Abstract BACKGROUND Tissue biobanks play a vital role in oncology research; however, disparities have been reported in enrollment. The representativeness of biobanks is an important consideration in the extrapolation of research to the broader population. We sought to identify associations between sociodemographic characteristics and pediatric brain tumor bank enrollment at Dana-Farber Cancer Institute and Boston Children’s Cancer and Blood Disorders Center (DFCI/BCCBDC) to identify opportunities for improving equitable and representative enrollment. METHODS We conducted a retrospective cohort study of all patients qualified for enrollment in DFCI/BCCBDC Integrated Tissue and Clinical Data Bank for Patients with Neurological Disorders from 2014-2017 and extracted sociodemographic information from the medical record. The primary outcome was biobank enrollment, and the secondary outcome was approach for enrollment among non-enrolled patients. Exposures of interest included sex, language, interpreter need, race, and ethnicity. RESULTS We identified 652 eligible patients of which 580 (89%) enrolled and 72 (11%) were non-enrolled. Non-enrolled patients were more likely to be non-English speaking compared to enrolled patients (20% vs 11%; p = 0.0019) and were more likely to require an interpreter (15% vs. 8%; p = 0.025). Non-enrolled patients were more likely to identify as non-white/other/multiple races compared to enrolled patients (29% and 17%; p = 0.0023). Sex and ethnicity were not associated with enrollment. Of the 72 non-enrolled patients, 19 (26%) were approached to discuss enrollment but declined consent. There were no sociodemographic differences between approached and not approached patients. CONCLUSION While overall enrollment rates in this single center pediatric brain tumor bank were high, non-English primary language, interpreter need, and non-white race were significantly associated with non-enrollment. There were no differences in terms of approach for enrollment, suggesting that opportunities to improve representation may include understanding the reasons for declining enrollment and improving the consent process for patients from historically marginalized groups.
The American Academy of Pediatrics and the National Academy of Medicine recommend vitamin D supplementation for breastfeeding infants. However, compliance with this recommendation is poor. Maternal supplementation with vitamin D is a safe and effective alternative to achieving vitamin D sufficiency in breastfeeding infants, and mothers have indicated a preference for self-supplementation over infant supplementation. Research aim: We sought to explore Family Medicine clinicians' knowledge, attitudes, and practices regarding vitamin D supplementation recommendations for breastfeeding dyads.Fifty-six Family Medicine clinicians (including faculty physicians, resident physicians, and nurse practitioners/physician assistants) completed an online, anonymous survey regarding their knowledge and practices concerning vitamin D supplementation for breastfeeding infants.The vast majority of clinicians (92.9%) correctly identified the American Academy of Pediatrics' 2008 recommended dose for vitamin D supplementation in breastfeeding infants and estimated recommending vitamin D supplementation of exclusively breastfeeding infants 70.1% of the time. If all options were equivalent, clinicians would prefer to offer maternal or infant supplementation (50%) or maternal supplementation (37.5%) over infant supplementation (12.5%). Most (69.6%) preferred daily over monthly supplementation regimens.Family Medicine clinicians are knowledgeable regarding current recommendations for vitamin D supplementation in breastfeeding infants. They are also open to recommending maternal supplementation or offering parents a choice of maternal or infant vitamin supplementation.
Adolescents and young adults (AYAs) with advanced cancer identify normalcy as an important component of quality end-of-life care. We sought to define domains of normalcy and identify ways in which clinicians facilitate or hinder normalcy during advanced cancer care. This was a secondary analysis of a qualitative study that aimed to identify priority domains for end-of-life care. Content analysis of semi-structured interviews among AYAs aged 12-39 years with advanced cancer, caregivers, and clinicians was used to evaluate transcripts. Coded excerpts were reviewed to identify themes related to normalcy. Participants included 23 AYAs with advanced cancer, 28 caregivers, and 29 clinicians. Participants identified five domains of normalcy including relationships, activities, career/school, milestones, and appearance. AYAs and caregivers identified that clinicians facilitate normalcy through exploration of these domains with AYAs, allowing flexibility in care plans, identification of short-term and long-term goals across normalcy domains, and recognizing losses of normalcy that occur during cancer care. AYAs with cancer experience multiple threats to normalcy during advanced cancer care. Clinicians can attend to normalcy and improve AYA quality of life by acknowledging these losses through ongoing discussions on how best to support domains of normalcy and by reinforcing AYA identities beyond a cancer diagnosis.
10005 Background: Racial and socioeconomic disparities have not been comprehensively investigated in high-risk neuroblastoma (HR NBL). Prior Children’s Oncology Group (COG) investigations have demonstrated population-based disparities in late relapse rates among Black children, and trial-based disparities in relapse and survival among children living in poverty receiving post-consolidation immunotherapy. It is unknown whether these disparities persist in upfront trials for newly diagnosed patients. We leveraged COG data to investigate race, ethnicity, and socioeconomic disparities in a cohort of children with HR NBL treated on upfront clinical trials from 2007-2016. Methods: Retrospective cohort study of children enrolled on upfront COG HR NBL trials ANBL0532, ANBL09P1, and ANBL12P1. Race and ethnicity were the primary exposures categorized as: Black Non-Hispanic (BNH); Hispanic; Other Non-Hispanic (ONH); or White Non-Hispanic (WNH). Poverty was the secondary exposure, defined as household (public insurance only vs others), area (census-defined high-poverty ZIP code with >20% of population living below 100% Federal Poverty Level (FPL) vs <20% below 100% FPL), and rural (Census-defined rurality measures linked to ZIP code). Overall (OS) and event-free (EFS) survival from time of trial enrollment were plotted by Kaplan-Meier methods; associations with race/ethnicity and poverty were evaluated by log-rank tests. Results: Among 696 children, 16% were BNH, 11% Hispanic, 4% ONH, and 69% WNH. One-third (33%) of children were household poverty-exposed, 26% area poverty-exposed, and 15% rural-exposed. Tumor stage and biology did not differ by race/ethnicity or poverty measures. Five-year OS differed significantly by race/ethnicity (47% Hispanic vs. 50% ONH vs. 61% WNH vs. 62% BNH; p=0.047). Five-year OS was inferior among children exposed to household-poverty (53% vs. 63%; p=0.036) and neighborhood-poverty (54% vs. 62%; p=0.050) compared to unexposed children. There was no difference in OS by rurality. Similar directionality in 5-year EFS outcomes by race/ethnicity and poverty were observed without statistical significance. Conclusions: Race/ethnicity and poverty-exposure are associated with inferior OS outcomes among children with HR NBL despite uniform planned treatment on upfront COG trials. Investigation of mechanisms driving these disparities, including disparate early phase trial enrollment are ongoing to inform targeted health equity interventions to improve outcomes.
10025 Background: Poverty is associated with inferior psychosocial function among parents of children with cancer. Severe parental distress during treatment predicts future poor mental health for both parents and children. It is also associated with impaired parental cognitive bandwidth and executive function, which may have implications for treatment adherence. Efforts to identify poverty-exposures amenable to intervention are essential to improving survivorship quality of life for the > 90% of children with acute lymphoblastic leukemia (ALL) who will be long-term survivors. Household material hardship (HMH) is a targetable poverty exposure defined as at least 1 of 3 unmet basic needs including food, housing, or utilities. Dana-Farber Cancer Institute (DFCI) ALL Consortium trial 16-001 is the first pediatric oncology clinical trial to systematically evaluate HMH. We investigated the hypothesis that HMH exposure independently predicts severe parent psychological distress during ALL therapy. Methods: Patients with newly diagnosed ALL ages 1-17 years were enrolled on the DFCI 16-001 embedded HMH cohort study at 8 U.S. and Canadian centers. Secondary interim analyses used baseline (within 32-days of trial enrollment) and 6-mos parent-reported sociodemographic data, the Kessler-6 (K6) Psychological Distress scale, and trial-collected child and disease data. Severe psychological distress was defined as a K6 > = 13. Multivariable cox regression evaluated baseline HMH-exposure and parent distress at baseline and 6-mos adjusting for child’s initial ALL risk group (Very High Risk (VHR) vs other) and marital status (single vs dual parent). Results: Among 258 families with evaluable data, 34% reported baseline HMH. Families were predominantly English-speaking (54%) dual parent households (71%). Children were a median of 5.7 years (IQR 1.0-17.99) at diagnosis and predominantly non-Hispanic white (66%) with expected disease distribution by immunophenotype (84% B-cell). HMH (odds ratio (OR) 2.18, 95% confidence interval (CI) 1.0-4.31, p = 0.025) and VHR initial risk group (OR 2.32; 95% CI 1.06-5.06, p = 0.035) were independently associated with baseline severe psychological distress. Only HMH was independently associated with 6-mos severe psychological distress (OR 4.93, 95% CI 1.80-13.48, p = 0.002). Future analyses will investigate race and ethnicity associations with parental distress pending trial accrual for statistical power. Conclusions: HMH, a modifiable poverty exposure, is significantly associated with severe parent psychological distress at diagnosis that persists 6-months into pediatric ALL therapy. These findings identify a cohort at high risk of inferior mental health outcomes, and affirm the need for HMH-targeted interventions to support children and parents during cancer treatment to reduce poverty-associated outcome disparities in survivorship.
Black and Hispanic children with leukemia experience inferior survival compared to non-Hispanic White (NHW) children. Identifying modifiable social determinants of health can inform intervention targets to address inequities. We characterized the frequency of income poverty and household material hardship (HMH) by race/ethnicity in a clinical trial cohort with de novo acute lymphoblastic leukemia. Compared to NHW families, Black and Hispanic families reported more frequent HMH (19% vs. 47% vs. 68% respectively); low-income (27% vs. 52% vs. 74%), and combined low-income and HMH (12% vs. 37% vs. 52%) poverty exposures. Disparate poverty exposures are interventional targets to address racial/ethnic outcome inequities.
BACKGROUND Racial and ethnic minority children with cancer disproportionately receive intensive care at the end of life (EOL). It is not known whether these differences are goal‐concordant or disparities. The authors sought to explore patterns of pediatric palliative care (PPC) and health care utilization in pediatric oncology patients receiving subspecialty palliative care at the end‐of‐life (last 6 months) and to examine goal‐concordance of location of death in a subset of these patients. METHODS This was a retrospective cohort study of pediatric oncology patients receiving subspecialty palliative care at a single large tertiary care center who died between January 2013 and March 2017. RESULTS A total of 115 patients including 71 White, non‐Hispanic patients and 44 non‐White patients (including 12 Black patients and 21 Hispanic patients) were included in the analytic cohort. There were no significant differences in oncologic diagnosis, cause of death, or health care utilization in the last 6 months of life. White and non‐White patients had similar PPC utilization including time from initial consult to death and median number of PPC encounters. Non‐White patients were significantly more likely to die in the hospital compared to White patients (68% vs 46%, P = .03). Analysis of a subcohort with documented preferences (n = 45) revealed that 91% of White patients and 93% of non‐White patients died in their preferred location of death. CONCLUSIONS Although non‐White children with cancer were more likely to die in the hospital, this difference was goal‐concordant in our cohort. Subspecialty PPC access may contribute to the achievement of goal‐concordant EOL care.
