A Decomposition Method to Assess the Contributions of Geographic and Nongeographic Factors to White-Black Disparities in Health Care.

Black-white gaps in high-quality hospital use are documented, but the relative contributions of various factors are unclear. The objective of this study was to quantify the contributions of differences in geographic and nongeographic factors to the gap, using decomposition methods and data for coronary heart disease.
We identified white and black fee-for-service beneficiaries aged 65 or older who were hospitalized for acute myocardial infarction (AMI) or coronary artery bypass grafting (CABG) surgery during 2009-2011. We categorized hospitals with AMI mortality rates in the lowest quintile as high-quality hospitals. We first decomposed the white-black gap in high-quality hospital use into a component due to racial differences in region of residence and a within-region component. We then decomposed the within-region differences into contributions due to racial differences in geographic proximity to high-quality hospitals and due to nongeographic factors.
The white-black gap in high-quality hospital use was smaller for AMI than for CABG (1.7 percentage points vs. 7.5 percentage points). For AMI, region of residence contributed more to the gap than within-region differences (1.0 percentage point vs. 0.6 percentage points), while for CABG, within-region differences prevailed (2.0 percentage points vs. 5.4 percentage points). For both conditions, the within-region white-black difference in high-quality hospital use was mainly driven by nongeographic factors.
Decomposition methods are a useful tool in quantifying the contributions of various factors to the white-black gap in high-quality hospital use and could inform local policy aimed at reducing disparities in hospital quality.