Measurement of glycemic control in diabetic patients—an evaluation of risk adjustment using a primary care registry

Abstract

Background The ability to measure hemoglobin A1c in populations under the care of physicians or physician groups has become increasingly important with the advance of system-based interventions that can affect this outcome and with payment linked to levels of diabetes control. Objective To evaluate the effect of patient and system factors on the rate of glycemic control as measured by a hemoglobin A1c of <7% using a diabetes mellitus registry from osteopathic training programs in internal medicine and family medicine. Design Observational study. Participants A cohort of 4715 diabetes cases abstracted from the medical records of 127 residency programs nationally between 2003 and 2008. Measurements and main results: Associations between glycemic control and age, gender, medications used, insurance type, race/ethnicity, levels of appointment adherence, hypertension, and presence of evidence of nephropathy were evaluated. In bivariate and multivariate analysis, age, medication type, insurance type, level of appointment adherence, and presence of evidence of nephropathy had a statistically significant association with hemoglobin A1c control. Age was associated with increased levels of control, whereas use of insulin, insurance other than Medicare, non-Caucasian race, missing more than 20% of office visits, and the presence of microalbuminuria or nephropathy were associated with decreased glycemic control. Conclusions System and patient factors not under the control of the physician have a significant effect on levels of glycemic control. To evaluate physician or practice performance, methods of controlling for these factors need to be developed and implemented.