A recent Wall Street Journal editorial strongly challenged the notion that there is enormous waste in American health care. In the article the editors acknowledge that dramatic variation in health care spending exists across the country–but point out that the precise reason for that variation remains uncertain. They also note that much of the data about regional variation comes from the Dartmouth Atlas–and that work, they point out, is limited in that it only examines Medicare data. And they cite work from Richard Cooper at the Wharton School that directly challenges some of the Dartmouth Atlas conclusions–essentially arguing that the Dartmouth observed regional variation is actually simply an artifact of Medicare. They conclude that “Dr. Cooper’s assault on the Dartmouth Atlas is controversial but compelling. He argues that the less-is-more theory is based on the flawed premise that when a region’s outcomes did not improve as spending increased, the difference is simply classified as ‘waste’ – even if it isn’t.”
And yesterday, at a Brookings Institution-sponsored conference on comparative effectiveness, Peter Orszag, mentioned prominently in the WSJ article, initiated the beginnings of a response to the editorial–noting that he did not believe that the Wharton research was correct.
But the WSJ editorial and the Wharton researchers probably do have a point about the observed variation–the observed variation is indeed probably different in Medicare than it is in the commercial data. In fact, some of the analysis conducted under the PROMETHEUS payment approach comes to the same conclusion. Unfortunately, beyond that point, the editorial seems to miss the point.
The Prometheus analysis is based in part on private data using a national commercially insured population database. It examines both total costs as well as costs for defective care–essentially that waste the WSJ editors don’t seem to see. The Prometheus analysis using the commercial data shows that there is significant variation in costs, but they do not, in fact, see the same variation pattern in both commercial and Medicare. Why would that be? It is likely the same “squeeze the balloon” effect we see in almost all payment discussions (it’s one reason why an episode-based approach to payment reform that leaves “nowhere to hide” is important.) In those areas where Medicare reimbursement is very high–say, RWJF’s home state of NJ–there isn’t as much pressure to make up income gains from the negotiated private side. On the other hand, in Minnesota where the Medicare reimbursements are relatively low–there is great pressure to recoup in the negotiated private side.
The Prometheus folks do see, in some respects, higher private reimbursements correlating with relatively higher quality–and vice versa–which is, in many cases, different than or even the opposite of the Medicare regional variation findings. Francois de Brantes of Prometheus, however, points out that both the Dartmouth investigators and the Wharton researchers are looking at total cost–and he notes “that’s not what matters most–rather it’s what’s underneath those total costs that is so important–the ‘total cost’ number really still masks the waste.”
The Prometheus analysis is one way to look at what’s underneath–identifying potentially avoidable complications or defective care (it’s not entirely correct to say, “waste”–but it’s close). The commercial/Medicare variation correlation is significantly less important than the amount practices spend on that waste–or defects–and that defective care rate is huge across the board–about 50 % of chronic care cost in the Prometheus approach.
So, care provided under both commercial plans and Medicare, has tremendous room to reduce defects and waste–that’s the important point. It might not be quite as tidy as the Medicare-only version of the regional variation story–but it leads to the same conclusions–given the enormous hit to the federal budget for Medicare and Medicaid –together with the challenge of covering the uninsured–we can no longer afford to overlook opportunities to maximize health care value. And we do that by reducing the extremely high rate of defective care. Overall health reform does likely hinge on efforts to capture that value–and within that drive toward high value, payment reform is critical. And this conclusion reinforces the point that if the payment reforms don’t uniformly promote value–across the continuum of care–and across payers, they won’t drive value. Instead the costs will simply disappear from one point and reappear somewhere else. That’s really Dr. Cooper’s insight on the Dartmouth data.
But payment isn’t sufficient either. The editorial also highlights the need for regional efforts to support innovation–in fact it seems to conclude that some folks are actually trying to stamp out the regional innovation capabilities of the status quo. It’s true that regional innovation is important, but they’re only partially right on this point as well. We actually do need all kinds of ongoing local innovation if we are ever to experience high-value care–as opposed to our current wasteful, irrationally expensive, procedurally oriented, high volume care. But to promote that regional innovation–we don’t need the status quo approach–instead we need more and better publicly available information about the cost and quality of care so consumers and health professionals can make the health care market mechanisms work. Health professionals must have the resources to improve once they receive the market signals to improve. Consumers must have a powerful role in driving toward value. And, as noted, we need payment, built off these basic ingredients, that rewards efforts to reduce wasteful, defective care in the pursuit of high-value outcomes.
Michael W. Painter, J.D., M.D., is a physician, attorney,
health care policy advocate, and 2003-2004 Robert Wood Johnson Health
Policy Fellow. He is currently senior program officer and a senior member of the
RWJF Quality/Equality Team.