181 - Experiences of Discrimination in Healthcare and Influence on Beliefs about Research: Creating a Culture of Trust

Publication Number: 181.4880

Christine B. Turley, Atrium Health Wake Forest School of Medicine, Charlotte, NC, United States; Kathleen Barber, Levine Children's Hospital, Charlotte, NC, United States; Dustin M.. Long, Wake Forest School of Medicine of Wake Forest Baptist Medical Center, Winston-Salem, NC, United States; Hillary Herr, Levine Children's Hospital, Charlotte, NC, United States; Gang Liu, Levine Children's Hospital, Charlotte, NC, United States; Morgan Walls, Levine Children's Hospital, Charlotte, NC, United States; Jack Weaver, Levine Children's Hospital, Charlotte, NC, United States; Michael Duncan, Atrium Health Levine Children's Hospital, Charlotte, NC, United States; Tiffany Linville, Levine Children's Hospital, Charlotte, NC, United States; Nathan Fleishman, Atrium Health Levine Childrens Hospital, Charlotte, NC, United States; Dennis Schellhase, Levine Children's Hospital, Charlotte, NC, United States

Background: The national prevalence of patient-reported experiences of discrimination in US healthcare settings (EDHC) was recently shown to be 21%. The frequency of EDHC in parents/caregivers(P/CG) seeking care for their children has not been studied, and the impact of EDHC on research beliefs is unknown.

Objective: Evaluate the frequency of EDHC in P/CG attending pediatric healthcare visits and the relationship to beliefs about how one’s racial/ethnic (R/E) group is treated in research.

Design/Methods: An anonymous paper survey was given to P/CG of children at 3 specialty and 2 general pediatric clinics. The survey was offered 3 times at 6-month intervals (mid and final period data included here). Questions were based on previously validated surveys: 1 question each about research attitude and trust, prior research participation, belief about how one’s R/E group is treated in research, personal EDHC, and demographics. Univariate logistic regression models were used to determine association between EDHC, demographics and beliefs about R/E treatment in research. A machine learning procedure, oblique random forests (ORF) for classification, was used to predict beliefs about R/E treatment in research with the importance of demographics and EDHC on predictions calculated. Each factor’s importance was calculated using ANOVA; partial dependence was used to explain ORF’s predictions. IRB approval was obtained.

Results: 3424 surveys were distributed, 2470 returned (72%),1683 completed all non-demographic questions (68%), no skip pattern was noted. Demographic features are seen in Table 1. 36% of NHB/AA respondents reported EDHC, 16% for H, 3% for NHW (p < 0.0001). The odds of EDHC were 19 times greater for NHB/AA v. NHW, 6 times higher for H v. NHW. (Fig. 1) Using ORF, EDHC had markedly high importance in predicting the belief that one’s R/E group is treated worse in research (0.63), the next highest variable, race, was 0.12. When predicting the impact of EDHC on belief about treatment in research, each R/E showed the same pattern. Without EDHC, the most commonly predicted belief was that treatment in research was the same. With EDHC, a universally flattened pattern with marked reduction in belief about treatment being the same, NHB/AA with greatest negative impact. (Fig. 2)

Conclusion(s): EDHC are common in P/CG of pediatric patients for NHB/AA and H individuals. The importance of EDHC outweighs all factors in shaping belief about R/E treatment in research. This is the first example of a direct relationship of EDHC cascading into research beliefs. Eliminating EDHC is crucial to improving research beliefs and health equity.

Demographic Characteristics of Respondents from Pediatric Clinics


Figure 1. Likelihood of Experiencing Discrimination in Healthcare and Impact of Discrimination on Beliefs About Research, OR with 95% CI


Figure 2. Prediction of Belief About Negative Treatment in Research Based on Experience of Discrimination; Oblique Random Forests (ORF)