Tuesday, March 2, 2010

Six Reasons to Think Twice About Hospital Payment Bundling and the Role of Population-Based Care/Disease Management

The peripatetic Disease Management Care Blog decamped to the back of the Philadelphia Marriott Conference 4th Floor Room Franklin 9-10 and sat in on the final event of the Disease Management Colloquium today. This last session was a wide ranging panel discussion that addressed a number of hot reform topics, including the ingredients that comprise state-of-the-art primary care, the role of Washington DC in the regulation of health insurance premiums, physician liability insurance reform and provider payment systems. It was all great stuff, but it was the talk about money that really made the DMCB’s ears perk up, especially because it dealt with a topic that has largely gone unexamined in this blog: ‘bundled’ hospital payments.

Readers may recall this notion arose largely in response to the problem of an average 20% readmission rate for Medicare beneficiaries who have been discharged from a hospital. Since hospitals can financially benefit from those readmissions, the concept of ‘bundling’ was invented. The good news is that bundling pays for a hospitalization and for all the necessary post-discharge care for a period of time (for example, 30, 45 or even 60 days). The bad news is that the bundling doesn’t pay for a readmission if that becomes necessary during that same period of time. Fans of bundled payments like it because the up-front extra payment should incent hospitals to dedicate resources to helping reduce avoidable readmissions. That's why bundling is one ingredient in the Democratic health reform proposals.

The details of bundling are important. The DMCB points out that the fair calculation of the amount of the bundled payment should include 1) the cost of the original admission, 2) the cost of the extra hospital resources (for example, case management and home monitoring), 3) the post-discharge provider care (such as the physician appointments), 4) some reasonable profit margin and, 6) last but not least, the cost of the unavoidable readmissions. The DMCB suspects a lowest a readmission rate can go to is around 8% of hospital discharges (if anyone knows of a lower rate, please share)

With that as background, readers may think this is a grand idea and another reason to support the Democratic vision for health reform. That may be true, but until the Colloquium, the DMCB hadn’t heard the arguments against the idea, which are outlined below. They’re not only worth consideration, but they're a handy way to upend any smug friends, family or colleagues who believe that Mr. Obama’s proposals represent an unassailable intellectual achievement with no unintended consequences:

1. Most hospitals just don’t have the necessary data management infrastructure: building or buying the post-discharge care in its myriad forms, managing the underlying risk transfer, understanding the co-morbidities of the patients and appropriately configuring all the interventions in a complex ecosystem of outpatient services is a very information-intense exercise. This is simply out of reach for the average hospital Board of Trustees, CEO, Director of Nursing and Chief of Staff.

2. Most hospitals also don't have the personnel or the expertise: even if they understood what needed to be done, the how behind creating and maintaining care management programs (job descriptions, work flows, policies and procedures etc) and building provider alliances would be very daunting.

3. This presupposes that the original admission was necessary in the first place: there is enormous regional variation in the decision to admit to a hospital. Bundled payments would simply enable more of the same shenanigans.

4. This could easily fail: there is a very real possibility that abundant optimism combined with budgetary shortfalls could lead to a public payer ‘perfect storm” that squeezes bundled payments and leaves vulnerable hospitals in the lurch. A potentially good idea could be set back years if it had to be repealed.

5. Actuarial neutrality: This was only brought up indirectly in by the Panel, but the DMCB thinks that the sausage making of bundled payments will end up somewhere between covering only the necessary services and nothing more on one end and covering what is happening anyway on the other end. Since this may ultimately just be a reshuffling of the money, the panel pointed out that post-discharge care coordination works well enough as a stand-alone initiative (for example, here, here and here). Why not just pay for it and be done with it?

6. Bad behavior
: This wasn’t brought up by the panelists, but the DMCB wonders if hospitals could end up with a financial incentive to deny inpatient care for patients that genuinely need to be readmitted to the hospital. After all, bundling sure looks like a form of capitation and we know how the physicians responded when they were forced to accept bad incentives. Could a hospital send someone home from the ER with advice to take two aspirins and call in the morning?

The DMCB also notes that much of the downsides described above can be amply mitigated by population-based care management. In fact, any literature search on post-discharge care coordination is likely to yield multiple articles that include the term 'disease management.' The DMCB says why not? Once again, non-physician professionals to the rescue with remote or in-person tailored interventions that can help persons and their family become experts in self-care. 'Nuff said.

The good news is that that the Democratic Senate bill appears to either 'pilot' or 'demo' the bundled payment concept. The DMCB hopes, after listening to this panel, that if health reform passes, that these six cautions and the role of population-based care management are closely examined before the HHS Secretary decides to implement it nationwide.

1 comment:

Bruce Pyenson said...

Thanks, Jaan, for your terrific insights.

I'd add another issue with bundled payments -- they can create an incentive to increase admissions. This was an issue with the earlier, aborted, "Provider Partnership" CMS Demo of the late 1990s. Under that demo, the bundle would have included physician payments plus the DRG payments. The idea was to create a common hospital-physician per-patient case rate, so the physicians could be paid an incentive to reduce LOS and to reduce hospital costs. However, some hospital organizations realized bundled payments could be used to incent physicians to increase admissions.

My preference for Accountable Care Organizations is to create regional benchmarks (that decrease over time) for statistics like Ambulatory Sensitive Admission, Readmissions, and Preference Sensitive Admissions.


Bruce Pyenson, FSA, MAAA
Milliman, Inc.