Measuring racial disparities in the quality of ambulatory diabetes care

Abstract

Background: Improving the health of minority patients who have diabetes depends in part on improving quality and reducing disparities in ambulatory care. It has been difficult to measure these components at the level of actionable units.

Objective: To measure ambulatory care quality and racial disparities in diabetes care across groups of physicians who care for populations of ambulatory diabetes patients.

Research design: Prospective cohort analysis using administrative data.

Subjects: Using fee-for-service Medicare claims data from 2003 to 2005, we link patients to their principal ambulatory care physician. The patients are then linked to the hospital where their physicians work or have their patients admitted, creating physician-hospital networks.

Measures: Proportion of recommended diabetes testing received by black and nonblack diabetes patients.

Results: Blacks received 70% of recommended care compared with nonblacks who received 76.9% (P < 0.001). However, for black and nonblack patients, variation in the quality of care exceeds the racial gap in treatment. The network-specific performance rates for blacks and nonblacks were highly correlated (r = 0.67, P < 0.001), but 47% of blacks, versus 31% of nonblacks, received care from the third of networks with lowest quality. Physician-hospital networks with higher overall quality, or patients with higher socioeconomic status, were no less likely to exhibit black-white disparities.

Conclusions: It is possible to measure, benchmark, and monitor the quality of minority care at the level of networks responsible for ambulatory care. Consequently, it should be easier to provide patients with information on network performance and to design policies that improve the quality of minority-serving providers.

FIGURE 1

Ambulatory diabetes care for non-Black and Black patients across physician-hospital networks. Forty-five degree line represents equal rates for Blacks and non-Blacks. Solid circles indicate networks with statistically significant racial disparities (P < 0.05).

FIGURE 2

Racial disparities in ambulatory diabetes care across terciles of ambulatory care quality.

FIGURE 3

Delivery of ambulatory diabetes care for non-Black and Black patients at physician-hospital networks grouped by (A) teaching and (B) nonteaching networks.