Feasibility and acceptability of social determinants of health data collection in the context of a Children’s Oncology Group trial.

Abstract

10010 Background: Children who are Black, Hispanic or of proxied low socioeconomic status experience inferior survival outcomes despite receipt of uniform treatment on Children’s Oncology Group (COG) trials. Investigation of mechanisms underlying these disparities and evaluation of interventions to address them has been limited by inadequate data. Prior COG trials have collected child race, ethnicity, insurance and zip code—non-modifiable data elements that are prone to misclassification bias—as proxies for exposure to adverse social determinants of health (SDOH). Collection of parent-reported SDOH data provides key information not previously assessed in COG trials, including modifiable exposures such as household material hardship (HMH): food, housing, utility or transportation insecurity. We report the preliminary feasibility and acceptability of the first COG trial to prospectively embed SDOH data collection. Methods: ANBL1531 (NCT03126916), the current frontline, groupwide COG trial for patients with high-risk neuroblastoma, was amended in September 2019 to include an opt-in, correlative aim to evaluate parent-reported SDOH. Parents of patients < 18 years could opt-in during initial ANBL1531 consent to complete a brief, single-timepoint survey at their enrolling site during Induction Cycle 1. Survey domains included parent-reported demographics, HMH, household income, and social support. Paper/pencil surveys available in English, Spanish or French were self-completed or read-aloud by site personnel and could be administered in any language with an interpreter. Surveys were submitted to a central disparities study team for data cleaning and entry into an electronic database. Results: As of December 31, 2022, 360/413 eligible participants opted-in to the embedded SDOH aim across 101 COG sites (87.2% consent rate). Among participants, 316 surveys (87.8%) were completed a median of 11 days post-enrollment. Four (1.1%) parents declined survey completion (3 withdrew consent). Reasons for non-completed surveys included: survey still pending within protocol-specified window (n = 10), institutional COVID-19 restrictions (n = 6), site scheduling issues (n = 6), patient came off ANBL1531 protocol (n = 5) and unknown (n = 13). Data missingness for HMH, the primary SDOH of interest, was 0.8%. Surveys were administered in English, Spanish, French, Vietnamese, Arabic, Mandarin and Tamil. Conclusions: Collection of parent-reported SDOH data as an optional, embedded component of a groupwide COG trial is feasible and acceptable, as demonstrated by remarkable willingness to participate and infrequency of missing data. Incorporation of this approach in future National Clinical Trials Network trials is essential to inform impactful health equity research, including the development and evaluation of targeted interventions to address outcome disparities.