Use of do-not-resuscitate orders in an intensive care setting.

STUDY OBJECTIVES Do-not-resuscitate (DNR) orders have been espoused for the enhancement of patient autonomy, avoidance of futile medical intervention, and cost containment. Outcomes of cardiopulmonary resuscitation (CPR) in the intensive care setting have been dismal, with few patients surviving to discharge. This study compares patients who died in medical and surgical ICUs in a DNR status with those who died after attempted CPR. DESIGN Retrospective chart review of all patients who died in the medical and surgical ICU in a 2-year period. MEASUREMENTS AND RESULTS A total of 195 cases were reviewed during the specified time period; 108 patients had undergone attempted resuscitation, and 87 patients died in a DNR status. There were no significant differences when preadmission disability, source of admission, location (medical ICU vs surgical ICU), chronic medical conditions, acute diagnosis, sex, and weight were considered. Patients who were designated "DNR" were significantly older than patients who underwent CPR (mean age, 65.7 years vs 58.9 years; p = 0.005). The DNR-designated patients were in general more severely ill as measured with the APACHE II system (mean score, 23.5 vs 20.7; p = 0.004), which was accounted for primarily by greater alterations in level of consciousness as measured with the Glascow Coma scale (mean score, 10.0 vs 12.1; p = 0.001). CONCLUSIONS Among patients dying in the medical and surgical ICUs in the authors' institution, only age and level of consciousness discriminated patients who died in a DNR status from those who died after attempted CPR.