625: DISPARITIES IN PICU RESEARCH PARTICIPATION ARE INFLUENCED BY APPROACH, RACE, AND LANGUAGE

Abstract

Introduction: Representative inclusion in critical care research is important for generalizability. There is little data on disparities in enrollment in studies in the pediatric intensive care unit (PICU), where illness severity and time-sensitivity may uniquely affect participation. We aimed to determine whether there were racial, linguistic, and socioeconomic disparities in consent rates in PICU research. We hypothesized the existence of disparate consent, part of which may be explained by differing rates of approach. Methods: Consent logs were reviewed for all studies enrolling patients at the Children’s Hospital of Philadelphia PICU between 2011-2020. Exposures included race/ethnicity (White, Black, Hispanic, and Other), language (English, Spanish, Arabic, or Other), religion (None, Christian, Muslim, Jewish, or Other), and social deprivation index (SDI, composite of multiple socioeconomic indicators). A priori confounders included sex, age, length of stay prior to eligibility, illness severity, patient origin prior to PICU, elective or emergent admission, and observational or interventional study. Logistic regressions tested associations between individual exposures (race, language, religion, SDI) and two outcomes: consent to research among all eligible patients, and consent to research among those approached for consent. Results: Of 2351 eligible children, consent was less likely to be obtained from families of Black (adjusted OR=0.64, 95% CI 0.52-0.78) and Other race (aOR=0.59, 95% CI 0.47-0.74), Arabic (aOR=0.45, 95% CI 0.28-0.71) and Other language (aOR=0.13, 95% CI 0.06-0.28), and Muslim children (aOR=0.54, 95% CI 0.39-0.76). There was no association between SDI and consent. Among children approached for research, lower consent odds persisted for Black race (aOR=0.66, 95% CI 0.52-0.84) and Other language (aOR=0.36, 95% CI 0.13-0.96). Conclusions: There are racial, linguistic, and religious disparities in consent rates for PICU studies partly attributable to disparate rates of approach. Our findings show that disparities in participation in PICU research reflect both disparate rates of contact with researchers as well as disparities in actual consent, suggesting opportunities for improvement in reducing barriers to approach and building trust in marginalized populations.