Abstract
Introduction: The association between health care payor source and HbA1c in patients with type 2 diabetes (T2D) treated at federally qualified health centers (FQHCs) has been evaluated only in cross-section. FQHCs provide an opportunity to compare the association of glycemic control between populations with or without public or private insurance followed over time. Methods: In this observational study of primary care data, we compared HbA1c values by antecedent insurance status among adults with T2D, before and after adjustment for age, sex, and primary spoken language. Results: Among 9,848 patients evaluated, mean HbA1c differed (P<.05 for each pair except Medicaid vs self-pay) by payor source with Medicare patients having the lowest (7.44%), Medicaid the highest (8.01%), and privately insured (7.84%) and self-pay patients (7.99%) intermediate. Greater age and being female (P<.001) were associated with lower HbA1c. Language was independently associated with HbA1c (P<.001), with Spanish-speaking patients having the highest HbA1c (7.97%), English-speakers intermediate (7.92%), and others lowest (7.46%) (Spanish vs English not significant, others P<.05). Age, sex, language, clinic location, time between first and last visits, and number of visits were each associated with HbA1c, but in a fully adjusted model, significant differences by payor group persisted. Conclusion: In this study of FQHCs patient data, HbA1c measured after nearly 2 years of payor source observation differed by payor. Measured covariates did not fully explain the differences. Uninsured patients and those with Medicaid had worse HbA1c than those with Medicare or private insurance. These results may be used to inform another study with more available variables to determine what specific factors may mediate the association between payor source and HbA1c.