Pediatric Palliative Care Programs in US Hospitals

The National Palliative Care Registry (“the Registry”) has been collecting self-reported, annualized data on palliative care program structures and processes since 2008.²⁵  Participation in the Registry is free and voluntary for all specialty palliative care programs in the United States. Programs are recruited through word of mouth, e-mail marketing, presentations, publications, and conference presence.

The Registry platform allows numerous palliative care program members to work collaboratively on a single program dashboard while preventing the submission of duplicate surveys. The fourth edition of the National Consensus Project Clinical Practice Guidelines for Quality Palliative Care represent best practices in the field; elements in the Registry align with domain 1: structure and processes of care.^13,25  The Registry elements are organized into 3 categories: core program information, optional program information, and specialty modules (see Supplemental Information).

Program leadership maintains responsibility for the integrity and validity of entered data. Registry administrators monitor data entries, identify outliers and significant changes in program reporting over time, compare submitted hospital data to the data in the American Hospital Association Annual Database for consistency, and follow-up with programs on missing or erroneous data. In addition to this support, participating programs receive annual internal and peer comparison reports that can be accessed at any time.²⁵ 

In 2019, a panel of PPC experts led the development of pediatric-specific data elements for the Registry’s hospital survey, including a PPC specialty module and the opportunity for programs to self-identify as a PPC program. Although there are no required questions in the Registry, programs that identify as a PPC program are encouraged to submit all core program information and the PPC specialty module.

Because of the Registry’s annual, retrospective nature, programs enter data on the previous year’s operations. In this study, we focused on data submitted for the 2018 calendar year by programs that specifically self-identified as a PPC program. In the analysis, we focused on the core program information and the PPC specialty module to increase the usage and generalizability of the study. In alignment with the Registry’s historical reporting standards, optional program information questions that met an a priori 70% response rate were also included.

To control for variation in hospital size, PPC penetration rates were calculated by dividing the total number of initial consults by the total number of annual pediatric hospital admissions. Full-time equivalent (FTE) staff were standardized to 10 000 pediatric hospital admissions by multiplying total FTE staff by 10 000 and dividing by the hospital's total annual pediatric admissions. To test data normality, we used the Shapiro-Wilk test and presented medians and ranges for data not normally distributed. Box and whisker plots were used to examine FTE staff by discipline. Descriptive statistics and bivariate analyses, including simple linear regression, were performed by using IBM SPSS Statistics 24 (IBM SPSS Statistics, IBM Corporation). The UpSet plot was created with UpsetR.²⁶ 

Fifty-four inpatient programs, representing 29 states and the District of Columbia, submitted data and identified as a PPC program (11% of all participating programs) (Fig 1). All core program information and PPC specialty module questions received a 100% response rate, except the subsequent visits (response rate = 72%). Program funding (response rate = 87%) was the only optional program information element that met the 70% inclusion rate.

The majority of programs operated in hospitals that were not for profit (89%), urban (83%), and teaching (100%) (Table 1). These hospitals varied widely in size, from 75 to 700 beds (median 257). The largest cohort (n = 27, 50% of total) operated within independent, freestanding children’s hospitals. The total number of PPC programs across all hospital types is unknown, so the total prevalence could not be determined.

Although the oldest participating program started in 1997, most (N = 33, 61%) were established between 2008 and 2018. New program development peaked in 2011 and 2012 when 25% of participating programs were established (Fig 2). For the 47 programs reporting funding sources, 98% used hospital support, 74% used revenue from billing, and 72% used revenue from grants or philanthropy.

Forty-six programs (85%) were available for consultation to all pediatric oncology, pediatric intensive care, pediatric cardiac intensive care, and neonatal intensive care beds in their hospitals. Programs that reported limited availability to these units were restricted by either operational design or insufficient staffing. Most programs reported limited availability, with only 46% (25 of 54) of programs providing 24/7 coverage (defined in the survey as 24/7 telephone support and availability for consultation Monday through Friday) to referring clinicians, patients and families, and hospice providers (Table 2). Eighty-two percent of PPC programs reported that hospice services were readily available for pediatric referrals.

Clinically, 85% of programs reported providing services beyond inpatient PPC consultation (Table 2). Eighty-seven percent reported staff participation in nonclinicaland academic pursuits (Table 3). More than two-thirds reported staff pursuits of extramural grants and philanthropic funding for clinical care or research.

Nearly all programs offered services to children from birth to the 25th year of life. Less than half offered prenatal services or services to patients aged >25 years (Table 2).

Programs reported a median of 266 annual initial consults (range 36–1466). When standardized to hospital size, PPC programs reached a median penetration rate of 2.7% (range 0.3%–10.3%). In bivariate analyses, a greater number of annual initial consultations was significantly associated with a greater number of core interdisciplinary FTE staff (b = 34.6; P = .001), longer-standing programs (b = −14.6; P = .011), 24/7 availability for patients and families (b = 125.5; P < .000), and larger hospitals (b = .6; P = .003).

Programs also reported a median of 742 billable subsequent visits annually (range 25–10 000), corresponding to a median of 3.5 billable subsequent visits per initial consult (range <1–29).

Program staffing was highly variable, with programs using a wide range of professional disciplines (Table 4). Programs reported a median of 3.8 FTE staff per 10 000 hospital admissions (range 0.7–12.1) across the core interdisciplinary team of physicians, advanced practice registered nurses (APRNs), registered nurses (RNs), social workers, and chaplains (Fig 3).

Only 20 programs (37%) met national recommendations for team composition, with at least some FTE staff for the physician team (medical doctor or doctor of osteopathic medicine), nursing (RN or APRN), social work, and chaplaincy (Fig 4). Programs not meeting these recommendations were most often missing chaplaincy (52%) or social work (33%). Four programs (7%) were staffed by only 1 of the recommended disciplines, and 7 (13%) reported having no disciplines with at least 1.0 FTE staff member. Although 52 programs (96%) reported having FTE physicians, only 42 (78%) had physicians that were board-certified in hospice and palliative medicine.

Nearly two-thirds (63%) of respondents reported subjectively feeling moderately or extremely concerned about burnout risk, and 60% reported being short-staffed and requesting additional staff to meet current patient care demands. Comparatively, only 1 program reported not being concerned about burnout risk, and 19% reported being stable and not requesting additional staff.

This descriptive study reveals ongoing variability among PPC programs in the United States related to program structure, staffing, and services. When compared with previous studies, these results reveal a continually evolving field, with larger PPC teams seeing more annual consults.¹⁰ 

This study also updates critical benchmarks and draws attention to the inherent risks associated with program development in a rapidly evolving field, including issues related to program sustainability, clinician burnout, and quality of care. As national benchmarks for staffing, care delivery models, and consult volume may be instrumental in fostering program sustainability, PPC leaders can use this study to advance program development, align program structure with national standards, monitor inherent risks to program sustainability, facilitate conversations about resource needs, and guide future research initiatives to advance practice standards.²⁷ 

Across the 54 participating programs, only one item was consistent: providing care to neonates, infants, and children. Notably, programs reported a wide range of annual initial consults, with larger consult volumes associated with longer-standing programs, programs with more core staff, 24/7 availability to patients and families, and programs in larger hospitals. The range of core FTE staff per 10c000 hospital admissions (0.7–12.1) reveals the significantly different approaches that health care systems use to staff their programs.

These results highlight numerous challenges for PPC programs. Many programs are not yet meeting recommended practice standards, such as offering 24/7 availability to patients and families and staffing that includes physician, nursing, social work, and chaplaincy disciplines.^1,4–7  Rather than focusing on meeting these recommendations, programs may pragmatically determine staffing and care delivery models on the basis of organic program growth, the clinical units covered, and currently available funding and staffing. In addition, although the number of annual consults helps define the overall program workload, the clinical demand represented by these numbers may simply reflect the workload capacity of available staff rather than the total clinical demand for PPC services within the health care system. To clearly define ideal PPC staffing and care delivery models, future research efforts should focus on how specific models impact clinical demand, workload, quality, and both staff and patient and family outcomes.

Finally, the Patient Protection and Affordable Care Act of 2010 made it possible for children to qualify for hospice services while concurrently pursuing disease-directed treatments.²⁸  Despite this expanded access to hospice services, 40% of programs still reported resistance from hospice organizations to the provision of concurrent care, including refusing to enroll patients under the concurrent care clause. Future studies should also focus on how these barriers to access hospice care impact patient and family, staff, and health care system outcomes.

Most programs reported that staff are responsible for multiple discrete tasks, including direct patient care, education and training, academic endeavors, and pursuit of extramural grants and philanthropic funding. It is unclear how these tasks impact staff workload and stress or how these responsibilities compare with other pediatric disciplines. A more in-depth investigation into how these tasks are operationalized and compare with other pediatric disciplines would be beneficial in determining PPC staffing needs and the subsequent impact on patient and family outcomes.

Moreover, the majority of programs reported concern for burnout and an inability to meet clinical demand with available staffing, supporting concerns for PPC clinician burnout. However, further research into modifying factors (staffing and care delivery models, presence of interdisciplinary teams, clinician and program workload requirements, unmet clinical demand, culture and acceptance of PPC services, and financial support) is necessary to better understand these risks and design interventions to mitigate them.

Financial constraints remain a significant barrier to the growth and development of high-quality PPC programs.²⁰  Programs in this study use a variety of funding sources, with the majority reliant on philanthropy and hospital support. However, significant gaps remain in understanding (1) how programs allocate these funds across clinical and academic initiatives, (2) the financial sustainability of different clinical models, (3) the impact of reliable funding on program sustainability and staff wellness, and (4) the sufficiency of payment models for PPC services. Combined with our findings on program stability and burnout risk, these results raise the concern that some PPC programs may already be at significant risk of closure or restructuring because of operational challenges.²⁹  Ensuring both the availability and financial sustainability of PPC programs will be critical to safeguarding ongoing high-quality care for seriously ill children and their families.

This study has some notable limitations. As the total number of inpatient PPC programs in the United States remains unknown, we cannot determine how representative the results are of all PPC programs nationally. All Registry data are self-reported, with some questions requiring subjective responses; results are, therefore, subject to reporting errors and biases. Although the study sheds light on many aspects of PPC, including hospice and concurrent care, a more in-depth investigation is limited because of the data set. Finally, associations could not be meaningfully explored through multivariable models because of the small sample size.

In addition, in this study, we focused on PPC programs before the coronavirus disease 2019 pandemic. The long-term impact of the coronavirus disease 2019 pandemic on PPC programs remains uncertain because many critical variables continue to evolve. In early studies, researchers have suggested that the pandemic may worsen the moral distress and well-being of PPC clinicians.^30,31  In addition, even with these stressors and the potential increased need for PPC services, some programs are reporting staffing model challenges because PPC staff are being diverted to cover other clinical services deemed more critical during the pandemic. Most PPC programs appear to be experiencing challenges related to morbidity, mortality, cultural acceptance, and economic pressures on health care systems and other funding streams; however, it is currently unclear if these variables will ultimately be beneficial or detrimental to challenges facing PPC programs. This study offers a baseline for future comparisons to understand how PPC programs are impacted by the pandemic.

Most importantly, the authors acknowledge that the aim of this study was limited to describing PPC program characteristics. More work is needed to define which care delivery and staffing models lead to the highest quality of care and are most effective at meeting the clinical demand. This study suggests that an annual national registry can help meet this need. Moving forward, the Palliative Care Quality Collaborative will provide the opportunity for PPC programs to collect standardized operational and patient-level data, creating opportunities for standardized practice evaluation, collaborative quality improvement initiatives, and evaluation of different staffing and care delivery models on patient and health care system outcomes.^32,33 

In this study, we found that variability in PPC program staffing and care delivery models persist. Many programs are not meeting national guidelines related to minimum standards of practice, and findings indicate that long-term survival may be at risk for some programs. Future research is needed to define the most effective care delivery models, the staffing requirements of those models, and how different models impact patients and families, staff, and health care system outcomes. Additional opportunities exist to define patient care outcomes, refine standards of practice, and understand the clinical demand, workload, and sustainability of PPC programs. This work will be imperative to providing high-quality PPC services, ensuring the sustainability of PPC programs, and improving care for seriously ill children and their families.

We thank Rebecca Kirch, Dr Emma Jones, Rachael Heitner, and the PPC Research Network for their contributions to the PPC content in the National Palliative Care Registry.

APRN

advanced practice registered nurse

FTE

full-time equivalent

PPC

pediatric palliative care

RN

registered nurse