Use of myelosuppressive chemotherapy and colony-stimulating factor (CSF) prophylaxis: A longitudinal assessment of practices participating in the Centers for Medicare and Medicaid Services (CMS) oncology care model (OCM).

Abstract

12 Background: The CMS OCM is a 5-year program (7/2016 – 6/2021) focused on improving oncology care by incentivizing practices via financial and performance accountability. Little is known about practices participating in the OCM and whether patterns of care have changed over time. Methods: We used a retrospective cohort with structured elements from national electronic health record dataset provided by Flatiron Health (1/2012-12/2017). Data were de-identified to prevent patient and provider re-identification. Study population included adults with breast, colorectal, lung, or ovarian cancer or non-Hodgkin lymphoma who received chemotherapy with intermediate/high-risk for febrile neutropenia (FN). Patient characteristics, treatment, and supportive care within OCM and non-OCM practices were evaluated overall and by calendar year. Results: Study population included 42,699 patients at 53 OCM practices, and 25,613 patients at 149 non-OCM practices. OCM practices were larger (mean: 806 vs. 172 patients); age (mean: 62 vs. 62 years), cancer type (breast cancer: 53% vs. 52%), chemotherapy (high FN-risk: 49% vs. 48%), and other characteristics were comparable between patients in OCM and non-OCM practices. Use of high (vs. intermediate) FN-risk chemotherapy in 2012, 2015, and 2017 was 48%, 50%, and 48% in OCM practices and 50%, 50%, and 48% in non-OCM practices. CSF prophylaxis use, while higher in OCM practices, decreased over time in both subgroups (OCM: 70% in 2012 to 65% in 2017; non-OCM: 63% in 2012 to 58% in 2017). Use of pegfilgrastim (vs. short-acting CSFs) was > 94% across years in OCM and non-OCM practices. Use of the on-body injector (vs. prefilled syringe) increased from 26% of pegfilgrastim use in 2015 to 73% in 2017 in OCM practices and from 16% to 63% in non-OCM practices. Conclusions: Although OCM practices are larger and more commonly use CSF prophylaxis than non-OCM practices, trends in use of high FN-risk chemotherapy and declining prophylactic support appear to be similar between subgroups. Additional research is needed to evaluate whether such changes impacted health outcomes.