• What does “all-payer” really mean?

    Since seeing Joe Newhouse’s suggestion of an all-payer health system and recalling Uwe Reinhardt’s advocacy and Maryland’s implementation of one, I’ve been trying to figure out what “all-payer” really means. What are the variations in all-payer systems? How do they work? To what extent do they reduce costs, enhance access, affect quality, etc.? In recent days, I’ve found answers to some of those questions.

    From a 2009 Health Affairs article by Robert Murray, it’s clear that one Maryland-style all-payer system is an administrative pricing one. Regulators pour over hospital cost reports and, based on them, set prices. But the implication of Newhouse’s brief mention of all-payer is that he preferred it to a single-payer system because it included price signals. So there must be more to it!

    A very old, 1992 Health Affairs paper by Paul Ginsburg and Ken Thorpe does reveal a little more. It’s about how to make the difficult transition from the current price-discriminatory system to an all-payer one, while maintaining price competition. There seem to be two key elements. One is that prices still vary across hospitals. So there’s a little price competition in that sense. However, a single hospital cannot vary its price for a given service across payers. That’s what all-payer means.

    Ginsburg and Thorpe also suggest the possibility of insurers negotiating discounts from an administratively set cap on a given service. That permits competition and is a big price signal, but doesn’t sound like it belongs in the all-payer category. It’s just price caps. Maybe I misunderstand their ideas.

    This brings me to Reinhardt’s 2009 Health Affairs blog post on the subject. (Aside: I overlooked this initially because Health Affairs blog posts don’t show up in database searches of the academic literature, or not prominently anyway, since they don’t get cited in other papers. This strikes me as a real problem for the research blog model that Health Affairs is promoting. I think they should consider bundling  a subset of their blog posts–maybe the ones with the most links, tweets, and comments–into occasional or periodic special issues.)

    For starters, one could use the diagnosis-related groups (DRGs) and resource-based relative value scale (RBRVS) now used by Medicare as the first-stage relative value scales, which could be refined over time on the basis of either cost or imputed value.

    If price competition among providers were desired, one would allow each individual provider to set their own monetary conversion factor for their relative value scale and compete on that simple, one-dimensional price indicator. …

    One could also, however, have these conversion factors negotiated between associations of providers and associations of insurers with a region (e.g. a state) and make them binding on all providers and insurers in the region, as is now done in some European countries — notably Germany — which operate all-payer systems within regions.

    Thus, to do away with the unwieldy and unseemly price discrimination now prevalent in American health care, a physician or a hospital would charge all insurance carriers or patients the same price for identical procedures. The system would work best if there were not a large number of uninsured people and if the public insurance programs — Medicare, Medicaid, and the Children’s Health Insurance Program, or CHIP — were part of the arrangement.

    Very clear! I get it now. I can see the dimensions in which an all-payer system can vary to build in price competition. In the rest of Reinhardt’s post he describes the various benefits of an all-payer system and how it can serve as a transition to bundled payments.

    One should not be too optimistic about what an all-payer system can do. It’s principally a system for purging price discrimination from the system, which would also eliminate cost shifting, to the extent it exists. Beyond that it would aid in the goal of price transparency, a necessary condition for a well- functioning market, and it would enormously simplify billing and reimbursement. All good stuff.

    But will all-payer bend the cost curve? I think one can’t be certain without pinning down the details. With prices that are administered in some fashion, the curve can be bent by fiat. Under a competitive model, the curve does whatever the market allows. As I like to say, “The market does what the market is.”

    Finally, since price discrimination currently exists between public and private payers–the former pay far less than the later, per admission or service–movement to an all-payer system means higher prices for Medicare and Medicaid. Isn’t that a problem for federal and state budgets? That’s got to be addressed before policymakers can really get behind this idea.

    Share
    Comments closed
     
    • I wonder if one could cobble together a test of this using Medicaid HMO’s? It’s sort of a crude analogy; but in some Medicaid HMO only markets it might work. The per patient price is a constant, the contracts between the carriers vary but often with a floor of paying them no lower than what FFS Medicaid pays. Profitably is only achieved through increasing enrollment and so most carriers’ efforts are around that, e.g., building networks and providing additional services.