SAN FRANCISCO – Elderly patients with preexisting Do Not Resuscitate directives appear to be less likely to pursue rescue from complications following emergency surgery than similar patients without such orders, according to an analysis of data from the National Surgical Quality Improvement Program.
When patients with preoperative DNRs were propensity-matched with non-DNR patients, major complication rates were similar – 42% for the DNR group and 41% for the non-DNR group. However, 37% of DNR patients died, compared with 22% of non-DNR patients, Dr. John E. Scarborough reported at the annual meeting of the American Surgical Association. The investigators adjusted for baseline differences in level of illness to create a propensity-matched cohort of 1,053 patients in each group.
"While we called this outcome failure-to-rescue, we believe that term to be misleading. The term implies that rescue from complications is attempted but is unsuccessful. ... We had no reason to believe that the DNR patients in this well-matched cohort were any less capable of being rescued than non-DNR patients," Dr. Scarborough said. "Instead, we believe that the DNR patients in the matched cohort were less likely than non-DNR patients to pursue rescue from complications."
This conclusion is supported by the finding that DNR patients were significantly less likely to undergo reoperation within 30 days of the index procedure (odds ratio, 0.67).
The authors used participant files from the National Surgical Quality Improvement Program (NSQIP) for 2005-2010, involving medical records for 25,558 patients. These patients were at least 65 years old and underwent an emergency operation for one of 10 common surgical diagnoses. The primary predictor variable was preoperative DNR status, which was defined as "an order signed or cosigned by an attending physician in the 30 days prior to surgery ... regardless of whether the DNR order was subsequently rescinded immediately before the index operation." Other predictor variables included patient demographics, chronic comorbid disease burden, acute physical condition at presentation, and complexity of the emergency operation.
Outcome variables included the 30-day postoperative mortality rate and the 30-day major complication rate – organ/space surgical site infection, wound dehiscence, deep vein thrombosis, pulmonary embolism, pneumonia, reintubation, ventilator use longer than 48 hours, cardiac arrest, myocardial infarction, sepsis, shock, coma longer than 24 hours, prosthetic/graft failure, and bleeding. The failure-to-rescue rate was defined as the mortality rate among patients who had one or more major complications.
A total of 1,061 patients had DNR orders, and 24,497 patients did not. The overall 30-day mortality rate for patients with a DNR order was 37% (395/1,061). The overall 30-day morbidity for patients with a DNR was 42% (446/1,061).
Patients with DNR orders were older: 22% were at least 90 years of age, compared with 5% of the non-DNR patients. They were also sicker, with significantly greater rates of non–independent functional status, cognitive dysfunction, known malignancy, congestive heart failure, chronic obstructive pulmonary disease, ASA physical status class 4, preoperative hypoalbuminemia, and septic shock.
"Although the DNR patients were sicker, we did not find any overt evidence that they were treated less aggressively than non-DNR patients in the preoperative period," said Dr. Scarborough, of the department of surgery at Duke University in Durham, N.C. There was no significant difference between DNR and non-DNR patients in terms of preoperative mechanical ventilation (6% vs. 5%, respectively); and DNR patients were significantly more likely to receive a preoperative transfusion.
There was also no indication that DNR patients were treated less aggressively in the operating room. Operative time was significantly longer for DNR patients, and DNR patients underwent procedures at least as complex as, if not more than, procedures for non-DNR patients.
Invited discussant Dr. Ronnie A. Rosenthal asked how these data could be used to improve the way families are counseled before operations involving elderly patients with DNRs. Dr. Rosenthal is surgeon-in-chief at the VA Connecticut Healthcare System in New Haven.
"What we hope this study provides is a more reliable and sturdy resource for surgeons to counsel such patients than merely explaining to them what the average outcomes are," said Dr. Scarborough. He noted that the oncology literature suggests that patients who better understand their prognosis are in a better position to evaluate whether they want to pursue more aggressive treatment or treatments that have a lot of side effects.
Dr. Norman Estes, chair of the surgery department at the University of Illinois in Peoria, questioned how much of a role surgeons should play in advance planning. "I think that sometimes the advance directive creates a self-fulfilling prophecy for the patient."
Dr. Scarborough noted that advance directives need to be signed by the attending physician. "As to whether the surgeon should be more engaged in the conversation, I guess I would say that it depends on the surgeon. This is a very delicate conversation and obviously one that requires a fair amount of time," he said. Other physicians – such as geriatricians and palliative care physicians – are often more skilled at handling these conversations.