Stuart Guterman, of The Commonwealth Fund, say(s):
May 12, 2009

To Marianne Udow-Phillips of Center for Healthcare Research and Trans:

Thanks for your comments on the column. We agree that health care will become better and more efficient only when providers work together and take joint responsibility for the welfare of their patients. Bundled payment is meant to encourage that type of change in how health care is provided, but we agree that payment changes must be combined with the development of strategies to enable physicians, hospitals, and other providers to work together better. We have been monitoring efforts around the country to develop those types of strategies, and appreciate your suggestions.

Stuart Guterman, of The Commonwealth Fund, say(s):
May 12, 2009

To Edward Gamache of Deckerville Community Hospital:

Thanks for your comments on the column. We agree that, while it is crucial to align payment incentives with more efficient use of health care resources, we must at the same time ensure that the quality of care and the outcomes produced by that care are maintained or improved. We currently are engaged in an initiative to reduce hospital readmissions in three states, and we certainly will be monitoring outcomes to make sure that both better care and slower cost growth result from those efforts.

notmd, of montefiore medical center, say(s):
April 30, 2009

When we analyzed our readmissions, close to 40% of the readmits were from nursing homes. This would appear to be an appropriate next step in providing care to the patient however the high readmission rate does not validate this conclusion. If you spoke to the nursing homes they would state that they are not reimbursed at a level to support acute care. If the bundling reimbursement given to hospitals does not compensate the nursing homes appropriately ,the relationship between post-acute providers and acute providers will become hostile and the patient will be caught in the middle.

Randall Williams, of Pharos Innovations, say(s):
April 29, 2009

I wholeheartedly agree with the observations and analysis presented by the authors. However, we know that the vast majority of "avoidable" hospitalizations in the Medicare population occur in populations with a short list of chronic conditions, such as heart failure, COPD, depression, frailty. Unfortunately, a payment reform model that time limits incentive alignment between acute care settings and outpatient providers runs a significant risk of underpperforming the improvement opportunity. By taking away the incentive after a defined period of time patients will be exposed to limited ongoing services that have otherwise proven to reduce hospitalization rates dramatically, such as care coordination/ case management, and remote patient monitoring through telehealth technologies. Thirty-day bundling is simply too short...56% of all Medicare discharges are readmitted in 12 months! This number rises to 56% in 6 months for those with CHF and other chronic medical conditions.

Marianne Udow-Phillips, of Center for Healthcare Research and Trans, say(s):
April 29, 2009

The concept of bundled payments is terrific and works very well in integrated health systems - especially academic medical centers like Geisinger or a place like the University of Michigan, here in Michigan. The approach is much more technically challenging, however, when applied in community-based settings. In that regard, there need to be mechanisms to create virtual networks of physicians who work together to improve quality. Blue Cross and Blue Shield of Michigan is engaged in some very creative approaches along those lines in their physician group incentive program. Adding these strategies to the recommendations included here are important because most people in this country still get their care through community-based providers that are not yet organized into integrated delivery systems.

Edward Gamache, of Deckerville Community Hospital, say(s):
April 29, 2009

Before declaring victory on the “health cost war” you may want to look at the VA Healthcare system readmission rates. They are not reimbursed on a fee for service system that would match the alleged negative private sector incentive system. You may find that the characteristics of readmissions are tied more to adopted practice patterns developed in university-based medical schools and not payment incentives as described in this article.

It is also important to develop cohort measures as you implement such programs, i.e. mortality rates, to monitor the creation of unintended consequences of such global payment systems. In looking at ambulatory sensitive conditions I have seen demographic patterns of low mortality in high admission demographic areas and higher mortality in low admission areas. The last thing we need is a 5-year readmission project followed by a mortality reduction project.