Racial/ethnic differences in medical mistrust among breast cancer patients: �What are the roles of network characteristics?

YAMILÉ MOLINA, MS, MPH, PHD

ASSISTANT PROFESSOR

SCHOOL OF PUBLIC HEALTH

CANCER CENTER

CENTER FOR RESEARCH ON WOMEN & GENDER

UNIVERSITY OF ILLINOIS AT CHICAGO

Topics

• Background: Substance + Networks
• Methods: 2-tiered Egocentric Data
• Results: Tons O’ Tables
• Discussion: Interpretation, Limitations & Next Steps

Background: Why Care About Medical Mistrust in the Breast Cancer Context?

Medical Mistrust

Anxiety & Depression During and Post Diagnosis

Lack of Genetic Counseling, Screening, & Diagnostic Care; Later Stage at Diagnosis

Inadequate Breast Cancer Treatment Uptake

Lower Perceived Quality of Care & Satisfaction with Care

Bickell et al., 2009; Hillen et al., 2011; Katapodi et al., 2010; Molina et al., 2014; Sheppard et al., 2013; Yang et al., 2011

Historical & Collective Context

Vicarious Experiences with Healthcare

Social Support/Pressure

Personal Experiences

Background: Why Care About Race/Ethnicity in the Breast Cancer Context?

As of 2009; Luckett, 2011; Yanez, 2011

Background: Networks & Medical Mistrust 🡪 Does Race/Ethnicity Matter?

Vicarious Negative Interpersonal Healthcare Experiences

Historical & Collective Traumas & Systemic Exclusion in Medical Settings

Social Support & Assurances to Avoid HealthCare System

Medical Mistrust

Historical & Collective Access to High-Quality Systems

Vicarious Positive Interpersonal Healthcare Experiences

Social Support & Assurances to Trust Healthcare System

Potential Differences in Collective/Network Experiences and Perceptions About the Medical System

(Molina et al., 2015; 2016a)

Personal Positive Interpersonal Healthcare Experiences

Personal Negative Interpersonal Healthcare Experiences

Background: Networks & Medical Mistrust 🡪 Does Race/Ethnicity Matter?

Historical Collective Experiences

Vicarious Interpersonal Healthcare Experiences

Medical Mistrust

Potential Differences in the Perceived Importance of Close, Tight-Knit Communities/Networks

(Molina et al., 2013; 2015a,b; 2016b; 2017a,b)

Personal Interpersonal Healthcare Experiences

Historical Collective Experiences

Vicarious Interpersonal Healthcare Experiences

Personal Interpersonal Healthcare Experiences

Research Question: �Do Relationships Between Network Characteristics & Medical Mistrust Among Breast Cancer Patients Vary by Race/Ethnicity?

Methods: Breast Cancer Care in Chicago Study (PI: Rauscher)

• Population-based study of patients who:
• Experienced an in situ/invasive breast cancer diagnosis during 2005-2008
• Reported race/ethnicity as non-Latina White (NLW), NL Black, or Latina
• Resided in Chicago at time of diagnosis

​

• 989 interested patients (56% of contacted)
• completed computer-assisted personal interview, including network information
• identified/recruited up to 5 people within their social networks to complete identical computer-assisted personal interviews

​

• 886 participants had complete data and included in analytic sample

Methods: Measures

• Medical mistrust: 10-item Health Care System Distrust Scale (α= 0.70) administered to patients who reported regular providers and/or providers

​

• Network: Network name generator (people who have provided help since diagnosis); name interpreters (race/ethnicity, gender) and name inter-relaters (relationships between people)
• Composition: for gender and for race/ethnicity
• Density: summation of existing relationships between people within network divided by the maximum number of possible relationships

​

• Other variables:
• Age
• Socioeconomic status: household income, education, census-tract level concentrated disadvantage, census-tract level concentrated affluence
• Healthcare access: regular provider (Yes/No), healthcare insurance (No/Public/Private)
• Physical exam within past year (Yes/No)

Results: Study Sample Characteristics

**p≤.001; ****p ≤.0001

Results: Study Sample Characteristics

**p≤.001; ****p ≤.0001

Results: Main Effects of Patient Race/Ethnicity & Networks

¹Other variables: network density, age, household income, education, concentrated disadvantage, concentrated advantage, having a regular provider, insurance status, & no physical exam. **p≤.001; ****p ≤.0001

Results: Network x Race/Ethnicity Effects

Other variables in addition to main effects and network variables: age, household income, education, concentrated disadvantage, concentrated advantage, having a regular provider, insurance status, & no physical exam **p≤.001; ****p ≤.0001

Results: Stratified Models

Other variables in addition to main effects and network variables: age, household income, education, concentrated disadvantage, concentrated advantage, having a regular provider, insurance status, no physical exam, type of detection. **p≤.001; ****p ≤.0001

Discussion

• Preliminary findings
• Latina breast cancer patients’ social networks appear to be somewhat less dense and more heterogeneous relative to other women’s networks 🡪 maybe due to Chicago dynamics
• Tighter knit communities appear to be associated with lower medical mistrust among Latinas in particular 🡪 maybe due to cultural emphases and/or interactions with immigrant identity
• These particular network characteristics may not be as powerful predictors for Black patients’ mistrust 🡪 maybe due to greater influence of socio-contextual factors and/or more complex network processes

​

• Limitations:
• No ‘cut-off’/magnitude guidelines for medical mistrust
• Sampling/measure administration
• Name generator focused on post-diagnosis support

​

• Ego-report ties
• Potentially not ‘independent’
• Relatively ‘old’ data

Next Steps

• Present this @ INSNA next Friday & Submit Brief Report by Sept 2017
• BCCC data
• Characterize if network characteristics are associated with ‘patterns of mistrust’ across the care continuum
• Incorporate alters’ medical mistrust and networks in models relating to ego medical mistrust
• Examine the “weight” of ethnic and gender concordance among Latino egos and their alters
• Related NCI grants awarded/written while SN&H Fellow
• K01: Testing the effectiveness of a train-the-trainer breast cancer intervention on Latinas’ communication and network structure re: breast cancer relative to an educational intervention
• R21: Examining how an individual-level intervention may have affected communication and behaviors re: breast cancer among Black breast cancer survivors

Thanks, DNAC SN&H! �(Especially Jake & Tyson & Jim)