Determinants of sustained uncontrolled blood pressure in a national cohort of persons with diabetes

Background: Randomized clinical trials have demonstrated that strict blood pressure (BP) control in diabetes reduces cardiovascular morbidity and mortality. Previous observational studies have confirmed that hypertension is inadequately controlled in the general population of the United States. In this study we evaluated the prevalence and determinants of severe, sustained, uncontrolled hypertension in a national cohort of persons with diabetes. Methods: We identified 64,105 veterans from the national Veterans Administration diabetes registry for whom BP, survey, laboratory, and medication data were available. Using mean BP from three visits in fiscal year 2000, we determined the prevalence of sustained BP readings >= 160/100, >= 140/90, or >= 130/80 mm Hg. We determined predictors of the three thresholds using demographic variables, self-reported medical comorbidities, estimated glomerular filtration rate, and number of BP-lowering medications. Results: Over a mean interval of 131.0 days (+/- 81.4), we found that 6,347 (9.9%) of the 64,105 veterans with diabetes had mean BP >= 160/100 mm Hg. Similarly 25,924 (40.4%) had a mean BP >= 140/90 mm Hg, and 38,296 (59.7%) had a mean BP >= 130/80 mm Hg. Independent predictors of mean BP >= 160/100 mm Hg included age, ethnicity, education level, cardiovascular comorbidities, alcohol use, and number of BP-lowering medications. Conclusions: Administrative databases can be used to identify patients with sustained uncontrolled hypertension within health care systems. Our findings suggest important patient-level factors that can be targeted for quality improvement programs in diabetes.