of sepsis. An end-of-life decision was made for 29% of the patients who died in the intensive care unit. Intensive care unit and hospital mortality rates were 6.1% and 10.3%, respectively, overall, and 65.1% and 82.2%, respectively, in patients with an end-of-life decision. In multivariate analysis, older age, admission from another hospital, cirrhosis, sepsis syndromes, simplified acute physiology score II, and sequential organ failure assessment scores were independently associated with end-of-life decisions.

CONCLUSIONS

Twenty-nine percent of patients who die in the surgical intensive care unit have an end-of-life decision. Severe sepsis/septic shock was associated with a 16-fold increased likelihood of having an end-of-life decision.