Association between ethnicity, language and metformin prescriptions in U.S. community health center patients with diabetes.

BACKGROUND: Metformin treatment is a recommended first-line medication for patients with type 2 diabetes. Latino patients are subject to factors that may modify their level of diabetes care, including medication prescription. We evaluated the odds of and times to metformin prescription among non-Latino whites, English-preferring Latinos, and Spanish-preferring Latinos with diabetes. METHODS: We constructed a retrospective cohort of 154,368 adult patients from 835 community health centers (CHCs) across 20 states who were diagnosed with diabetes during the study. Patients were from non-Latino white, English-preferring Latino, and Spanish-preferring Latino ethnic/language groups. We modeled adjusted odds of metformin prescription and adjusted hazards (time-to-event) of metformin prescription after diabetes diagnosis and high hemoglobin A1c (HbA1c > 9) test results. RESULTS: English-preferring Latinos had similar odds of metformin prescription (Odds Ratio (OR) = 1.01 (95% CI = 0.93, 1.09)), slightly lower time to metformin prescription after diabetes diagnosis (Hazard Ratio (HR) = 1.06(95% CI = 1.04, 1.09)), and similar time to metformin prescription after a high HbA1c result (HR = 1.04 (0.99, 1.09)) compared to non-Latino whites. Spanish-preferring Latinos had higher odds of metformin prescription (OR) = 1.42 (95% CI = 1.33, 1.52), and less time to prescription after diabetes diagnosis (HR = 1.18 (1.15, 1.20)) and after a high HbA1c result (HR = 1.15 (1.11, 1.20)). CONCLUSIONS: Our analysis of metformin prescription patterns among non-Latino whites, English-preferring Latinos, and Spanish-preferring Latinos did not suggest a lower or slower tendency to prescribe metformin in Latino patients. Understanding disparities in diabetes diagnosis may require further investigation of medication adherence barriers, diet and exercise counseling, and multi-level influences on diabetes outcomes in Latino patients.

Atherosclerotic Cardiovascular Disease Primary and Secondary Prevention in Latino Subgroups.

BACKGROUND: Studies assessing equity in the prevention of atherosclerotic cardiovascular disease (ASCVD) for Latinos living in the USA collectively yield mixed results. Latino persons are diverse in many ways that may influence cardiovascular health. The intersection of Latino nativity and ASCVD prevention is understudied. OBJECTIVE: To determine whether disparities in ASCVD screening, detection, and prescribing differ for US Latinos by country of birth. DESIGN: A retrospective cohort design utilizing 2014-2020 electronic health record data from a network of 320 community health centers across 12 states. Analyses occurred October 1, 2022, to September 30, 2023. PARTICIPANTS: Non-Hispanic White and Latino adults age 20-75 years, born in Cuba, Dominican Republic, El Salvador, Guatemala, Honduras, Mexico, and the USA. EXPOSURES: Ethnicity and country of birth. MAIN MEASURES: Outcome measures included prevalence of statin eligibility, of having insufficient data to establish eligibility, odds of having a documented statin prescription, and rates of statin prescriptions and refills. We used covariate-adjusted logistic and generalized estimating equations logistic and negative binomial regressions to generate absolute and relative measures. KEY RESULTS: Among 108,672 adults, 23% (n = 25,422) were statin eligible for primary or secondary prevention of ASCVD using American College of Cardiology/American Heart Association guidelines. Latinos, born in and outside the USA were more likely eligible than Non-Hispanic White patients were (US-born Latino OR = 1.55 (95% CI = 1.37-1.75); non-US-born Latino OR = 1.63 (95% CI = 1.34-1.98)). The eligibility criteria that was met differed by ethnicity and nativity. Latinos overall were less likely missing data to establish eligibility and differences were again observed by specific non-US country of origin. Among those eligible, we observed no statistical difference in statin prescribing between US-born Latinos and non-Hispanic White persons; however, disparities varied by specific non-US country of origin. CONCLUSION: Efforts to improve Latino health in the USA will require approaches for preventing and reversing cardiovascular risk factors, and statin initiation that are Latino subgroup specific.

Answering calls for rigorous health equity research: a cross-sectional study leveraging electronic health records for data disaggregation in Latinos.

INTRODUCTION: Country of birth/nativity information may be crucial to understanding health equity in Latino populations and is routinely called for in health services literature assessing cardiovascular disease and risk, but is not thought to co-occur with longitudinal, objective health information such as that found in electronic health records (EHRs). METHODS: We used a multistate network of community health centres to describe the extent to which country of birth is recorded in EHRs in Latinos, and to describe demographic features and cardiovascular risk profiles by country of birth. We compared geographical/demographic/clinical characteristics, from 2012 to 2020 (9 years of data), of 914 495 Latinos recorded as US-born, non-US-born and without a country of birth recorded. We also described the state in which these data were collected. RESULTS: Country of birth was collected for 127 138 Latinos in 782 clinics in 22 states. Compared with those with a country of birth recorded, Latinos without this record were more often uninsured and less often preferred Spanish. While covariate adjusted prevalence of heart disease and risk factors were similar between the three groups, when results were disaggregated to five specific Latin countries (Mexico, Guatemala, Dominican Republic, Cuba, El Salvador), significant variation was observed, especially in diabetes, hypertension and hyperlipidaemia. CONCLUSIONS: In a multistate network, thousands of non-US-born, US-born and patients without a country of birth recorded had differing demographic characteristics, but clinical variation was not observed until data was disaggregated into specific country of origin. State policies that enhance the safety of immigrant populations may enhance the collection of health equity related data. Rigorous and effective health equity research using Latino country of birth information paired with longitudinal healthcare information found in EHRs might have significant potential for aiding clinical and public health practice, but it depends on increased, widespread and accurate availability of this information, co-occurring with other robust demographic and clinical data nativity.

Statin Eligibility and Prescribing Across Racial, Ethnic, and Language Groups over the 2013 ACC/AHA Guideline Change: a Retrospective Cohort Analysis from 2009 to 2018.

BACKGROUND: It is uncertain if the American College of Cardiology/American Heart Association (ACC/AHA) 2013 guidelines for the use of HMGCoA reductase inhibitors (statins) were associated with increased statin eligibility and prescribing across underserved groups. OBJECTIVE: To analyze, by race, ethnicity, and preferred language, patients with indications for and presence of a statin prescription before and after the guideline change. DESIGN: Retrospective cohort study. SETTING: Multistate community health center (CHC) network with linked electronic health records. PATIENTS: Low-income patients aged ≥ 50 with a primary care visit in 2009-2013 or 2014-2018. MAIN MEASURES: (1) Odds of each race/ethnicity/language group meeting statin eligibility via the National Cholesterol Education Program Adult Treatment Panel III Guidelines in 2009-2013 or the ACC/AHA guidelines in 2014-2018. (2) Among those eligible, odds of each group in each period with a statin prescription. KEY RESULTS: In 2009-2013 (n = 109,330), non-English-preferring Latino (OR = 1.10, 95% CI = 1.03, 1.17), White (OR = 1.41, 95% CI = 1.16, 1.72), and Black patients (OR = 1.25, 95% CI = 1.11, 1.42), were more likely than English-preferring non-Hispanic Whites to meet guideline criteria for statins. Non-English-preferring Black patients, when eligible, were no more likely than non-Hispanic Whites to have statin prescriptions (OR = 1.16, 95% CI = 0.88, 1.54). In 2014-2018 (n = 319,904), English-preferring Latino patients (OR = 1.02, 95% CI = 0.96-1.07) and non-English-preferring Black patients (OR = 1.08, 95% CI = 0.98, 1.19) had similar odds of statin prescription to English-preferring non-Hispanic White patients. English-preferring Black patients were less likely (OR = 0.95, 95% CI = 0.91-0.99) to have a prescription than English-preferring non-Hispanic Whites. CONCLUSION: Across the 2013 ACC/AHA guideline change in CHCs serving low-income patients, non-English-preferring patients were consistently more likely to be eligible for and have been prescribed statins. English-preferring Latino and English-preferring Black patients experienced reduced prescribing, comparatively, after the guideline change. Further work should explore the contextual factors that may influence guideline effectiveness and care equity.

Racial, ethnic, and language differences in screening measures for statin therapy following a major guideline change.

Atherosclerotic cardiovascular disease (ASCVD) disproportionally affects racial and ethnic minority populations. Statin prescribing guidelines changed in 2013 to improve ASCVD prevention. It is unknown whether risk screening for statin eligibility differed across race and ethnicity over this guideline change. We examine racial/ethnic/language differences in screening measure prevalence for period-specific statin consideration using a retrospective cohort design and linked electronic health records from 635 community health centers in 24 U.S. states. Adults 50+ years, without known ASCVD, and ≥ 1 visit in 2009-2013 and/or 2014-2018 were included, grouped as: Asian, Latino, Black, or White further distinguished by language preference. Outcomes included screening measure prevalence for statin consideration, 2009-2013: low-density lipoprotein (LDL), 2014-2018: pooled cohort equation (PCE) components age, sex, race, systolic blood pressure, total cholesterol, high-density lipoprotein, smoking status. Among patients seen both periods, change in period-specific measure prevalence was assessed. Adjusting for sociodemographic and clinical factors, compared to English-preferring White patients, all other groups were more likely to have LDL documented (2009-2013, n = 195,061) and all PCE components documented (2014-2018, n = 344,504). Among patients seen in both periods (n = 128,621), all groups had lower odds of PCE components versus LDL documented in the measures' respective period; English-preferring Black adults experienced a greater decline compared to English-preferring White adults (OR 0.81; 95% CI: 0.72-0.91). Racial/ethnic/language disparities in documented screening measures that guide statin therapy for ASCVD prevention were unaffected by a major guideline change advising this practice. It is important to understand whether the newer guidelines have altered disparate prescribing and morbidity/mortality for this disease.