Medical Distrust in Applied Genomic Sequencing: Prediction, Perceptions, and Patient outcomes

Abstract

Background: A growing collection of research shows medical distrust negatively impacts a wide range of health behaviors and outcomes including: rates of preventive screening, treatment adherence, and satisfaction with care. Despite evidence suggesting medical distrust may impact genomic sequencing research participation, limited work has been conducted to explore who may be affected by medical distrust and how health outcomes associated with genomic sequencing are impacted. Objective: The goal of this work is to: 1) Identify a prevalence estimate and predictors of medical distrust in a clinical research sample of adults undergoing exome sequencing to diagnose hereditary cancer syndromes. 2) Analyze the impact of medical distrust on patient perceptions of utility and feelings associated with the receipt of genetic results. 3) Assess the impact of medical distrust on the uptake of recommended care and patient-initiated health behavior changes as well as willingness to share genetic information with health care providers and family members. Methods: Data for these analyses was collected as part of the Cancer Health Assessment Reaching Many (CHARM) project conducted by Kaiser Permanente Northwest (KPNW). This project targets AHRQ priority populations in a multi-site implementation study of clinical exome sequencing for heredity risk of cancer syndromes and includes a nested treatment arm testing literacy-focused genetic counseling tailored to low health literacy populations. Participants were recruited based on their familial history of cancer though in-clinic contact and via text messages and email. Measures were collected in-clinic via tablet using web-based surveys during the baseline timepoint and submitted online during two subsequent timepoints. Analyses: Prevalence estimates were assessed using cut points from previous literature on medical distrust. We used three models of prediction selection to determine predictive patient-level characteristics: backward stepwise regression and two least absolute shrinkage and selection operator model, varying the level of conservativeness of selection. To analyze the relationship between distrust and patient reports of utility and feelings associated with return of genetic test results we used linear multivariable modeling. A final multiple logistic regression was used to assess the impact health care system distrust has on the likelihood patients adopt recommended health behavior changes based on genomic test results and whether patients share those results with family members and providers. Results: The point prevalence estimate showed medical distrust to be as present in a medical genomic research context (32%) as it is in other medical contexts. African American race/ethnicity, poor access to health care, transgendered/non-binary gender identity, worse mental health status, lower income and higher education predicted high medical distrust. We found no impact of medical distrust on feelings related to the return of genomic results there was, however, a negative relationship between medical distrust and perceptions of utility of genomic results. There was inadequate data to analyze an effect between medical distrust and following medically recommended advice. Personal health behavior change based on genomic results were not impacted by level of medical distrust. Higher medical distrust was related to lower odds of sharing genomic results with family members, but no relationship between medical distrust and discussing results with health care provider was observed. Discussion: Patient socio-demographic, health status, and access to health care variables predict medical distrust. While medical distrust may not impact feelings related to receipt of medical results, it seems to lower the perceived utility of genomic sequencing. This may illustrate a mechanism by which medical distrust discourages positive health behaviors. The negative impact of medical distrust on familial information sharing undermines a unique benefit of genomic medicine. However, the lack of association between medical distrust and personal health behavior change and discussion of results with health providers may indicate that information gained from genomic sequencing could be especially resilient to the negative impacts of medical distrust.