Adherence to ASTRO Choosing Wisely Recommendations for IMRT after Lumpectomy Across Insurance Types

S. A. Patel,L. M. Halasz,J. McDougall,C. Fedorenko,Q. Sun,B. Goulart,J. A. Roth

DOI: https://doi.org/10.1016/j.ijrobp.2017.06.1583

2017-01-01

Abstract:In 2013, ASTRO issued choosing wisely recommendations (CWr) to not routinely use intensity modulated radiotherapy (IMRT) to deliver whole breast radiotherapy as part of breast conservation. A previous study demonstrated substantial variation in IMRT use across SEER registries among Medicare fee for service (FFS), but no study has evaluated adherence in Medicare advantage, commercial plans, or compared IMRT use across insurance types. In addition to the SEER-Medicare database, we used a novel linkage between the Washington Cancer Surveillance System (CSS) and claims from two large commercial insurance plans to explore IMRT use after lumpectomy around the period of the CWr. CSS records for patients with first primary diagnosis of Stage I/II breast cancer (2008-2015) were linked with Premera Blue Cross and Regence Blue Shield claims. SEER records (12 sites) for patients age ≥65 with first primary diagnosis of Stage I/II breast cancer (2007-2012) were linked with Medicare claims (2007-2013). Included cases had coverage for ≥12-months before/after diagnosis (or until death), lumpectomy 1 month before to 6 months after diagnosis, and radiation therapy within 6 months of lumpectomy. We calculated summary statistics and evaluated the association between insurance type and IMRT use following lumpectomy with generalized estimating equations (GEE) clustered on institution and adjusted for demographic and clinical characteristics. We identified 894 cases with Medicare FFS, 370 cases with Medicare Advantage, and 678 cases with commercial insurance (age <65). Overall, mean age was 73.4, 89.5% were white, and 50.6% were left-side cases. Table 1 shows the proportions receiving IMRT across insurance types. GEE analyses adjusted for age, stage, grade, race, laterality, and diagnosis year (2007-2012 vs. 2013-2015) demonstrated significantly less IMRT use in Medicare Advantage (OR=0.29, p<0.01) and commercial insurance (OR=0.21, p<0.01) vs. Medicare FFS. Left-side cases were strongly associated with IMRT use (OR=1.68, p<0.001). There was a non-significant trend towards IMRT use after the CWr vs. before (OR=1.42, p=0.25). This is the first study evaluating IMRT use across insurance types. We found that practice in Western Washington appears aligned with the ASTRO CWr overall, though there was significantly more use in Medicare FFS. For comparison, a prior analysis of use in Medicare FFS 2000-2005 showed proportions receiving IMRT as high as 23%. Across insurance types, the vast majority of IMRT use was in left-sided cases--likely reflecting reimbursement policies allowing for cardiac sparing purposes. Additional comparisons of IMRT use across insurance types should be undertaken in other regions to inform targeted policies to promote high-value cancer care.Abstract 2973; Table 1Insurance TypeOverall IMRT %Left Use%Right Use %Medicare FFS5.8%8.7%2.6%Medicare Advantage8.1%13%2.4%Commercial (Age <65)5.3%9.1%1.5%All Patients6.3%10.5%1.8% Open table in a new tab