The Reform Conundrum | [node:field-byline] | Healthcare Blogs Skip to content Skip to navigation

The Reform Conundrum

July 24, 2009
by richard
| Reprints
Not quite two months ago, Atul Guwande published an article in The New Yorker magazine that has already attained something close to the status of legend.

Its findings have been cited by President Obama who reportedly made it required reading for his staff. It is cited by congressional leaders and it seems to come up at just about every healthcare conference I’ve attended over the last two months.

Why then do its main lessons seem to be, for the most part, absent from any of the current legislative proposals? The main finding of the article was that two towns less than 100 miles apart, with similar populations, can and do have radically different healthcare spending patterns, chiefly, it seems, because doctors make decisions based partially on economic factors, not only medical ones.

Many physicians, after all, are small business proprietors and they are rational economic agents. Hospitals too are rationale economic creatures. Provides will do, in other words, what they are paid to do. The most prevalent payment model, the fee for service model, dictates that providers are paid on the basis of volume and it also dictates that things that are reimbursed more favorably, like major procedures for example, will be done more frequently than things that are not reimbursed very well, like preventive services for example.

So even if we increase the number of people covered by insurance (and we should), and even if we find a way to pay for it via a tax, a fee to insurers, a reduction in allowable deductions, etc., covering more lives under the current fee for service system will only escalate the rate of healthcare spending, not curtail it. Finding a way to pay for coverage and finding a way to keep costs from continuing to rise are related, but different problems.

Systems like Geisinger Health and Mayo Clinic are frequently cited as examples of institutions who have found a way to provide high quality care in an efficient manner. What lessons from these organizations have filtered into the reform legislation?
Have these organizations become more efficient by capping readmissions at the 25th percentile? No.
Have these organizations become successful by relying on an external advisory board – a GeisingerPAC - to set payment rates? No.
Have they created Clinical Effectiveness Boards to decide what will be paid for and what will not? Although I’m sure these organizations are successful in part because they adhere to the best principles of evidence based medicine, I don’t believe they have set up boards to enforce this.


What these organizations have done instead, is they have found a way to remove or at least lessen the perverse economic incentives to generate waste. That is, they have removed, in part, the incentives to generate more visits, more procedures, more hospital days simply because they are paid for.

Some attempts to remove perverse economic incentives have made it into current legislative proposals in the form of the “accountable care organization” which would require groups of providers to come together to receive a unified payment for providing services to a specific group of Medicare beneficiaries. Also, the “medical home” model which requires a greater coordination of care across the continuum has been proposed as another type of pilot.

These two models could have real potential, but the current proposals leave such things as defining the providers who are “accountable” intentionally vague and the both models are proposed as pilots and would be voluntary in any event.

Real reform will mean taking away or lessening incentives to simply do more of what gets paid for. It will require a movement away from paying one visit, one procedure, one stitch at a time.

This is the true reform conundrum: No one in his or her right mind would design a health care delivery system as today’s “pay-by-the-stitch” model, and yet no one is naïve enough to think we can dismantle it – at least not overnight.

But a proposal that purports to achieve healthcare reform should contain some plan, or at least a framework, that shows us how to get from here (the unsustainable pay-by-the-stitch model) to there (a model in which we pay for health).

Sadly, no where is such a plan to be found.



Thanks for your comment.

I am not familiar with the Archimedes model, but I intend to learn more about it thanks to your excellent post. It sounds like it is designed to take into account the big picture, and to look at benefits as well as costs, which is what public policy analysis must do.

The CBO scoring mechanism is a terrible way to decide public policy, and it doesn't purport to do so, but the discussions of late seem to make this a tacit assumption. It's worth rembering that our Federal Government, in addition to having a regulatory and policy setting role, is also the largest payor in health care. The CBO considers no savings "real" unless they are savings dirctly to the "payor." Perhaps the payors of the remaining 50% of healthcare costs, the employers, should do thier own "scoring."

In any event, if any one stakeholder were to announce, "I am unconcerened with savings, unless the beneftis accrue directly to me," (Imagine the AMA or AHA making such a statement), then we would immediately dismiss the approach as "self interested."
Of couse the government (we taxpayers) must maintain self interest in this area, but healthcare policy needs to examine much more, and perhaps the Archimedes model would be a useful tool.

Are you familiar with the approaches Kaiser Permanente is using, and their investment in what has become the Archimedes model? I identified the same problem you're describing in my post "Obama Asks Bloggers For Help With Health Care System Overhaul." Actually, Dr David Eddy does propose a plan. And the model includes the care decisions driven by the economics and behavioral factors you and Atul Gawande appropriately cite. And, of course, were cited by Jack Wennberg 20 years ago, with his small area variation work and the Dartmouth Atlas.

I'm interested in your thoughts on the role of such models. There are several and they've been used for many years.