Political Animal

Blog

May 02, 2013 10:40 AM Data Points and Dogma

By Ed Kilgore

By now, if you read MSM or conservative web sites, you’ve probably heard that a new randomized study of Medicaid in Oregon (published yesterday in the New England Journal of Medicine) shows the program has no impact on the health of people receiving it, which, some say, means the program is a worthless waste of money, or that its expansion via Obamacare should be resisted.

Kevin Drum, who has been following this development closely, reports the results rather differently:

In a nutshell, Oregon held a lottery a few years ago in which some people received Medicaid coverage and others didn’t. Today’s study is a two-year followup, and the headline result is that “Medicaid coverage generated no significant improvements in measured physical health outcomes.” But it turns out that “significant” is doing a lot of heavy lifting here, and the headline is extremely misleading.
In fact, the study showed fairly substantial improvements in the percentage of patients with depression, high blood pressure, high cholesterol, and high glycated hemoglobin levels (a marker of diabetes). The problem is that the sample size of the study was fairly small, so the results weren’t statistically significant at the 95 percent level.
However, that is far, far different from saying that Medicaid coverage had no effect. It’s true that we can’t say with high confidence that it had an effect, but the most likely result is that it did indeed have an effect. The table below shows the point estimates. Note also that in all cases, the use of prescribed medication went up, in some cases by a lot.
Bottom line: It’s more likely that access to Medicaid did improve health outcomes than that it had zero or negative effects. It’s just that the study was too small to say that with certainty. For laymen, as opposed to stat geeks, the headline result of the Oregon study was “Possibly positive but inconclusive,” not “Had no effect.”

Although most of the chatter involves the impact (or lack thereof) of Medicaid coverage on this brief list of health indicators, there’s no controversy whatsoever over one positive impact, as noted by Aaron Carroll and Austin Frackt:

Financial hardship matters. Here Medicaid shined. It hugely reduced out of pocket spending, catastrophic expenditures, medical debt, and the need to borrow money or skip payments.

Along with the undisputed finding in Oregon that Medicaid coverage also boosted use of preventive services, that’s a pretty big deal, since two of the principal purposes of Medicaid are to help low-income folk pay for health services and make regular physician visits instead of showing up in the emergency room for uncompensated care once a condition has become acute.

I don’t know anyone, however liberal, who believes Medicaid is a perfect health insurance program; its balkanization among states with different coverage and eligibility levels, and different degrees of administrative competence, makes it difficult to generalize. But the idea that this very limited Oregon study justifies increased balkanization of Medicaid via a block grant or “state flexibility” (typically code for letting states dump large numbers of people from the program altogether) represents a total non sequitur. More studies are welcome, but if this is supposed to be the silver bullet “proving” Medicaid is useless, then clearly, a data point is being turned into dogma by those who have already made up their minds that poor people should be left to their own virtually non-existent resources, or should continue to show up in emergency rooms when their blood pressure or A1C levels are far out of control.

Ed Kilgore is a contributing writer to the Washington Monthly. He is managing editor for The Democratic Strategist and a senior fellow at the Progressive Policy Institute. Find him on Twitter: @ed_kilgore.

Comments

  • David Martin on May 02, 2013 11:00 AM:

    In the Florida legislature, some study or another (I haven't seen any actual citation) has been held up as showing that Medicaid is worse than no insurance at all--in other words, Medicaid works like some Stalinist factory that turns valuable raw materials into useless junk.

    A Republican legislator from Sebring added that expanding Medicaid would displace private charity, the proper source of medical aid to those who can't afford it. She might actually like reading Ed Kilgore's piece on "Communitarian Versus Individualistic Conservatism."

    In any case, the legislature won't pass Medicaid expansion this year. The New York Times has a well-written story. For non-subscription coverage, check the Tampa Bay Times.

  • bleh on May 02, 2013 11:07 AM:

    Oh please. Most of the rightwing frothing and foaming turns on use of the word "significant," which the authors used in the technical statistical sense -- ie, their estimate of the effect, while positive, could not be shown to be nonzero at a 95% confidence level -- and the wingnuts are taking to mean "any at all."

    It's typical silly, sophomoric wingnut word-games. It's the sort of thing that wins points in fifth grade.

    And there's no point in wasting precious electrons trying to explain this, much less the deeper point that an OR study of cardiovascular and diet-related health spanning TWO YEARS has an EXTREMELY limited applicability.

    You want their real opinion, ask whether they are willing to give up Medicaid for their elderly parents who can't afford nice nursing homes on their own, or whether they're willing to give up their own Medicare. Then suddenly there will be all sorts of value.

  • James on May 02, 2013 11:27 AM:

    Jesus, Ed. If you are going to post about the results of a hugely important health study, the least you should do is to seek out people with actual expertise in statistics and formal research -- an epidemiologist, medical experts, at least someone who with background in interpreting formal research data, instead of a bunch of people who pull opinions out of their ass to fill up a blog post. You aren't helping matters by piggybacking on Drum et al. Put some effort into it.

    Brad DeLong is having a discussion and highlighting comments by people who actually know what they are talking about. (Note: Prof DeLong pulled his first-off opinion out of his ass this morning and has regretted it, as shown in his twitter feed.)

    The best one-off comment at DeLongs' place is: http://delong.typepad.com/sdj/2013/05/oregon-medicaid-lottery-says-medicaid-expansion-definitely-worth-it-hoisted-from-incidental-economist-comments-weblogging.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+BradDelongsSemi-dailyJournal+%28Brad+DeLong%27s+Semi-Daily+Journal%29


  • OKDem on May 02, 2013 12:49 PM:

    Don Taylor at Reality Based Community goes into the fundamental issue of applying the results for Portlandia to the country in general.
    One was the clumsy way the diabetes results were reported but still seem to imply Portland handles diabetes better than the rest of the country.
    Two may explain part of the diabetes anomaly; it appears Portland in general sees a doctor nearly twice as often as the US average.

    http://www.samefacts.com/2013/05/medicaid/the-portland-oregon-medicaid-study/#more-40506

  • mr.peabody on May 02, 2013 1:01 PM:

    @James: People don't read Ed's blog for complicated statistical data. If people want that, there are plenty of blogs for that perspective, as you found.
    Ed was more or less pointing out the usual Conservative misrepresentation of the study. That's what he does. No need to bash him for it.
    Maybe you should start your own blog and cover politics from a statistical point of view.

  • Oregonian on May 02, 2013 2:18 PM:

    Oregon's governor, John Kitzhaber, was an ER doctor before going into politics. In the 8 years* he wasn't Oregon's governor, he founded the Archimedes Movement (http://en.wikipedia.org/wiki/Archimedes_Movement) and served as Director of the Center for Evidence Based Policy at Oregon Health & Science University. So Oregon is the perfect place to try new healthcare policies.

    *First two terms 1995-2003, current term 2011-current (Oregon only allows two consecutive terms; Kitzhaber is the first to win a third, obviously non-consecutive)

  • Th on May 02, 2013 2:31 PM:

    The study reported that people went to the doctor, were diagnosed and received treatment. That this treatment was only slightly effective says much about our medical treatment and nothing about the funding of treatment. The financial situation of the participants is the only item that has anything to do with funding and it improved substantially.

  • RIC HEIVILIN on May 06, 2013 3:00 AM:

    This so called Oregon report was Koch brothers owned lock stock and barrel once reputable Cato institute. I keep posting and writing warning people that Cato is nothing more than an A.L.E.C. tool. It is truly a shame that Cato has become a crossroads sold their soul to the devil unreputable propaganda machine never to be trusted again.