Influence of provider and health system factors on uptake and disparities in breast cancer gene expression profile testing.

Abstract

Listen

Translate article

250 Background: Gene expression profiles (GEPs) have been rapidly and widely adopted to guide chemotherapy decisions in early hormone receptor-positive (HR+) breast cancer, but not all eligible patients receive testing. Patient-level predictors of testing have been reported. However, the role of provider and insurance factors in GEP use has not been well described. Racial disparities in access to testing are also under-studied. Methods: We used state cancer registry data linked to claims from multiple insurance payers to identify 11,958 patients eligible for GEP testing between 2005 and 2012. For each patient, we characterized specialty and breast cancer volume for the provider deemed most likely to make the testing decision. We identified claims for the most commonly used GEP (OncotypeDX) and used adjusted incidence rate models to examine predictors of testing, stratified by nodal status (N0 versus N1). Results: Overall, 26% of eligible patients were tested with rates increasing over time. In multivariable analyses among N0 patients, black women were 12% less likely than clinically similar whites to be tested. Patients treated by a general surgeon were less likely to be tested than those treated by oncologic specialists, while those seen by middle-volume providers were significantly more likely to be tested than patients of the highest-volume providers. Black women with N1 disease were 27% less likely to be tested. Among N1 patients, testing was most common among patients of highest-volume providers and declined consistently with provider volume. Private insurance increasing the likelihood of testing for N0, but not N1 disease. Conclusions: Insurance status and provider characteristics appear to influence the likelihood of GEP testing for early breast cancer, but do not fully explain racial disparities in access to testing. Black women are less likely than whites to receive testing in either the node-negative or node-positive setting after adjustment for clinical factors. These date suggest a need to develop quality programs that enable uptake of appropriate breast cancer genomic tests, with a focus on elimination of disparities.