Objective: Identifying modifiable covariables that reduce demographic disparities in controlling type 2 diabetes could inform efforts to improve health equity.
Research design and methods: This retrospective study utilized electronic health record data on 22,285 adults with type 2 diabetes seen at 110 outpatient clinics in the Southeast U.S. from 2004-2008. Demographic differences in diabetes control and modifiable covariables which reduce those disparities were quantified using descriptive and logistic regression analysis.
Results: Patients were 55.8 +/- 14.6 (SD) years old, 57.5% women, 61.0% white: 39.0% black and had baseline body mass index 34. +/- .3 kg/ m2 and HbA1c 7.61 +/- 1.9%. The percentage with HbAlc <7% was higher in Whites than blacks (55.6% vs. 44.7%, P < .0001) and rose with age in all patients from 45.3% at <50, to 50.0% at 50-64, and 59.6% at > or =65 years, P < .001. white vs. black race (odds ratio [OR] 1.59, 95% confidence interval [CI] 1.51-1.68) and age/ 10 years (OR 1.20/10 years, 95% CI 1.17-1.22) were predictors of HbAlc <7% in univariable logistic regression. In multivariable analysis, three modifiable covariables (initial HbAlc, therapeutic inertia, visit frequency) accounted for 47.9% of variance in diabetes control. When accounting for these modifiable covariables, the independent impact of race/ethnicity (OR 1.21, 95% CI 1.13-1.30) and age (OR 1.13, 95% Cl 1.11-1.16) on HbA1c control declined.
Conclusions: Race and age-related difference in diabetes control declined significantly when modifiable covariates were considered. Greater attention to early diagnosis and treatment, ensuring regular healthcare visits and overcoming therapeutic inertia could improve diabetes control and health equity.