Rethinking Palliative Interventions in Critical Care—When More Is Not Better

Kieran L. Quinn, MD, PhD^1,2,3; Rebecca Rodin, MD, MSc⁴; James DownarJAMA Intern Med. Published online December 16, 2024. doi:10.1001/jamainternmed.2024.6848

The intensive care unit (ICU) may be the hospital’s unrecognized epicenter for people with palliative care needs. Several existing clinical trials have tested palliative care interventions for critically ill patients and their families. These trials aimed to improve communication about goals of care, support complex decision-making, and enhance palliative care involvement.¹ Collectively, the results of these trials were variable and largely disappointing.

The randomized clinical trial by Cox et al² published in this issue of JAMA Internal Medicine is a welcome effort to advance the evidence in support of palliative care for critically ill older adults. The authors conducted a multicenter, parallel-group randomized clinical trial of 151 dyads of critically ill older adults with elevated palliative care needs and their family members. The complex intervention was multifaceted, consisting of a mobile application–based communication tool that displayed family-reported needs over 7 days, coached ICU attending physicians on addressing those needs, and prompted palliative care consultation if their needs did not improve within 3 study days. Two mandatory family meetings were also prescribed as part of the intervention. The primary outcome was the change in palliative care needs between study days 1 and 3. The study also measured several other patient-centered outcomes including goal concordance, quality of communication, and psychological distress. Ultimately, the palliative care needs of the intervention group were reduced over time. However, the improvement was no better than in the control group, despite a higher number of palliative care consultations and family meetings; these consultations and meetings might have taken more time and effort from families with no measurable clinical benefit.

The researchers in this clinical trial² should be commended for many reasons. First, they used a pragmatic, scalable digital health intervention; a resource-intensive intervention would have been unfeasible outside well-funded and well-staffed institutions. Second, they focused on participants with elevated needs. This strategy increased the likelihood of benefit and improved the efficiency of using limited specialized palliative care resources while excluding people with low baseline needs to avoid a floor effect. Third, they used person-centered outcomes and focused on both patients and their families.

Some important limitations may help to explain the lack of benefit in the intervention group in this clinical trial.²Randomization occurred at the patient level rather than the unit level, raising the possibility of contamination. Physicians who were reminded about the palliative needs of one patient could have been more likely to address the palliative needs of other patients, which would bias the results toward the null. Conversely, it is possible that the intervention did not change physician behavior at all, or that drawing attention to types of distress that are difficult to treat (eg, existential distress) would lead to little benefit and possible frustration from all parties, mitigating any other benefits. Furthermore, for the patients who ultimately received a palliative care consultation, it is not clear how the intervention enhanced collaboration beyond simply triggering the referral. In an increasingly busy and fragmented clinical environment, it is challenging to find ways to truly work together, rather than alongside one another. Studies evaluating the effects of triggered palliative care referral thus far have shown a modest effect at best, except to demonstrate improvement in the rate and timing of consultation.³

So why do so many palliative care interventions fail to improve outcomes in the ICU? Critical care and palliative care are sometimes thought to involve opposing philosophies—one focused on increasing quantity of life, the other on increasing quality of life. Yet the growth of the modern palliative care movement from the 1970s to today has been mirrored by a progressive shift in critical care practice toward less aggressive, comfort-focused health care in people with advanced, incurable illness. Critical care clinicians have made notable advances in the management of symptoms such as pain, dyspnea, agitation, and delirium, using guidelines and an evidence base that is far more developed than that of palliative care. Staffing ratios are also relatively high, so multidisciplinary teams have more time to spend with each patient. What then can palliative care clinicians add to the so-called “insensitive” care unit?

The acute setting of the ICU differs from the traditional model of palliative care in outpatients with cancer. In the outpatient setting, palliative care involvement often develops longitudinally over weeks to months. That time is essential to establish strong therapeutic relationships based on a foundation of trust, and to improve complex outcomes like quality of life. A recent meta-regression of 39 trials of specialist palliative care in adults with advanced illness found the greatest improvements in quality of life occurred after 3 to 9 months of treatment.⁴However, even when specialist palliative care was provided for that duration, only 1 in 14 people were found to experience a clinically important improvement. Put another way, a measurable benefit accrued only after 4 to 11 person-years of specialist palliative care. In the ICU context, where follow-up is much shorter (3 to 7 days in the present trial by Cox et al),² time is something that patients and care teams simply do not have.

The randomized clinical trial by Cox et al² and the existing evidence base do not prove that palliative care interventions in the ICU are ineffective. Instead, it emphasizes that methodical study and understanding of the full spectrum of needs of diverse populations of people with critical illness, coupled with the careful development and evaluation of all components of complex novel palliative care interventions are needed. Critically ill older adults with terminal illnesses may have a differing set of palliative care needs than critically ill younger adults with septic shock or trauma, or older adults receiving outpatient treatment for cancer. A heterogeneous array of interventions and outcomes founded on cancer-based models of palliative care (regardless of how they are tailored or targeted) may offer limited benefit above what contemporary ICU care can deliver to patients and families. Palliative care researchers should instead pivot to focusing on the development of novel and scalable models of care that collectively address the modifiable needs of critically ill patients in a timely way, which are poorly understood at present.

A multipronged effort is needed to meaningfully move the needle. This should include more comprehensive study and detailed measurement of individual physical, psychological, sociocultural, and other palliative care needs instead of using screening tools that merely identify these unmet domains of needs at a superficial level. Prospective measurement using novel tools, combined with in-depth qualitative work, will provide a more fulsome landscape of the heterogeneous needs of critically ill adults, including needs potentially modifiable through targeted interventions. The guided design and evaluation of innovative complex palliative care interventions in the ICU using the Medical Research Council framework⁵ will further enable a richer understanding of the contributions of each element of the intervention to inform iterative improvements and implementation. Ultimately, these efforts will accelerate the creation of new knowledge and novel interventions to yield larger returns.

The status quo of studying usual palliative care may no longer be sufficient. As the ICU continues to evolve, embracing an evolution of new palliative care interventions that truly address the complex needs of critically ill patients and their families will help ensure that the future of care is both compassionate and effective.