In Reply I agree with Dr Bourque and colleagues that current methods of risk adjustment are imperfect and that prolonging life-sustaining measures in patients with little or no likelihood of recovery can be harmful. In contrast, I disagree that the inclusion of 30-day mortality data in hospital value–based purchasing will result in underuse of hospice services. A recent survey of Medicare beneficiaries indicates an increase in hospice use at the time of death.1 Regardless of the specific effect, I am in agreement with Bourque et al that incentivizing specific outcomes has the potential to influence physician behavior and health care delivery.
Kupfer JM. Mortality Reporting and Pay-for-Performance Programs—Reply. JAMA. 2013;310(13):1405. doi:10.1001/jama.2013.277987