Compassionate deactivation of ventricular assist devices in children: A survey of pediatric ventricular assist device clinicians' perspectives and practices

Deactivation of Left Ventricular Assist Devices at the End of Life: Narrative Review and Ethical Framework

Left ventricular assist devices (LVADs) have become an increasingly common advanced therapy in patients with severe symptomatic heart failure. Their unique nature in prolonging life through incorporation into the circulatory system raises ethical questions regarding patient identity and values, device ontology, and treatment categorization; approaching requests for LVAD deactivation requires consideration of these factors, among others. To that end, clinicians would benefit from a deeper understanding of: 1) the history and nature of LVADs; 2) the wider context of device deactivation and associated ethical considerations; and 3) an introductory framework incorporating best practices in requests for LVAD deactivation (specifically in controversial situations without obvious medical or device-related complications). In such decisions, heart failure teams can safeguard patient preferences without compromising ethical practice through more explicit advance care planning before LVAD implantation, early integration of hospice and palliative medicine specialists (maintained throughout the disease process), and further research interrogating behaviors and attitudes related to LVAD deactivation.

Home Milrinone in Pediatric Hospice Care of Children with Heart Failure

CONTEXT

The symptom profile of children dying from cardiac disease, especially heart failure, differs from those with cancer and other non-cardiac conditions. Treatment with vasoactive infusions at home may be a superior therapy for symptom control for these patients, rather than traditional pain and anxiety management with morphine and benzodiazepines.

OBJECTIVES

We report our experience using outpatient milrinone in children receiving hospice care for end-stage heart failure.

METHODS

Retrospective review of a contemporary cohort of all patients at Lucile Packard Children's Hospital, Stanford who were discharged on intravenous milrinone and hospice care between 2008 and 2021. Clinical data, including cardiac diagnosis, milrinone dose and route of administration, total milrinone days, symptoms reported, rehospitalization rates, concurrent therapies and complications were analyzed.

RESULTS

Among 8 patients, median duration of home milrinone infusion was 191 (33, 572) days with the longest support duration 1,054 days. All (100%) patients were also receiving diuretics at the time of death. Five (63%) were receiving no other pain control medications until the active phase of dying. From milrinone initiation to last outpatient assessment, a reduction in the number of patients reporting respiratory discomfort, abdominal pain, weight loss/lack of appetite, and fatigue was observed. Six (75%) died at home.

CONCLUSION

We used milrinone with oral diuretics effectively for symptom control in children with heart failure on palliative care. Our experience was that this combination can be used safely in the outpatient setting for long-term use without the addition of opiates, benzodiazepines, or supplemental oxygen in most cases.

End-of-Life in Pediatric Patients Supported by Ventricular Assist Devices: A Network Database Cohort Study

OBJECTIVES

Most pediatric patients on ventricular assist device (VAD) survive to transplantation. Approximately 15% will die on VAD support, and the circumstances at the end-of-life are not well understood. We, therefore, sought to characterize patient location and invasive interventions used at the time of death.

DESIGN

Retrospective database study of a cohort meeting inclusion criteria.

SETTING

Thirty-six centers participating in the Advanced Cardiac Therapies Improving Outcomes Network (ACTION) Registry.

PATIENTS

Children who died on VAD therapy in the period March 2012 to September 2021.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Of the 117 of 721 patients (16%) who died on VAD, the median (interquartile range) age was 5 years (1-16 yr) at 43 days (17-91 d) postimplant. Initial goals of therapy were bridge to consideration for candidacy for transplantation in 60 of 117 (51%), bridge to transplantation in 44 of 117 (38%), bridge to recovery 11 of 117 (9%), or destination therapy (i.e., VAD as the endpoint) in two of 117 (2%). The most common cause of death was multiple organ failure in 35 of 117 (30%), followed by infection in 12 of 117 (10%). Eighty-five of 92 (92%) died with a functioning device in place. Most patients were receiving invasive interventions (mechanical ventilation, vasoactive infusions, etc.) at the end of life. Twelve patients (10%) died at home.

CONCLUSIONS

One-in-six pediatric VAD patients die while receiving device support, with death occurring soon after implant and usually from noncardiac causes. Aggressive interventions are common at the end-of-life. The ACTION Registry data should inform future practices to promote informed patient/family and clinician decision-making to hopefully reduce suffering at the end-of-life.

State of the science and future research directions in palliative and end-of-life care in paediatric cardiology: a report from the Harvard Radcliffe Accelerator Workshop

Workshop proceedings, priorities, and recommendations from the "State of the Science and Future Directions in Palliative and End-of-Life Care in Pediatric Cardiology," a Harvard Radcliffe Accelerator Workshop, are detailed. Eight priorities for research were identified, including patient and family decision making, communication, patient and family experience, patient symptom measurement and management, training and curriculum development, teamwork, family hardships and bereavement, and ethical considerations. Barriers to research in this area were also identified: lack of outcome/measurement tools, lack of research funding, small population sizes, lack of effort/protected time for research, undervalued research topic by field and colleagues, and heterogeneous research participant diversity. Priorities and barriers were mostly consistent with those reported by the field of paediatric palliative care at large. These collective, consensus-based findings from diverse, multidisciplinary leaders in the field, as well as parent representatives, provide a catalyst for scientific advancement specific to paediatric and end-of-life care in paediatric cardiology.

Compassionate Deactivation of Pediatric Ventricular Assist Devices: A Review of 14 Cases

CONTEXT

Compassionate deactivation (CD) of ventricular assist device (VAD) support is a recognized option for children when the burden of therapy outweighs the benefits.

OBJECTIVES

To describe the prevalence, indications, and outcomes of CD of children supported by VADs at the end of life.

METHODS

Review of cases of CD at our institution between 2011 and 2020. To distinguish CD from other situations where VAD support is discontinued, patients were excluded from the study if they died during resuscitation (including extracorporeal membrane oxygenation), experienced brain or circulatory death prior to deactivation, or experienced a non-survivable brain injury likely to result in imminent death regardless of VAD status.

RESULTS

Of 24 deaths on VAD, 14 (58%) were CD. Median age was 5.7 (interquartile range (IQR) 0.6, 11.6) years; 6 (43%) had congenital heart disease; 4 (29%) were on a device that can be used outside of the hospital. CD occurred after 40 (IQR: 26, 75) days of support; none while active transplant candidates. CD discussions were initiated by the caregiver in 6 (43%) cases, with the remainder initiated by a medical provider. Reasons for CD were multifactorial, including end-organ injury, infection, and stroke. CD occurred with endotracheal extubation and/or discontinuation of inotropes in 12 (86%) cases, and death occurred within 10 (IQR: 4, 23) minutes of CD.

CONCLUSION

CD is the mode of death in more than half of our VAD non-survivors and is pursued for reasons primarily related to noncardiac events. Caregivers and providers both initiate CD discussions. Ventilatory and inotropic support is often withdrawn at time of CD with ensuing death.

Pediatric Palliative Care in the Heart Failure, Ventricular Assist Device and Transplant Populations: Supporting Patients, Families and Their Clinical Teams

Heart failure is a life-changing diagnosis for a child and their family. Pediatric patients with heart failure experience significant morbidity and frequent hospitalizations, and many require advanced therapies such as mechanical circulatory support and/or heart transplantation. Pediatric palliative care is an integral resource for the care of patients with heart failure along its continuum. This includes support during the grief of a new diagnosis in a child critically ill with decompensated heart failure, discussion of goals of care and the complexities of mechanical circulatory support, the pensive wait for heart transplantation, and symptom management and psychosocial support throughout the journey. In this article, we discuss the scope of pediatric palliative care in the realm of pediatric heart failure, ventricular assist device (VAD) support, and heart transplantation. We review the limited, albeit growing, literature in this field, with an added focus on difficult conversation and decision support surrounding re-transplantation, HF in young adults with congenital heart disease, the possibility of destination therapy VAD, and the grimmest decision of VAD de-activation.

Variability in clinical decision-making for ventricular assist device implantation in pediatrics

BACKGROUND

Minimal data exist on clinical decision-making in VAD implantation in pediatrics. This study aims to identify areas of consensus/variability among pediatric VAD physicians in determining eligibility and factors that guide decision-making.

METHODS

An 88-item survey with clinical vignettes was sent to 132 pediatric HT cardiologists and surgeons at 37 centers. Summary statistics are presented for the variables assessed.

RESULTS

Total respondents were 65 (72% cardiologists, 28% surgeons) whose centers implanted 1-5 (34%), 6-10 (40%), or >10 (26%) VADs in the past year. Consensus varied by patients' age, diagnosis, and Pedimacs profile. Highest agreement to offer VAD (97%) was a mechanically ventilated teenager with dilated cardiomyopathy. Patients stable on inotropes were less likely offered VAD (11%-25%). SV infant with Pedimacs profile 2 had the most varied responses: 37% offered VAD; estimated survival ranged from 15% to 90%. Variables considered for VAD eligibility included mild developmental delays (100% offered VAD), moderate-severe behavioral concerns (46%), cancer in remission >2 years (100%), active malignancy with good prognosis (68%) or uncertain prognosis (36%), and BMI >35 (74%) or <15 (69%). Most respondents (91%) would consider destination therapy VADs in pediatrics, though not currently feasible at 1/3 of centers. Factors with greatest influence on decision-making included HT candidacy, families' goals of care, and risks of complications.

CONCLUSIONS

Significant variation exists among pediatric VAD physicians when determining VAD eligibility and estimating survival, which can lead to differences in access to emerging technologies across institutions. Further work is needed to understand and mitigate these differences.

Palliative Care Engagement for Pediatric Ventricular Assist Device Patients: A Single-Center Experience

Outcomes in pediatric patients with ventricular assist devices (VADs) for advanced heart failure (HF) are improving, but the risk of associated morbidity and mortality remains substantial. Few data exist on the involvement of pediatric palliative care (PPC) in this high-risk patient population. We aimed to characterize the extent of palliative care involvement in the care of patients requiring VAD placement at our institution. Single-center retrospective chart review analyzing all VAD patients at a large pediatric center over a 4 year period. Timing and extent of palliative care subspecialty involvement were analyzed. Between January 2014 and December 2017, 55 HF patients underwent VAD implantation at our institution. Pediatric palliative care utilization steadily increased over consecutive years (2014: <10% of patients, 2015: 20% of patients, 2016: 50% of patients, and 2017: 65% of patients) and occurred in 42% (n = 23) of all patients. Of these, 57% (n = 13) occurred before VAD placement while 43% (n = 10) occurred after implantation. Patients who died during their VAD implant hospitalization (24%, n = 13) were nearly twice as likely to have PPC involvement (62%) as those who reached transplant (38%). Of those who died, patients who had PPC involved in their care were more likely to limit resuscitation efforts before their death. Four patients had advanced directives in place before VAD implant, of which three had PPC consultation before device placement. Three families (5%) refused PPC involvement when offered. Pediatric palliative care utilization is increasing in VAD patients at our institution. Early PPC involvement occurred in the majority of patients and appears to lead to more frequent discussion of goals-of-care and advanced directives.