Discussion

We described the rate and timing of palliative care consultation in children receiving extracorporeal membrane oxygenation support and compared demographic information, clinical characteristics, and outcomes of children who did and did not receive a consult. Our study has three main findings. First, palliative care consultation occurred infrequently. Second, the consultation often occurred after a prolonged length of stay in the setting of high disease severity. Third, mortality rates were high regardless of consultation status.

Less than one in five children receiving extracorporeal membrane oxygenation had a palliative care consult despite recent guidelines that all children supported on extracorporeal membrane oxygenation are eligible for palliative care consultation. ^{Reference DellaVolpe, Barbaro and Cannon7,Reference Friebert and Osenga8} The low rates of consultation may be related to the heterogenicity of disease within the extracorporeal membrane oxygenation population. Despite the high risk of complications inherent to the highly technical therapy, prognostication of outcomes beyond survival is very difficult, even for experienced providers. ^{Reference Costello, Mussatto and Cassedy27} In addition, critical care physicians have expressed concerns that palliative care creates conflicts among patient’s care teams, adds more providers to an already long list of specialists, and gives families the impression of giving up or losing hope. ^{Reference Morell, Thompson and Rajagopal28–Reference Balkin, Kirkpatrick and Kaufman30} A recent survey regarding the benefits and barriers of an automatic palliative care consultation for children supported on extracorporeal membrane oxygenation showed similar concerns by ICU physicians. ^{Reference Delgado-Corcoran, Wawrzynski and Mansfield31}

Unfortunately, the majority of children who need extracorporeal membrane oxygenation support are placed emergently on the device, not allowing enough time for a clear and thoughtful conversation between the clinical team and the patient’s family to fully absorb the goals and clinical risks, and benefits of this emergent therapy. ^{Reference Mulaikal, Nakagawa and Prager32,Reference Makdisi and Makdisi33} Not surprisingly, parents of extracorporeal membrane oxygenation survivors and non-survivors have indicated the need for improved communication with the critical care team, particularly regarding extracorporeal oxygenation weaning and decannulation. ^{Reference Curley and Meyer34} Even though critical care physicians possess expertise in communication and primary palliative care principles, their proficiency in these areas varies. There are guidelines in place that are intended to assist critical care physicians in addressing serious news, prognostic awareness, and interventions with families of children supported on extracorporeal membrane oxygenation. ^{Reference Moynihan, Purol and Alexander35} Nonetheless, the specialised palliative care team has the expertise and training to facilitate these difficult conversations with families to elucidate goals of care, facilitate decision-making, and advance care planning. ^{Reference Richards, Starks and O’ Connor14,Reference Ciriello, Dizon and October15} This is especially important considering that there are multiple specialists involved in the care of these patients and critical care physicians may face time-limiting constraints with the care of other high-intensity critical care patients. ^{Reference Richards, Starks and O’ Connor14,Reference Ciriello, Dizon and October15} The discussions facilitated by the palliative care team improve parental satisfaction with care, reduce families’ anxiety, decrease decision regret, and reduce the use of non-beneficial treatments. ^{Reference Kaye, Weaver and DeWitt9–Reference Vern-Gross, Lam and Graff13,Reference Hancock, Pituch and Uzark36,Reference Marcus, Santos and Ciapponi37} The use of an automatic referral model has the potential to standardise the consultation process, speed the time to referral, and decrease the variability of referral practices in children supported on extracorporeal membrane oxygenation. ^{Reference Zalenski, Jones and Courage17,Reference Lutmer, Humphrey and Kempton21–Reference Doorebos, Starks and Bourget23,Reference Nelson, Curtis and Mulkerin38,Reference Boss, Nelson and Weissman39}

Our second finding demonstrated that children were in the hospital 3 days before cannulation or in the device for almost a week and yet, the consults occurred nearly 1 month after extracorporeal membrane oxygenation cannulation in the setting of high disease severity, missing an opportunity for palliative care involvement earlier in the disease trajectory.

These findings suggest that palliative care involvement was not viewed as an acute need and was only considered when the clinical course became protracted. Families of children supported on extracorporeal membrane oxygenation report acute stress related to concerns regarding the chances of survival, pain and suffering, and long-term quality of life of their child. ^{Reference Curley and Meyer34,Reference Tramm, Ilic and Murphy40–Reference Garcia Guerra, Robertson and Alton43} Recent publications show that children supported on extracorporeal membrane oxygenation had significantly lower health-related quality of life scores when compared to healthy cohorts. ^{Reference Feudtner, Nye and Boyden44–Reference Sealey, Breen and O'Connor46} Furthermore, children are likely to have several complex chronic conditions after decannulation that put their parents and their siblings at risk of poor mental and physical outcomes. ^{Reference Feudtner, Nye and Boyden44} These findings support the involvement of the palliative care team in providing longitudinal psychosocial-spiritual support to children and their families to improve overall quality of life, increase family satisfaction with care, and reduce health care utilisation. ^{Reference DellaVolpe, Barbaro and Cannon7,Reference Grunauer, Mikesell and Bustamante47}

Lastly, involving palliative care early allows time for the palliative care team to build rapport, and assess the needs and goals of the family while providing psychosocial support to the family and health care team through the disease’s trajectory as well as at the end of life care especially considering that the mortality rates among children supported on extracorporeal membrane oxygenation are high. ^{Reference Machado, Hollander, Murray and Philip48–Reference Goloff and Joy50} None of the children who died within 48 hours of extracorporeal membrane oxygenation cannulation received a palliative care consultation, yet these children were in the hospital for at least 3 days prior to cannulation, providing a window of opportunity for early consultation. Providers may not recognise this window as an appropriate time to consult palliative care; however, earlier consultation in children with life-threatening conditions allows time for the palliative care team to build relationships and facilitate important conversations with families about their values and goals of care. ^{Reference Bogetz, Ullrich and Berry49,Reference Goloff and Joy50} In our study, mortality rates were similar to ones reported by the Extracorporeal Life Support Organization ^{Reference Di Nardo, MacLaren and Marano1,Reference Laussen and Guerguerian2,Reference Maratta, Potera and van Leeuwen24,Reference Brown, Moynihan and Deatrick25} and did not differ by palliative care involvement. However, it is clear that there are missed opportunities for palliative care to support bereaved families. Serial assessments of the family after the child’s death provide the palliative care team the opportunity to screen and refer those families at risk for complicated grief and other poor bereavement outcomes. ^{Reference Garcia Guerra, Robertson and Alton43} Additionally, children who have undergone extracorporeal membrane oxygenation treatment are at risk for late mortality and more complex chronic conditions. ^{Reference Coleman, Goldman and Moffett3} Palliative care can assist these families navigating complex chronic conditions by providing symptom management, counselling, and facilitating decision making across changes in the child’s illness and in different settings.

Our study has limitations. As this study includes data from a single centre, the generalizability of our findings is limited. We collected data from an administrative database and are limited by the variables available in the data set. For example, we do not have data to determine when extracorporeal support was considered in relation to the device deployment. Likewise, we do not have patient-centered outcomes and the psychological impact of interdisciplinary palliative care on providing longitudinal psychosocial-spiritual support to families of extracorporeal membrane oxygenation survivors and non-survivors. In addition, there may be under or over-reporting of complex chronic conditions related to possible misclassifications in the International Classification of Disease version 10 Clinical Modification coding.