[Skip to Navigation]
Comment & Response
March 2016

Discretionary Interpretations of Accountable Care Organization Data—Reply

Author Affiliations
  • 1Department of Health Care Policy, Harvard Medical School, Boston, Massachusetts
JAMA Intern Med. 2016;176(3):412. doi:10.1001/jamainternmed.2015.8506

In Reply Dr Saver raises an important question regarding our study1 of the Medicare Pioneer Accountable Care Organization (ACO) program, which was associated with modest reductions in the use of low-value services in its first year. Specifically, how can we know whether a reduction in the use of low-value services reflects an overall improvement in value? Citing the RAND Health Insurance Experiment (HIE),2 he notes that increased patient cost sharing caused reductions in both high-value care and low-value care. Thus, evaluating an intervention by only measuring changes in the use of low-value services may provide an incomplete picture of overall value because use of high-value services may also change. While we agree with this general point, in the case of the Pioneer ACO program, the evidence suggests that ACOs disproportionately reduced the use of low-value services.

Add or change institution