Within-practice Medicare spillovers to non-Medicare patients: Evidence from the Oncology Care Model

Abstract

ABSTRACT Objectives Little is known about whether the Oncology Care Model (OCM) affected care for non-Medicare patients. Study Design The quasi-experimental design here compares participant providers with non-participant providers during the first year of the program compared with the prior year. Methods Relative different levels of office visits and total costs across four cancers – breast, lung, colon, and prostate cancer – are measured for patients with Medicare Advantage, private insurance, and Medicaid and compared with the Medicare population. Results OCM participation is associated with 11.6% fewer Medicare breast cancer visits (estimate = −0.11, 95% CI: −0.20 to −0.01, p = 0.03) and 9.3% fewer Medicaid breast cancer visits (estimate = −0.10, 95% CI: −0.21 to 0.02, p = 0.09). Practices seeing more Medicare patients drive this result; every 10 percentage points more Medicare patients seen by OCM providers correlate with 6.1% fewer Medicaid visits (estimate = −0.66, 95% CI: −1.39 to 0.06, p = 0.07). Some patterns like lower costs did not extend to non-Medicare patients while others absent in Medicare were measured in non-Medicare patients. Conclusions This analysis suggests that care patterns are different under the OCM for both Medicare and non-Medicare lives, but that these differences are modest and inconsistent.