Diabetes screening and monitoring among older mexican-origin populations in the US

Context: The prevalence of both diagnosed and undiagnosed diabetes is nearly twice as high among Latino adults than among non-Latino white adults. However, limited evidence exists on whether patients are monitored for diabetes adequately across all Latino subgroups, especially individuals of Mexican-origin.

Objective: To examine the rate of diabetes screening and monitoring among Latino individuals as compared to non-Latino white individuals, and how those rates vary based on estimated percentages of Mexican-origin population in the local neighborhood (a proxy for Latino Mexican subgroup).

Study Design: Retrospective observational study.

Setting or Dataset: OCHIN electronic health record data linked with neighborhood-level Latino subgroup data obtained from the American Community Survey.

Population studied: Adults aged ≥50 years who have had ≥1 ambulatory visit in study clinics within the study period (2012-2017), and who had census tract-level geocoded addresses available in the EHR (N=105,257).

Intervention/Instrument: Binary indicator of Latino vs. non-Hispanic white. Additionally, we used the Jenks natural breaks method to produce three categories of population percent Mexican-origin by neighborhood: low (0-22%), medium (22-55.7%) and high (55.7-98%).

Outcome Measures: Odds of HbA1c monitoring and screening tests. Logistic regression models were adjusted for patient-level sociodemographic covariates and clustered by primary clinic.

Results: Latino patients monitored for diabetes had 26% higher odds of receiving an HbA1c test than non-Latino white patients with diabetes (OR=1.26, 95% CI=1.10-1.45). Compared to non-Latino white patients with diabetes, Latino patients with diabetes living in a low percent Mexican-origin neighborhood had 40% higher odds of receiving HbA1c monitoring, however, those living in neighborhoods with higher percent Mexican-origin had similar odds of monitoring compared with non-Latino white patients.

Conclusions: These findings suggest the varying health behavior and healthcare utilization aspects by the Latino population in the ethnic enclaves. They also have implications for differences in primary care evaluation and management, where heterogeneity of service delivery has already been demonstrated.