Prostate cancer in the medicare shared savings program: are Accountable Care Organizations associated with reduced expenditures for men with prostate cancer?

Objectives: To assess whether Medicare expenditures for men with incident prostate cancer, treated in Accountable Care Organizations (ACOs) differ from those of men treated in non-ACOs.
Methods: Using the 20% Medicare sample, total charges for 1 year following an initial diagnosis of prostate cancer were abstracted from Medicare claims. Prostate cancer expenditures were calculated by subtracting total charges from the year prior to diagnosis. Propensity score weighting was used to balance baseline characteristics of men treated in ACOs and non-ACOs, and between treatment modalities (radiation, prostatectomy, and expectant management). A propensity score weighted regression model was then used to estimate mean expenditures for men with prostate cancer treated in ACOs and non ACOs and to test the association between ACO status and costs.
Results: We identified 3297 men treated in ACOs for localized prostate cancer versus 24,088 in the non-ACO cohort. The weighted total charges for each treatment modality were $32,358 (radiation), $27,662 (prostatectomy), and $11,134 (expectant management). In our propensity score weighted regression model, the association between charges and ACO status was not significant, nor was the interaction between treatment type and costs. This was true both overall, and in a stratified analysis by treatment type.
Conclusions: There was no significant difference in Medicare spending on prostate cancer care based on provider ACO affiliation, regardless of treatment type. Although the effects of ACOs on clinical care are complex, this study adds to a growing body of evidence suggesting that ACOs fail to achieve significantly lower charges in certain clinical settings.