Background: Do-not-resuscitate (DNR) orders instruct medical personnel to forego cardiopulmonary resuscitation in the event of cardiopulmonary arrest, but they do not preclude surgical management. Several studies have reported that DNR status is an independent predictor of 30-day mortality; however, the etiology of increased mortality remains unclear. We hypothesized that DNR patients would demonstrate increased postoperative mortality, but not morbidity, relative to non-DNR patients undergoing the same procedures.
Methods: Using the American College of Surgeons National Surgical Quality Improvement Program database for 2007-2013, we performed a retrospective analysis to compare DNR and non-DNR cohorts matched by the most common procedures performed in DNR patients. We employed univariable and multivariable logistic regression to characterize patterns of care in the perioperative period as well as identify independent risk factors for increased mortality and assess for the presence of "failure to rescue."
Results: The most common procedures performed on DNR patients were emergent and centered on immediate symptom relief. When adjusting for preoperative factors, DNR patients were still found to have increased incidence of postoperative mortality (odds ratio 2.54 [2.29-2.82], P < .001) but not postoperative morbidity at 30 days. In addition, cardiopulmonary resuscitative measures and unplanned intubation were found to be less frequent in the DNR cohort.
Conclusions: These findings suggest that increased mortality is the result of adherence to goals of care rather than "failure to rescue."