Internal medicine residents reported being less likely to consider certain aggressive interventions outside of CPR on patients with do not resuscitate (DNR) and do not intubate (DNI) orders, according to a study.

These findings have researchers worried about a trend of doctors ignoring patient preferences, especially those who may have DNRs but do not want to ignore other treatment options, according to Elizabeth K. Stevenson, MD, of the Division of Pulmonary and Critical Care Medicine, North Shore Medical Center, Salem, Mass., and her colleagues.

“DNR/DNI patients were less likely to receive many invasive procedures, surgical consultations, or transfer to the ICU,” wrote Dr. Stevenson and her colleagues. “[D]ecisions to withhold many types of care not specified in DNR/DNI orders is concerning, given that the majority of patients with a DNR/DNI status in registry studies indicated they would accept other interventions beyond CPR and intubation.”

Researchers surveyed 553 internal medicine residents in the United States using an Internet survey that presented four vignettes describing clinical situations. Participants were asked to rank how likely they would be to employ listed intervention methods, from “strongly agree” to “strongly disagree,” in each scenario (Ann Am Thorac Soc. 2017, Apr;14[4]:536-42) .

Two different versions of the survey were randomly assigned, varying only in terms of which vignettes included patients with a DNR/DNI order.

Of the interventions listed for each scenario, decisions to transfer patients to the intensive care unit and suggest surgery consultations showed the strongest association with code status.

“Residents were significantly less likely to indicate they would provide invasive procedures (including central venous catheter placement, esophagogastroduodenoscopy, colonoscopy, bronchoscopy, dialysis, and surgery consultation) to patients who had a status of DNR/DNI compared with Full Code,” the investigators noted. “In contrast, decisions to pursue noninvasive diagnostic or therapeutic interventions (CT scans, administration of oxygen or intravenous fluids, blood cultures, and initiation of anticoagulation) did not significantly differ by patient code status, with high levels of use across all vignettes.”

In one vignette involving surgical consultation for an 80-year-old woman with septic shock secondary to Clostridium difficile infection, 89.1% of residents recommended a consult for full-care patients, while 77.7% recommended one for a patient with a DNR/DNI (P = .0008).

Despite these findings, 94%-96% of participants reported willingness to consult with patients on their preferences before treatment decisions, which Dr. Stevenson and her peers found somewhat comforting, although it did not completely assuage them.

“Although the ideal approach would be to have more comprehensive discussion and documentation of patients’ goals of care in the outpatient setting, realistically, many patients will neither have had such discussions nor [have] completed advance directives before hospitalization,” investigators wrote.

The study was limited by the size of the sample, which numbered approximately 2% of the active internal medicine residents in the United States. The researchers recognized that these scenarios were theoretical, and that practicing physicians may act differently when faced with a medical situation in real life. The study also was limited by the concentration of respondents within a single program, as shared experiences or teachers may cause similar responses to theoretical situations, they wrote.

One of the study’s authors reports grants from the National Institutes of Health. The other investigators report no relevant financial disclosures.

On Twitter @EAZTweets