UC Berkeley

Racial and Ethnic Differences in Cardiovascular Disease Medication Management for Patients with Diabetes

by

Ana Helena Traylor

Doctor of Philosophy in Public Policy

University of California, Berkeley

Professor Stephen Raphael, Chair

Extensive research documents inequities in the quality of health care provided to members of racial and ethnic minority groups in the United States. Cultural differences between patients and health care providers may contribute to health disparities by increasing the likelihood of physician bias, patient distrust and patient-provider miscommunication.

This dissertation uses data from Kaiser Permanente's Northern California Diabetes Registry of 2005 to 1) examine racial and ethnic disparities in cardiovascular disease (CVD) risk factor management for patients with diabetes, 2) explore the prevalence and predictors of patient-physician racial/ethnic match and 3) examine the association between racial and ethnic match and CVD risk factor management for patients with diabetes. I use a cross-sectional observational design and conduct a series of logistic regression models that adjust for patient, physician and medical facility characteristics.

I found significant differences in CVD risk factor control, treatment intensification for patients in poor control and CVD risk factor medication adherence. Compared to white patients, African American patients were less likely to be at target for diabetes (Hemoglobin A1c), hyperlipidemia (LDL-c) and Hypertension (SBP). Latino and Asian patients were less likely to be at target A1c levels, but more likely to be at target LDL levels. Spanish speaking patients were also less likely to be at target A1c levels.

Racial differences in treatment intensification were complex. When not at target levels, African American and Asian patients were less likely than white patients to have treatment intensified for A1c medications. However, African American patients were more likely to have treatment intensified for SBP and Asian patients were more likely than white patients to have treatment intensified for LDL and SBP medications. There were no disparities in intensification for Latinos patients; in fact Latino patients were more likely to have treatment for LDL intensified. Similarly, Spanish-speaking patients were more likely to have treatment intensified for LDL.

I found that African American and Latino patients were least likely to be racial matched, though Spanish speaking Latino patients were more likely than English speaking Latino patients to have a same race physician. Compared with patients who were assigned a physician by the health care organization, patients who chose their physicians were more likely to have a same race provider. While statistically significant for all racial and ethnic groups, this relationship was strongest for African American and Latino patients. Availability of a same race provider was the strongest predictor of patient-physician race concordance for African American and Latino patients.

I examined the association between race concordance and intermediate CVD risk factor outcomes for African American and Latino patients. Race and language concordance did not impact risk factor control or treatment intensification. However, race/ethnicity concordance was marginally associated with better adherence to medication for African American patients and language concordance was marginally associated with adherence for Spanish speaking Latino patients.

By allowing for more race and language concordance between patients and providers, increased minority representation in the medical professions is hypothesized to improve the cultural competence of health care delivery. Given wide and persistent disparities in health for African American and Latino patients, this dissertation examined the hypothesis that by increasing opportunities for race, ethnicity and language concordance, race-conscious medical school and workforce diversity efforts might lead to improvements in public health and a reduction in health disparities.

The results further highlight the need for continued efforts to measure, understand and address racial and ethnic disparities. The results presented here suggest that increasing the number and proportion of underrepresented minorities might lead to important improvements in patient adherence to medication. However, these efforts alone, will not eliminate gaps in CVD medication management for patients with diabetes.