Wednesday, January 27, 2016

A question for healthcare economists (update)

I have a question for healthcare economists. Bear with me through a few bullet points to get there. [Note update at end: there's an answer of sorts in Kenneth Thorpe's analysis of Bernie Sanders' plan.]

  • It's well known that the U.S. pays far more for healthcare -- per capita, per procedure, and as percentage of GDP --  than any of its developed-world peers. According to the OECD, the US spent 16.4% of GDP on healthcare in 2013. The next highest spenders, Holland and Switzerland, spent 11.1% of GDP.
  • A main reason, if not the driving reason, is that in the U.S. healthcare payers are fragmented, diminishing their buying power. Private insurers and self-funded employers pay rates that may average 150-160% of Medicare rates.

  • Medicaid rates are far lower than Medicare's. Yet, U.S. public spending on healthcare as a percentage of GDP exceeds the median total healthcare spending in OECD countries, including total spending in Japan, the U.K., Australia and New Zealand  -- while leaving 30 million uninsured and perhaps 60 million underinsured.
  • At the same time, we're told that hospitals would go bankrupt and physician practices would not meet their costs if they had to depend entirely on patients with government-provided insurance.
  • It would appear that high costs are baked into providers' business models. Some of the excess cost is administrative, some is in high compensation for providers, and some is due to high cost facilities and equipment and high utilization of high-cost equipment, procedures and medications. A Commonwealth Fund look at U.S. Health Care from a Global Perspective shows far higher use of MRIs, CT scans and prescription drugs in the U.S. than in peer nations.*
  • Current cost control measures baked into the ACA and the 2015 doc fix legislation reforming Medicare payment focus mainly on squeezing out wasteful care. But let's leave utilization aside for the moment. By other measures -- e.g., hospital and physician use --  U.S. utilization is not high. And in Japan, where overall spending is very low, imaging use if very high (and dirt cheap, I seem to recall reading somewhere).

Here's my question: to get U.S. healthcare spending to a level comparable to that of relatively high-cost peers like Switzerland or Holland, what would the average or universal payment rate need to be as a percentage of current Medicare rates? (That's leaving aside drug costs, which account for about 11% of healthcare spending.)

In other words, what's our target?

I realize that to the extent that payment methods are changing, the target is a moving one. Still, it seems to me worth asking: what uniform, blended or average rate for all payers right now would bring U.S. spending in line with, say, Switzerland's? What would the rate need to be if healthcare insurance were truly universal?

Arguably Switzerland is the wrong target. The U.S. is going to be paying more than other countries for healthcare for decades at least. Alternatively, then: what uniform rate rate now would be needed to sustainably provide affordable healthcare to everyone in the country? Of course, that question implicates our rate of taxation, a question on which consensus is, shall we say, a tad elusive.

It will probably take decades for the U.S. to tack its way to an all-payer system, or something like it -- that is, a system in which all institutions that pay for healthcare pay the same rates, or at least pay coordinated rates that move together like currencies tied to the dollar (say, private insurers paying 120% Medicare).

Whatever the target may be, we can't get there quickly -- only by squeezing growth in prices as well as usage.

If I've framed the question wrong, please let me know...

UPDATE, 1/28: Healthcare economist Kenneth Thorpe of Emory, who helped design Hillary Clinton's health plan in 1993 and wrote a single payer proposal for  Vermont in 2006, has an answer of sorts to my question. Pricing Bernie Sanders' all-payer plan (at almosty twice the price Bernie's chosen economist, Gerald Friedman, placed on it), Thorpe writes that Medicare currently pays hospitals at a rate covering 89% of their costs, with private insurers presumably making up the shortfall. Thorpe estimates that a single payer system would have to pay hospitals 105% of cost, which comes out to 118% of Medicare,  Later in the analysis, that 105%-of-cost is assumed as the rate to be paid to providers generally. But it may be that Medicare pays physicians and other providers a percentage of cost other than the 89% Thorpe cites for hospital inpatient care.

* The Commonwealth Fund brief, published in October 2015, cites somewhat different OECD spending figures than the ones I linked to. I think that's because the OECD chart provided at the link excludes investment -- it's limited to spending on actual provided healthcare.


  1. I do not have any first hand knowledge of medical billing, but my impression is this:
    Cutting the rate of reimbursement only causes providers to manipulate and cherry-pick the fee schedule. They simply find the fees that have not been cut and are still more generous than the work involved.

    For that reason, I am skeptical that a percent of medicare will be effective cost control.

    1. Changing the fee schedule does not produce a 1:1 reduction in spend as there is some coding adaption that you posit, but it does reduce the fees available especially if there is an aggressive audit function (ie, we expect to see 7 to 12% of claims at Level 5 , you're submitting 29% of your claims as level 5, let's see the case notes....)