Interpreting Glycated Hemoglobin (HbA1c), Blood Pressure, and Cholesterol Trajectories in Quality Improvement or Disparities Research

Outcome-based quality measures for persons with type 2 diabetes generally assess the proportion of patient populations who have levels outside of conventional thresholds—e.g., below 7.0 percent for HbA1c, below 130/80 for blood pressure, or below 100 mg/dL for low density lipoprotein (LDL). Intervention studies often assess pre-post changes over time for relatively short periods. Comparatively few intervention or quality improvement studies are assessed by analyzing clinical data such as HbA1c levels, blood pressure, or cholesterol levels over several years.

This analysis explores changes in the trajectories of clinical values as a method for assessing quality, intervention effects, or disparities over longer a period. Fewer make comparisons to other similar longitudinal datasets. These data may inform or serve as a benchmark for other studies of disparities in chronic illness care, or for quality improvement initiatives. The population comprises entire panels of patients with type 2 diabetes served by five inner-city community health centers (CHCs). This racially and ethnically diverse group of over 3,500 patients, about half of whom were African American, received diabetes care at the CHCs over multiple years. We collected data from medical records for a 5 to 7 year period and performed multivariate regression analyses to estimate trajectories for HbA1c, blood pressure, and total cholesterol, adjusting for demographic characteristics and participation in a diabetes quality improvement program.

Both HbA1c levels and cholesterol levels tended to improve markedly soon after participation began in either enhanced care or usual care regimens; blood pressure improved marginally. HbA1c levels worsened slightly over subsequent years; cholesterol levels and blood remained constant. Age tended to have a negative effect on HbA1c and a positive effect on blood pressure and total cholesterol. African Americans had the highest levels of HbA1c levels and blood pressure over the period; Asians had the lowest. Asians and Hispanics had the highest levels of total cholesterol over the period. The shape of the HbA1c trajectories for this population resemble that of the study population in the United Kingdom Prospective Diabetes Study, even though the characteristics of patients served in American health centers differs, and even though the respective study periods were over two decades apart. Within the health center population, differences among racial/ethnic groups in the three clinical indicators may represent important differences in physiological or social factors.