COMPLETE (Communication Plan Early Through End of Life): Development of a Research Program to Diminish Suffering for Children at End of Life

Abstract

While overall survival has improved significantly for children with cancer over the past 75 years, cancer remains the leading cause of death from disease among children and adolescents. Further, despite the many advances in medical and nursing care, children with cancer still experience significant physical and emotional suffering over the course of their illness, especially at the end of life (EOL). Children endure significant rates of high-intensity medical interventions (e.g., intubation, intensive care unit admission) at the EOL despite many parents, adolescents, and young adult patients identifying home as their preferred location of death. Hospice care has the potential to ease suffering at the EOL and facilitate home deaths, and yet, most children still die in acute care settings without hospice care. Numerous barriers prevent timely enrollment in hospice among children with cancer who are in the EOL period. This report describes the development and testing of a palliative care/EOL communication intervention designed to overcome some of these barriers and subsequently improve EOL outcomes (i.e., earlier hospice enrollment, less use of high-intensity medical interventions, reduced pain and suffering) among children with cancer and their parents (i.e., less emotional distress and uncertainty, improved hope and healthcare satisfaction).

Although overall survival rates for children with cancer have improved dramatically over the past 75 years, cancer remains the number one cause of death by disease among children aged 1 to 19 years. Cancer is responsible for more deaths during childhood than all other diseases combined (Kochanek et al., 2019). Further, despite the many advances in medical and nursing care, children with cancer still experience significant physical and emotional suffering over the course of their illness, especially at the end of life (EOL). Children endure significant rates of high-intensity medical interventions (e.g., intubation, intensive care unit admission) at the EOL despite many parents, adolescents, and young adult patients identifying home as their preferred location of death (Johnston et al., 2019). Hospice care has the potential to ease suffering at the EOL and facilitate home deaths, and yet, most children still die in acute care settings without hospice care (Johnston et al., 2020).

Several patient, family, clinician, and system-level barriers prevent timely hospice enrollment for children with a life-threatening illness. Earlier hospice enrollment requires that clinicians have frank and ongoing conversations with parents about any changes to their child’s illness, response to treatments, and prognosis. Clinicians often lack palliative care (PC)/EOL communication training and may be hesitant or uncomfortable discussing the child’s prognosis and the benefits of palliative and/or hospice care, fearing that parents will think they are “giving up” or abandoning their child despite ample evidence to the contrary (Kaye et al., 2016; Levine et al., 2017). Further, parental misconceptions about palliative and hospice care may limit conversations with clinicians (Haines et al., 2018). Thus, more research is needed to help break down these barriers and support clinicians in engaging parents in earlier conversations that focus on enhancing prognostic understanding and explaining how prognosis may evolve over time, ensuring that the treatment plan aligns with patient and family goals of care. Specifically, interventions are required to facilitate parental receptivity to altering the treatment plan and goals of care when cure becomes unlikely (e.g., accepting a focus on comfort and support prior to end of life), which may open the door to earlier hospice enrollment and thereby improved patient and parental outcomes.

The ORBIT Model for Behavioral Treatment Development (Czajkowski et al., 2015) provided the framework for creating an intervention to fill this identified gap. The ORBIT Model outlines flexible and iterative research processes that focus on pre-efficacy testing to identify essential elements, doses, and targets of behavioral interventions. Following this model, our first qualitative study funded by the Oncology Nursing Foundation (Research Fellowship Grant) explored bereaved parents’ memories of their conversations with clinicians about hospice care and symptoms of concern during their child’s last week of life (Hendricks-Ferguson, 2007). Parents of children who died from brain tumors reported receiving hospice information near the child’s death (96%) and that communication regarding EOL care was handled insensitively (50%). Parents wished they had, had earlier EOL discussions (~50%). These findings provided evidence of the need to initiate EOL conversations earlier in the disease trajectory and to ensure that clinicians are trained to empathically discuss EOL care and the importance of evaluating symptom distress during the EOL period.

We subsequently developed the COMPLETE I (Communication Plan: Early Through End of Life) intervention to address these parent-identified needs. The COMPLETE I intervention involved a series of three medical doctor/registered nurse (MD/RN)-guided goals-of-care conversations that began at diagnosis and continued over approximately 6 months. The MD/RN dyads used conversation guides and visual aids to help parents identify their own and their child’s values and make realistic goals-of-care decisions (cure, slow the cancer, or comfort) in the context of the child’s shifting prognosis. To deliver the COMPLETE I intervention, MD/RN dyads were trained on the COMPLETE I study protocol, implementation procedures, and palliative care (PC)/EOL evidence-based best practices. Dyads also participated in role-play activities with bereaved parents portraying parents of children with high-risk cancers. MD/RN dyads reported that our PC/EOL training activities were sufficient for them to deliver COMPLETE I and that it enhanced their EOL communication skills. However, they would have preferred role-playing with trained actors versus bereaved parents (Hendricks-Ferguson et al., 2015).

The COMPLETE I intervention was then evaluated with 13 parents of 11 children with high-grade malignant brain tumors to determine acceptability, feasibility, and preliminary efficacy (funded by the National Institutes of Health [NIH] and National Institute of Nursing Research [1R21NR011071-01A1]). The COMPLETE I intervention was feasible and acceptable to parents of children with poor-prognosis brain tumors. The MD/RN delivered sessions were incorporated into regularly scheduled clinic appointments, which resulted in 100% attendance at scheduled sessions. COMPLETE I was also associated with positive parental changes, noting significantly increased hope, significantly decreased uncertainty, a trend towards decreased emotional distress, and significantly diminished pain and procedural anxiety based on parents’ perceptions of their child’s responses to treatments and their condition (Hendricks-Ferguson et al., 2017). In post-intervention qualitative interviews, the parents also reported many positive benefits from participating in the COMPLETE I intervention. Positive benefits included having more realistic hopes and goals for their child, engaging in meaningful dialogue with their child’s oncology providers, and exploring future EOL options prior to a health crisis (Hendricks-Ferguson & Haase, 2019).

Based on pilot study results and in consultation with communication skills training experts, the COMPLETE I intervention training was enhanced to promote improved collaboration and delivery of the intervention by MD/RN dyads. Training refinements included: 1) three versus two days of training; 2) trained actors versus bereaved parents to role-play as parents; 3) booster communication skills training to support MD/RN dyad use of empathic statements and assessment of family goals and values; 4) interprofessional collaboration training; and 5) training off-site to ensure focused time for interprofessional team-building and learning. The revised training resulted in improved use of empathic statements compared with COMPLETE I (100 % versus 11%; p <.001) and improved clinician self-assessed preparedness for these discussions compared to baseline (Moody et al., 2020).

The revised intervention, COMPLETE II, was delivered by three newly trained MD/RN dyads to a sample of six parents of children with poor-prognosis cancers. Among children in COMPLETE I and II, approximately 85% were enrolled in hospice (median 39.5 days) versus the national average of 40%, and only 15% received high-intensity medical interventions at EOL versus the national average of 60% (Moody et al., 2020). Although non-significant due to the small sample size, both parent and child clinical outcomes trended in the expected directions with increased parental hope, decreased parental uncertainty and decreased pain in the children. Healthcare satisfaction remained high throughout.

The pilot work examining COMPLETE I and COMPLETE II provided compelling evidence that led to NIH/National Cancer Institute (NCI) funding (MPIs; Ferguson, Haase, and Moody) to conduct a cluster-randomized controlled trial (RCT) to evaluate the COMPLETE II intervention among parents of children with poor-prognosis cancers (R01CA235632-01A1) at seven pediatric oncology treatment centers. The aim of this current RCT is to evaluate the effectiveness of the intervention to improve child EOL outcomes (i.e., earlier hospice enrollment, less use of high-intensity medical interventions, reduced pain and suffering) and parental outcomes (i.e., less emotional distress and uncertainty, improved hope and healthcare satisfaction). The extent to which the effects of the intervention on child and parent outcomes are mediated by therapeutic alliance and parent/physician goals-of-care alignment will also be examined. Prior to activation of this RCT, the COMPLETE II intervention was retitled as the Informational Meetings for Planning and Coordinating Treatments (IMPACT) intervention to help foster parent receptivity to receive information about the study as parents are approached shortly after their child’s diagnosis.

For this RCT, MD/RN dyads were recruited from the solid tumor and central nervous system (CNS) tumor programs at the participating institutions. MD/RN dyads were randomized to deliver the IMPACT intervention or an Attention Control education intervention. The constrained randomization plan considered previous MD and RN PC/EOL communication skills training, number of patients seen annually per MD/RN team, and each team’s disease subspecialty (CNS versus non-CNS solid tumor). The Attention Control education intervention is a nurse-led intervention, which allows parents additional time to engage in new cancer diagnosis teaching guided by the NCI’s handbook, The Child with Cancer: A Parent’s Guide. Parents of children with poor-prognosis (anticipated overall survival less than approximately 25% at 5 years) CNS or non-CNS solid tumors are eligible to participate. Enrollment occurs within eight weeks of a new high-risk cancer diagnosis. Parents and their child, if interested and at parent discretion, participate in three one-hour study sessions with the trained interveners over the course of approximately six to nine months, correlating with disease evaluation time points.

This RCT examining the IMPACT intervention is the first NIH/NCI-funded study aimed to foster earlier hospice enrollment and reduce the suffering of children with poor-prognosis cancers at EOL by improving goals-of-care alignment early in the cancer course between the child’s clinicians and their parents. We hypothesize that parents receiving the intervention will be more receptive to having their child receive hospice care resulting in longer hospice enrollment than what is described in the published literature (i.e., approximately 2 months before a child’s death) (Dingfield et al., 2015). We also hypothesize that parents receiving the intervention will have significant improvement in hope and decreased uncertainty and that children whose parents received the intervention will experience fewer high-intensity medical interventions and significantly less pain and emotional suffering at EOL. In summary, this carefully designed RCT examining the efficacy of the IMPACT intervention supports the tenets of the NIH, NINR, NCI, and National Academy of Medicine to 1) foster earlier PC/EOL care for seriously ill children, 2) ensure that parents’ goals of care are congruent with the treatments their child receives, and 3) promote high-quality care at EOL.

Highlights.

• Despite significant advances in medical and nursing care, children with cancer and their parents still suffer physical and emotional distress during the end-of-life period.

• The COMPLETE I (Communication Plan: Early Through End of Life) intervention involves a series of three medical doctor/registered nurse (MD/RN)-guided goals-of-care conversations that begin at diagnosis and help parents identify their own and their child’s values and make realistic goals-of-care decisions (cure, slow the cancer, or comfort) in the context of the child’s shifting prognosis.

• Pilot testing of the intervention of the intervention demonstrated improved end-of-life outcomes with increased rates of hospice enrollment and decreased used of high-intensity medical interventions at end of life. Also, both parent and child clinical outcomes trended in the expected directions with increased parental hope, decreased parental uncertainty and decreased pain in the children at end of life.

• The COMPLETE intervention is currently undergoing efficacy testing in a multi-site randomized controlled trial, sponsored by the NIH/NCI.