By Elaine Chen, MD

Assistant Professor, Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, Section of Palliative Medicine, Rush University Medical Center, Chicago

Dr. Chen reports no financial relationships relevant to this field of study.

SYNOPSIS: When considering prognostication or limitation of care in critically ill patients with different cancers, oncologists may focus on cancer characteristics, whereas intensivists may focus on multiple organ failure.

SOURCE: Nassar AP Jr, et al. Oncologists’ and intensivists’ attitudes toward the care of critically ill patients with cancer. J Intensive Care Med 2017 Jan 1:885066617716105. doi: 10.1177/0885066617716105. [Epub ahead of print].

Approximately 20% of all ICU admissions present with cancer, and ICU mortality for patients with cancer is approximately 30%. Intensivists might be overly pessimistic about their patients with cancer, and oncologists might be overly optimistic about their patients’ survival. Metastatic cancer often is associated with refusal of ICU admission in some countries. The aim of this study was to identify whether oncologists and intensivists would make different decisions in clinical scenarios for two critically ill patients with cancer.

The authors created a survey that included demographic information and two hypothetical patient vignettes, which the authors administered to oncologists and intensivists at two academic cancer centers in Brazil. The case vignettes were identical, other than the type of cancer: metastatic pancreatic or triple-positive metastatic breast cancer. The patient developed septic shock after two days of inpatient therapy for community acquired pneumonia. Participants were asked about their decision-making at this juncture and again three days later when the patient was deteriorating despite mechanical ventilation, vasopressors, and renal replacement therapy.

At the first juncture, options included no ICU admission, ICU admission with discussion of eventual limitation of life support, and ICU admission without discussion of eventual limitation of life support. At the second juncture, options included comfort measures with withdrawal of life support, withholding (which investigators defined as maintenance of life support without additional measures), or full code. The authors hypothesized that oncologists would focus on cancer status and intensivists would focus on organ dysfunction when making decisions, and that oncologists would manage the patients differently, whereas intensivists would favor limitation of care for both patients.

Of 113 physicians invited, 106 completed the survey. Sixty were intensivists, and 46 were oncologists. More oncologists had attended end-of-life or palliative care courses, but more intensivists had participated in end-of-life or palliative care educational activities within the past year.

For the pancreatic cancer patient, more intensivists (75%) than oncologists (52%) opted for admission to the ICU with discussion of eventual limitation of life support (P = 0.02). Most of the oncologists who chose otherwise opted for no ICU admission. At three days, most respondents opted for withdrawal of life support (76%), with no difference between oncologists and intensivists.

For the breast cancer patient, no participant selected no ICU admission, and most opted for ICU admission with discussion of eventual limitation of life support, with more intensivists (78%) than oncologists (59%) selecting this option (P = 0.055). At three days, more intensivists (54%) than oncologists (21%) favored withdrawal of life support (P < 0.001).

For both patient types, those who had attended palliative care education at any time opted for withdrawal more frequently than those who had not (90% vs. 53% for the pancreatic patient, P = 0.018; 46% vs. 12% for the breast cancer patient, P = 0.007). More oncologists favored full code for the breast cancer patient compared to the pancreatic cancer patient (27% vs. 0%, P < 0.001), whereas there were no differences among the intensivists (13% vs. 3%; P = 0.094).

These results suggest that oncologists and intensivists use different clinical information in their decision-making processes. The authors hypothesized that oncologists may pay more attention to cancer survival rates, and intensivists pay more attention to organ dysfunction survival rates. In the pancreatic cancer patient, this led to the same conclusion of a dismal prognosis, but different conclusions in the breast cancer patient. While this simple study does not further evaluate decision-making processes, it does point out potential cognitive biases and areas of conflict between oncologists and intensivists. The study authors also noted that receiving updates on palliative care and end-of-life issues is associated with a higher tendency to withdraw life support.


This study provides important perspective to help intensivists be mindful in their conversations with referring colleagues. An intensivist’s perspective is appropriately narrowly focused on the patient’s current critical status. However, referring colleagues may have a long-term outpatient relationship with strong emotional connections and the perspective of the patient’s prior excellent functional status.1 Personally speaking, as a pulmonologist, my outlook for my chronic stable outpatients who are admitted to the ICU often is more positive than it would be for similar patients I first meet when they are critically ill.

Conflicts are common in critically ill patients requiring life-sustaining therapy. Up to 50% of these conflicts arise among hospital staff,2 and up to 20% of conflicts involve ICU staff and consultants.3 These conflicts may arise between oncologists and intensivists regarding the aggressiveness of management of critically ill patients with cancer. Conflict may not necessarily lead to negative outcomes, but can present an opportunity for constructive conversation.

This study notched an exceptionally high response rate of 94%. The survey was short and simple (four questions). The simplicity of the survey likely contributed to the high response rate but limited the extent of potential analysis. Other limitations include the fact that survey responses may not actually reflect clinical practice, and that only cancer centers in Brazil were included. Do these results generalize to community hospitals without cancer centers elsewhere in the world? Are there practice differences in Brazil compared to the United States? For example, refusing admission to the ICU from an acute care unit is a rare practice in the United States, regardless of presenting illness. Many oncologists in this study recommended deferring ICU admission for the pancreatic cancer patient, while intensivists recommended admission; this result warrants further investigation.

As cancer treatment improves, worldwide ICU use by cancer patients has increased, and outcomes have improved.4 Given the increasing use of ICUs by cancer patients, intensivists and oncologists should respectfully remember that perspectives are different and communicate regularly about their perspectives in hopes of reducing conflict.


  1. Bhatnagar M, Arnold R. The oncology-ICU palliative care interface #310. J Palliat Med 2016 Jul;19:785-786.
  2. Breen CM, et al. Conflict associated with decisions to limit life-sustaining treatment in intensive care units. J Gen Intern Med 2001;16: 283-289.
  3. Azoulay E, et al. Prevalence and factors of intensive care unit conflicts: The conflicus study. Am J Respir Crit Care Med 2009;180:853-860.
  4. Chen E. ICU outcomes and triage in elderly patients with advanced cancer. Critical Care Alert 2016;24:33-37.