The Paradoxical Association of Disaggregated Data Collection With Diabetes Control Among Latino Patients

BACKGROUND: Data disaggregation in Latino patients has been called for by national organizations. However, analyses of the association between the collection of this data point, especially the place of birth, and common disease-specific factors, including disease severity, are lacking. OBJECTIVE: To examine the relationship between the collection of place of birth data and diabetes control in a national network of community health centers (CHCs). PARTICIPANTS: Hispanic/Latino adult patients, aged 18 to 79 years, with diabetes. MEASURES: We described the following groups by demographic, clinical, and neighborhood factors, and compared odds of hemoglobin a1c (HbA1c) always <7% or ever >9%, stratified by preferred language: (1) patients whose clinics never collected place of birth; (2) those whose clinics did collect this information, but they personally did not have a country of birth in their record; and (3) those with country of birth documented. RESULTS: In our study population (n=81,107), Spanish-preferring Latinos with their place of birth recorded (HbA1c always <7: OR=0.75, 95% CI: 0.65-0.87; HbA1c ever >9: OR=1.68, 95% CI: 1.44-1.95) had worse HbA1c measures than Spanish-preferring Latino patients in clinics that did not collect country of birth. CONCLUSIONS: In a national CHC network, Spanish-preferring Latino patients with country of birth information in their records had less well-controlled HbA1cs than those who were served by clinics that did not collect these data. These surprising findings underscore the need to assess disaggregated data collection in Latino patients to better understand the data necessary for high-quality primary care in Latino communities.