Thursday, October 23, 2008

Share the savings

I was invited to be a panelist today at the MA Medical Society's annual leadership forum, entitled "Cost, Quality and Access, The Challenging Journey Continues." Secretary of Human Services Judy Ann Bigby provided an overview of the state's health care environment. This was followed by a national perspective from Susan Dentzer (picture above), Editor-in-Chief of Health Affairs, having joined that journal in May after ten years as the on-air health correspondent on PBS' NewsHour with Jim Lehrer. And then we heard from Elliott Fisher (also shown above), Director of the Dartmouth Institute for Health Care Policy and Clinical Practice.

There were a number of common themes between Susan and Elliott which were reminiscent of points I have heard before from other observers like Brent James. Much of this, though, is based on the actual work done by Elliott and colleagues at Dartmouth. Key points: Unnecessary variation in the delivery of health care and an inverse correlation between spending and clinical results. The "paradox of plenty", explained Elliott, suggests that those regions with more medical resources not only have higher costs, but worse technical quality; worse access to primary care; lower satisfaction with hospital care; worse communication among physicians; greater difficulty ensuring continuity; great perception of scarcity; and lower satisfaction by providers with their career.

Brent James had talked about "local medical mythology" as an explanation for the unexplained variation in medical treatment across regions. Elliott reinforced that but also tied it into the payment system by suggesting that the payment system ensures that existing capacity in a region will be fully utilized because physicians adapt to available resources with more referrals, more admissions, and more ICU stays. He advocated a reformulation of the current fee-for-service approach that would establish a target growth rate in overall health care spending and then share the savings between the insurers and the providers if the provider group achieved lower-than-expected growth in spending.

I'm sure I am not giving this all justice in this short summary, but I think you get the idea. Lots of food for thought. In particular, this shared savings approach offers an interesting alternative to the fully capitated type of plans that have recently been proposed in some forums.

Addendum: I found out they made a podcast of the panel discussion. Here it is.


Anonymous said...

Thanks for the summary of the MMS meeting, Paul. I like Fisher's idea of sharing savings if the growth of spending is below a target. But what does he think we should do if the actual growth rate in spending is greater than the target--just sigh or actually do something like cutting fees?

Anonymous said...

I'll ask him to reply directly, as I do not recall if he addressed that during his talk.

Anonymous said...

The shared savings models currently in place do not have a penalty for exceeding the target. There are two primary reasons for this "soft" approach. First, because patients are not required to lock-in to their primary provider (nor are they penalized by higher co-pays from seeking care outside the network), providers have limited tools to control out-of-system use. Second, a firm target or budget would require providers to take on financial risk.

We hope that a voluntary approach will encourage broader participation by providers. And because there is so much opportunity for savings we should be able to truly bend the curve. All that's needed is to close a few hospital beds!!

Anonymous said...

The above comment was from Elliott. The comment box closed before he could type his name!

Anonymous said...

Oh wait, actually this is his post, which he sent me by email because the other one didn't get through.

Our current thinking -- and the current approach of the Medicare demonstration project -- is to use a target that does not penalize groups for exceeding it. There are several reasons to favor this approach. First, because there is no lock-in of patients, or even an incentive for them to remain within the network, the physicians within the group have little control over where they seek care (although the evidence suggests that most care is provided "within" the group). Second, physician groups or physician-hospital networks are not widely prepared to manage risk, which a firm target would require. We believe that a soft-target, voluntary approach is more likely to attract broad participation by physicians.

There are likely to be several keys to success: having all payers involved so that the provider organization has a strong incentive to actually save (rather than reducing utilization in the patients under shared savings while increasing utilization in others!); a multiyear commitment that encourages strategic moves (converting acute care beds to nursing beds and allowing a few high-cost physicians to retire without being replaced); and, perhaps, linking the shared savings program to other innovations -- the medical home model, for example.

But we have lots to learn about how best to do this. Experimentation and evaluation will be important.
Elliott Fisher

Anonymous said...

Ever since training in Six Sigma and Lean Kaizen, I've tried to make standardization my mantra. I've been impressed, however, at how variations in patient expectations profoundly influence provider behavior. We shouldn't be surprised- eating patturns, exercise habits, demand for consumer goods and political views all vary among people, even across regions, so why wouldn't demand for health care vary similarly? Because response to illness is influenced by cultural factors, a component of the variation and many of the "medical myths" represent physicians and patients dancing to the same beat. People are unlikely to accept changes that push up against their expectations.