American Surgical Association Home American Surgical Association Annual Meeting
Search Meeting Site Only

Current Meeting Home
Past & Future Meetings




Back to Annual Meeting Program

Failure to Rescue or Withdrawal of Support: Explaining the Excessive Mortality of Elderly DNR Patients Undergoing Emergency General Surgical Procedures.
John E Scarborough*, Kyla M Bennett*, Theodore N. Pappas, Sandhya A. Lagoo-Deenadayalan*
Duke University Medical Center, Durham, NC

Objective: Preoperative do-not-resuscitate (DNR) status is a known predictor of postoperative mortality. Our objective was to determine potential causes for this association.
Methods: Patients age ≥ 65 years undergoing emergency operation for one of ten common indications were extracted from the 2005-2009 National Surgical Quality Improvement database. Propensity score techniques were used to match patients with and without preoperative DNR directives for operative diagnosis, patient age and comorbidities, preoperative physiologic status, and procedure complexity.
Results: 19,409 patients were evaluated for analysis (845 DNR, 18,564 non-DNR), 1,598 of whom were included in the matched cohort. No significant difference existed between DNR and non-DNR patients for any of the preoperative or intraoperative variables. DNR patients were more likely to die postoperatively than non-DNR patients despite the fact that they were less likely to develop a major complication (Table). DNR patients were less likely to undergo reoperation, and were more likely to die in the absence of a major complication, than non-DNR patients.
OutcomeDNRNon-DNRP Value*AOR **(95% CI)
Major Complication283/799 (35.4%)315/799 (39.4%)0.030.84 (0.68,1.04)
Reoperation59/799 (7.4%)97/799 (12.1%)0.0020.58 (0.40,0.82)
Overall283/799 (35.0%)192/799 (24.0%)<0.00011.78 (1.41,2.25)
If Major Complication161/283 (56.9%)125/315 (39.7%)0.0012.53 (1.41,4.73)
If No Major Complication122/516 (23.6%)67/484 (13.8%)0.021.63 (1.07,2.50)
*Using McNemar’s Chi-Square tests **Using conditional logistic regression

Conclusions: The independent effect of preoperative DNR status on mortality after emergent general surgery cannot be explained by comorbid conditions or by a higher incidence of major complications. Our findings suggest that elderly patients with preoperative DNR directives may be managed less aggressively in the postoperative period than patients without such directives, leading to a much higher mortality than would be expected based on clinical predictors.

Back to Annual Meeting Program


© 2022 American Surgical Association. All Rights Reserved. Privacy Policy.