Medicare payments to oncologists for chemotherapeutic agents have been a source of considerable controversy for several years. The Medicare Modernization Act (MMA) changed reimbursement methods, effectively reducing the markups oncologists receive, especially for expensive agents. In this work we propose a regression discontinuity design to study the effect of this change on the likelihood of prescribing chemotherapy to all cancer patients and, if chemotherapy is prescribed, the expensiveness of the agents used, and the setting in which chemotherapy is administered. If treatment changes are found, we will also consider the impact on indicators of untoward outcomes. In prior work we studied a similar question using data from before the MMA to assess the effect on treatment of variation in chemotherapy reimbursement to the same physician over time as well as across carriers. We found that more highly reimbursed oncologists were not more likely to prescribe chemotherapy to patients with metastatic cancer, but were more likely to use expensive agents. The complex nature of the recent reimbursement changes -- large reductions in payments for drugs but a significant increase in administrative fees -- makes predicting the MMA's effect from prior work difficult. Understanding its impact on chemotherapy treatment patterns has important implications not only for Medicare spending but also for the welfare of beneficiaries with cancer.
