Skip to content

By: Laura Petrillo, @lpetrillz


Decisions about life-sustaining treatment in the intensive care unit are among the most ethically fraught decisions in medicine, and disagreements between clinicians and families are not uncommon. To provide guidance to clinicians, the American Thoracic Society (ATS) collaborated with the American Association of Critical Care Nurses and other critical care societies to create a policy statement on “Responding to Requests for Potentially Inappropriate Treatment in the Intensive Care Unit,” to be published on June 1st in the American Journal of Respiratory and Critical Care Medicine.

Proponents of palliative care will be glad to see that the guidelines recommend an emphasis on good communication in the ICU and early, expert consultation from palliative care or ethics services to prevent conflict. There has been increasing appreciation of the value of palliative care in intensive care, and efforts likeIPAL-ICU are making strides to facilitate the collaboration of ICU and palliative care teams. But there is still a long way to go, and the ATS’s endorsement of early palliative care in the ICU is most welcome.

When conflicts arise despite proactive communication, the guidelines recommend using specific language to categorize the family request that led to the dispute. The authors use the term “potentially inappropriate treatment” to acknowledge that in some scenarios, the treatment that a family requests may have some physiologic benefit, but the clinician disagrees with the worth of that benefit or the means justified to achieve it. This is a move toward transparency about the competing value judgments inherent in these decisions, in contrast to the past when clinicians had sole power to deem a treatment “futile” if it could not bring about an outcome that the clinician considered meaningful. The authors resurrect the idea of futility, but narrow the definition to apply only to treatments with absolutely no chance of bringing about a physiologic effect, like providing CPR to a corpse with rigor mortis.

The guidelines go on to propose a multi-step process for resolving intractable conflicts that ensures fairness and justice for patients and families. The steps again emphasize good communication and transparency. The authors suggest review by a multidisciplinary committee, and providing families with the opportunity for extramural appeal and transfer if the committee ultimately upholds the clinician’s stance on the request.

Finally, the guidelines take a bold step to urge clinicians to lead the way in advocating for policies and legislation about when life-sustaining treatment should be used. The authors suggest that clinicians partner with patients and other “key stakeholders” to create rules that guide how ICU beds should be allocated when scarce, for example, and whether to ever provide intensive care to patients with advanced dementia or in a persistent vegetative state.


At first blush, the ATS guidelines appear to strike a cautious balance of patient autonomy, exercised by families as surrogates, and clinician judgement of best interest. The press release for the article quotes one of the authors, Douglas White, MD, as saying, “Clinicians should neither simply acquiesce to treatment requests that they believe are not in a patient’s best interest, nor should they unilaterally refuse to provide treatment.”

But on closer inspection, the guidelines take a firm stance on which requests are open to debate. The guidelines assert that clinicians can refuse to provide treatment with no physiologic benefit, and can refuse requests for “legally proscribed” treatments, such as providing ongoing life support to a patient diagnosed as brain dead, in a state where brain death is recognized. This may seem obvious, but there has been an erosion of previously clear lines since the tragic and widely publicized case of Jahi McMath, the 13-year-old patient with anoxic brain injury who was kept on mechanical ventilation at her family’s request despite a diagnosis of brain death in California.

In addition, the guidelines suggest that when time is so limited as not to permit the recommended conflict-resolution process, clinicians can refuse to provide treatments that they feel certain are “outside of the bounds of accepted practice,” leaving room for interpretation. There is also a loophole in states, like Maryland, with statutes that give clinicians the right to refuse to provide “medically ineffective treatment,” with the example of a clinician refusing to provide CPR in a metastatic cancer patient with multi-organ system failure.

There is considerable variation in what clinicians believe is within the realm of “accepted practice” when it comes to the provision of resuscitation and life-sustaining treatment. The recommendation that we examine these situations systematically to arrive at public policy about when to offer these treatments is completely reasonable and just, but bound to fan political flames and fear of rationing, given that even legal endorsement of advance care planning met with hysterical fear of “death panels”.  Even the idea that clinicians can raise an objection to “potentially inappropriate” treatments on an ethical basis, and consider distributive justice in the context of a decision about a single patient, is itself somewhat radical and counter to current practice.

On a more basic level, I wonder whether the question posed by the authors, “how to respond to family requests for potentially inappropriate treatment” truly captures the issue in question in ICU conflicts.  How often do families bring their loved ones to the hospital and ask for vasopressors or mechanical ventilation by name? Much more often, clinicians provide life-sustaining treatment in an emergency by default, or out of uncertainty, and the conflict arises about when to discontinue it. Or resuscitation is offered even in situations with low likelihood of benefit, and then clinicians are disappointed when patients or families choose it. Can families be faulted for opting for more aggressive care when the healthcare system itself presents such ambivalence about the “appropriateness” of these interventions?  The idea of creating boundaries as a society would be ideal, but if the process will be hamstrung by politics, perhaps clinicians should agree on boundaries among themselves that can be applied fairly and consistently. Or the use of trials of treatment could be more uniformly applied, so that families expect that once treatments are initiated, there will be periodic review of whether those treatments are meeting the patient’s goals.

Another idea that merits mention is the presentation of an alternative, so that families do not feel that the options are life-sustaining treatment or nothing. In a different guideline, published last year in the same journal, the ATS presented its “Choosing Wisely” list for critical care medicine. In it, the ATS again endorsed the idea of palliative care in the ICU with its recommendation, “Do not continue life support for patients at high risk for death or severely impaired functional recovery without offering patients and their families the alternative of care focused entirely on comfort,” and the additional comment, “Even among patients pursuing life-sustaining therapy, initiating palliative care simultaneously with ongoing disease-focused therapy may be beneficial.”

There is still much work to be done to establish the boundaries of “appropriate treatment” for critically ill patients, but the ATS guidelines present a great starting point for discussion and attempt to find routes to transparency and fairness in an ethically murky area.

Back To Top