More on Medicare’s Latest Data Dump

Yesterday, we noted the New York Times‘ analysis of hospital charges from the Centers for Medicare & Medicaid Services (CMS) latest data dump. The same data dump showed how the amounts Medicare paid to hospitals and other providers for different services. The Hill‘s Ferdous Al-Faruque has pointed out some extreme differences:

health-care-costsThe agency found wide discrepancies in how much services cost in different regions of the nation and within the same geographic area. In 2012 a major joint replacement surgery cost Medicare $15,901 in Baltimore while the same procedure cost $239,138 in Los Angeles, the report says.

This variation appears too extreme. If it is a quality difference, surely the lower-quality provider is so bad that it should not be accepting patients! The seeming arbitrariness of Medicare payments might be one good explanation for the variance in costs observed by the Dartmouth Health Atlas team.

Like the physician data dump, for which we praised CMS, this is a treasure trove of data. CMS has also presented the data in a reasonably user-friendly way. It took me less than ten minutes to figure out the dashboard, which allows users to make charts and tables of almost any shape and size.

Well done, CMS. Keep ‘em coming.

Comments (25)

Trackback URL | Comments RSS Feed

  1. Mitch says:

    More from the Hill article:
    The report also notes that a hospital in Newark, N.J, charged $32,750 to treat heart failures while another hospital in the same city charged $142,000 for the procedure.

    wow….

  2. Richard Franklin says:

    Well that’s funny, because the New York Times posted an article in April with regards to Big Data and why it’s bad. The Times may make all of the predictions it chooses, but it can’t hide from the facts.

    http://www.nytimes.com/2014/04/07/opinion/eight-no-nine-problems-with-big-data.html

    Although this article on data is not necessarily about healthcare, it still applies because predictions that are complex in their projections, regardless of the field, will have flaws.

  3. Steve says:

    CMS did something…good? I’m actually pleasantly surprised. Well done, CMS.

  4. Lacey says:

    The cost differences outlined in the article are crazy. Something has to be done.

  5. Devon Herrick says:

    This is why the WellPoint / CalPERS experiment is so valuable. CalPERS picked a midpoint for joint replacement (~$30,000) but allowed retirees to go anywhere. However, patients had to pay the difference for all costs above $30,000. It wasn’t very long before virtually all CalPERS retirees were getting joint replacement for $30,000.

    Medicare needs to work like this. Rather than set prices for 7,000 procedures, it should pick a reasonable mid point and charge seniors who run up charges in excess more. Seniors that find procedures for less should share in the gains (for use on deductibles and cost-sharing. That would unleash an army of 50 million seniors all checking prices.

    • Buster says:

      Currently, it’s difficult to establish a bundled payment for Medicare services. This would be a way to begin that process. Seniors asking the price would be told “we won’t know until we perform the service.” The senior would vote with their feet and leave. Hospitals would respond or go out of business.

      • Elizabeth says:

        In medical services, “voting with your feet” is often easier said than done. If it’s a small or elective procedure, it’s feasible. But it’s not quite as easy when it’s a large or emergency procedure. Many people feel they don’t have time to make a choice.

        • Buster says:

          “Many people feel they don’t have time to make a choice.”

          No problem! When they’re in the process of being discharged after their hip replacement, kindly pay the cashier $209,138 ($239,138-$30,000).

          P.S. That wouldn’t have to happen but once or twice before the news media picked up on it. The end result is seniors would find the time and hospitals would begin competing on price.

  6. bob hertz says:

    I severely doubt that Medicare actually paid wildly different amounts for the same procedure in different cities, if the codes were entered identically.

    What a hospital charges Medicare is a non-issue, I believe. What the hospital is paid is what matters.

    I am not in hospital operations so I could be wrong.
    But charges alone do not interest me.

    • John R. Graham says:

      I agree, with the caveat that the divergence between charges and payments is symptomatic of a generally messed up system.

      The example cited is how much Medicare paid. As Linda Gorman has stated above, we can describe this as “payment”, “expenditure”, or “spending”, or words like that but using the word “cost” as the reporter did, is not accurate.

  7. Big truck joe says:

    That hip replacement disparity CANNOT exist. Physicians get paid on a Professional and Technical component and hospitals get paid on a DRG (diagnostic related group). There are slight differences in payment for city vs rural reimbursement but nothing this drastic. Plus Baltimore and LA are both cities so their reimbursements should be similar. somebody isn’t practicing good journalism – or in other words just making sh!!t up to push an agenda.

    • Devon Herrick says:

      Presumably, the disparity is reflected (to some degree) in how much uninsured patients, employer health plans and insurers are charged for joint replacements.

      But you’re correct that newspaper articles often don’t explain that nobody actually pays list prices for surgical procedures.

      • John R. Graham says:

        I searched myself under DRG 469 – major joint replacement or reattachment. The range of Medicare payments went from to $6,185.48 (charged at $16,862.90) at Weirton Medical Center in West Virginia to $109,651,67 at University of Iowa (charged at $210,512,33).

        The highest priced five or so hospitals are outliers. Nevertheless, it is happening. (Unless CMS is misreporting the data.)

        • Devon Herrick says:

          I wonder if that could be due to the implant itself? I talked to an acquaintance employed by an insurer and he said the implants vary in price by a huge margin. He was hesitant to even give me an average cost for a hip replacement because the cost varies from $30,000 to $120,000.

  8. Bob Hertz says:

    By the term “outlier” are you implying that this case had special complications?

    The abuse of the “outlier” category was a big factor in numerous frauds involving heart care in McAllen Texas and elsewhere.

    Medicare has a lot of auditors who hover over doctors for $50 in excess fees, but apparently no one audits outliers.

    • John R. Graham says:

      No, i just mean that the distribution of costs on the right-hand side does not look (to my naked eye) approximately normal. It looks leptokurtic. (I did not look at the low-cost hospitals so whether it is symmetrical or skewed I have no impression.)

      If I had the inclination, I could download the data and run the descriptive statistics within the hour.

      That is why this is so valuable. CMS has effectively crowd-sourced the auditing of these facilities, to a degree.

      I have no clue about the cases themselves. The released data only covers charges and costs.

  9. Bob Hertz says:

    Actually the government has studied the use and abuse of ‘outlier’ claims, as it turns out. As I remember, the auditors estimated that $5 billion in outlier claims is being paid out each year.

    Not chicken feed, but still less than one per cent of total Medicare spending.

    What is more interesting, in a way, is that once again we see in health care that the more you charge, the more you make.

    Medicare actually made this worse in the early years after its enactment. Doctors and hospitals found that higher fees could become the ‘ordinary and reasonable’ standard.

    A lot of the discussions on this very blog are about different ways to change this dynamic. Reference pricing is a good start, for procedures that are not emergencies and not life-threatening.

Leave a Reply



If you want a picture to show with your comment, go get a Gravatar.