LAUSR

BACKGROUND Spending on cancer patients is substantial and has increased in recent years. Accountable Care Organizations (ACOs) are arguably the most important national experiment to control healthcare spending. How ACOs are managing patients with cancer at the end of life (EOL) is largely unknown. We conducted this study with the objective of determining whether becoming an ACO is associated with subsequent changes in EOL spending or utilization among patients with cancer. METHODS Using national Medicare claims from 2011-2015, we identified patients who died in 2012 (pre-ACO, n = 12,248) and 2015 (post-ACO, n = 12,248), assigning each decedent to a practice. ACOs were matched to non-ACOs within the same geographic region. We used a difference-in-differences model to examine changes in EOL spending and utilization associated with becoming an ACO in the Medicare Shared Savings Program for beneficiaries with cancer. RESULTS We found that the introduction of ACOs had no meaningful impact on overall EOL spending in cancer patients (change in overall spending in ACOs -$1,687 versus -$1,434 in non-ACOs, difference $253, 95% CI -$1,809 to $1,304, p = 0.75). We found no changes in total patient spending by cancer type examined or by spending categories, including cancer-specific categories, radiation oncology, chemotherapy, and hospice services. Finally, emergency department visits, inpatient hospitalization, intensive care unit admissions, radiation therapy, chemotherapy, and hospice use did not meaningfully differ between ACO and non-ACO patients. CONCLUSIONS The introduction of ACOs does not appear to have had any meaningful effect on EOL spending or utilization for patients with a cancer diagnosis.