and support for clinicians in this area. These challenges are not insurmountable and in several cases, implementing performance measurement would be the exact stimulus needed to overcome these challenges.

Currently providers are not at financial risk for poor-quality transitional care. Few penalties exist for poor performance. For example, the hospital attending physician receives additional payment on the day of discharge irrespective of how well prepared the patient is to resume self-care. Alternatively it could be argued that providers are financially rewarded for poor quality to the extent that this care leads to recidivism and additional billing opportunities. While performance measurement in general and pay-for-performance in particular could positively influence the alignment of financial incentives, there will likely be significant resistance from the health care industry in defense of the status quo.

There has been limited experience exploring what aspects of transitional care health plans, institutions, and clinicians can be held accountable. Existing Medicare Conditions of Participation articulate these responsibilities but these have not been strongly enforced. Accountability also raises unprecedented questions as to whether two institutions that have no formal or fiscal relationships can be held jointly accountable for a failed transition. It also raises questions pertaining to the definition of an episode of care. However, as was alluded to above, this is a case in which answers to key questions such as these would follow once progress is made towards promoting greater accountability by enacting performance measurement.

Pay-for-performance discussions ultimately lead to discussions regarding case-mix adjustment. Currently performance measures oriented towards outcomes may not be sophisticated enough to discern whether a poor-quality care transition experience was due to a failed “hand-off” or simply a matter of disease progression. Experience with case-mix adjustment has been limited. This situation may argue for preferentially relying on process-oriented and patient experience-oriented measures versus more outcome-oriented measures. It may also argue for a two-staged approach in which, initially, health systems or institutions are paid for doing certain tasks rather than being paid for how they performed. For example, payment for timely transfer of a discharge summary, followed by timely transfer of a discharge summary that meets certain criteria for content and accuracy.

To date, pay-for-performance activities have focused on a set of measurement items that could be easily audited using administrative data sources. Yet as detailed earlier, the leading performance measurement instruments do not fit this profile and it is unlikely that such measures are possible given the current content of administrative records. For example, measures such as ACOVE or medication reconciliation rely on chart review that is often impractical for most health care systems to produce in a reliable and timely fashion (Wenger and Young, 2003). However, the emer-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement