Race and Gender Differences in Patient-Centered Communication and Diabetes Management Across the Life Course

Abstract

In this dissertation, I present a series of papers that describe how race, gender, and age impact patient-provider experiences and the management of diabetes. In Chapter 2, I use multivariate linear regression and logistic regression to analyze the association between race, gender, SES, and other covariates on a composite measure of patient-centered communication fielded in the Health and Retirement Study (HRS) in 2011. I find that women are more likely than men to report more frequent experiences of patient-centered communication in a sample of older adults (ages 50+) from the HRS. I also find that in 2011, not having enough money for food is a predictor for less frequent patient-centered communication. Regarding racial differences, I found that non-Hispanic Black women had higher odds of reporting more frequent experiences of patient-centered communication than White or Hispanic women in 2011. Patient-centered communication is a component of healthcare quality and Chapter 1 provides a base year for understanding how communication may have changed for older adults over time. In Chapter 3, I use two waves of the Health and Retirement Study to continue describing healthcare quality for an aging population with an increasing chronic disease burden. I perform a change-score analysis for individuals that answered questions about patient-provider communication in both 2011 and 2019 to analyze whether PCC changed over time in the context of major health policy reform: the implementation of the Patient Protection and Affordable Care Act. I also use multivariate linear regression and logistic regression with the full sample that completed the 2019 HRS Health Survey to analyze whether Medicaid expansion has any association with healthcare communication. I find that Medicaid expansion is not associated with care communication for either the repeat or full sample. Further, comparing 2011 and 2019, I find that the racial and gender differences in patient-centered communication decline over time. Finally, in Chapter 4, I use the Behavioral Risk Factor Surveillance System to understand whether Medicaid expansion differently affected low-income Black and White people with diabetes. I use triple difference-in-differences models with national level data to understand if the gap between measures of diabetes management changed over time in expansion and non-expansion states for Black and White adults; the three differences are between race groups, between expansion and non-expansion states, and between time points. I also perform separate difference-in-difference analyses for Black and White low-income adults with diabetes in Kentucky (a Medicaid expansion state) and Tennessee (a non-expansion state). I find that nationally, more Black than White low income PWD increased getting professional foot checks at least once per year. When I compare two states, I find that Medicaid expansion affected different measures of diabetes management; for Black Kentuckians expansion increased the percentage points of people getting professional foot checks and HbA1c checked at least twice per year. For White Kentuckians, Medicaid expansion increased the percentage points of people receiving a flu shot and reporting good or better health. I suggest a theoretical framework for integrating theories about structural sexism and organizations as producers of hierarchies in Chapter 2 and describe theories that explain why individual states may produce different health outcomes through variation in policy implementation in Chapter 4. Together, these studies contribute to the literature about the impact of race, gender, and age on how people experience healthcare and health policy.