Interpreting The Oregon Medicaid Study: Health Is More Than Insurance

Here’s a very clear analysis of a very confusing study that came out last week and was framed in wildly different ways by various media. The Oregon Health Insurance Study was complicated, for sure, but the bottom line, argues physician John Lumpkin, in the current Health Affairs, is fairly simple: “Better health requires health insurance coverage, but it doesn’t end there.”



Published in The New England Journal of Medicine, the landmark Oregon study by researchers at Harvard and MIT offered a snapshot that compared Oregonians on Medicaid to those not on the public assistance program. (A 2008 lottery among low-income residents established the two groups, which effectively created a treatment and control arm of the experiment.)

The findings were mixed (generally not good for a headline): on the up side, after about two years on the program, patients showed improved mental health with a dramatic drop in depression among the newly insured, and more financial stability. It also found these patients had greater interaction with the health care system, and more preventive care, in general.

The down side, however, was that researchers found no measurable improvement in specific health outcomes, including high blood pressure, cholesterol and blood sugar control in people with diabetes.

Some interpreted these findings as proof that Medicaid is a total failure, and should not be used as a model for health reform. Others emphasized the positive.

Here’s Dr. Lumpkin’s concise interpretation:

So far, the Oregon Health Insurance Study shows us that people who obtained Medicaid coverage received more health care services in the first two years—especially needed preventive care—and had less depression and financial worries. Their health outcomes weren’t significantly better, but at least they are now participating in the health care system and getting the care they need, without plunging their families deeper into poverty. From this vantage point, the glass seems more than half full.

Please follow our community rules when engaging in comment discussion on this site.
  • JenInBoston

    Why would conditions like diabetes and hypertension be expected to improve?! They are “lifestyle diseases” and we do virtually nothing in America to change the lifestyle. We just toss pills at them, followed by surgery. Why on earth would Medicaid change this? Patients are lazy. Even doctors have gotten used to how fat and unhealthy and unwilling to change their habits people are. I recall talking to a colleague who had had lap band or stapling surgery, and then had lost and very slowly regained almost ALL the weight. She was about 5′ 3″ and 185 LBS and was telling me, “oh, I know I should exercise a little but it’s not like I’m obese, or something. I need to lose 10 or 20 pounds.” What?! Oh my gosh. Yes, you’re obese–not chubby, not curvy, not Rubenesque. Obese. America is in complete denial–even the goodhearted folks who apparently thought going on Medicaid would make people decide to start caring about their bodies. Yowsers.

    This is all in stark juxtaposition to the WBUR story yesterday about how elderly, low-income Chinese exercise together almost religiously in order to avoid medical conditions and treatments they cannot hope to pay for.

  • Josh Archambault

    Dr. Lumpkin’s statement that “Better health requires health insurance
    coverage” is not nearly nuanced enough. This most would agree that “cadillac” insurance looks very different from a mini-med policy.

    The real debate in the Medicaid program is how to best use public dollars in
    order to assist low-income families and the disabled. From my perspective, it
    is not a matter solely of the amount of money being spent, I would be happy to
    continue spending if we knew we were getting good results. Oregon has questioned that basic argument.

    Even Jon Gruber seems to have reframed his opinion based on this experiment:

    “I would view this study as somewhat weakening the argument for universal
    coverage based on health improvements,” Gruber said, “and greatly strengthening the argument based on financial security and mental well being.”

    This is like referencing a study that finds kids going to school results in
    very little learning, but improves socialization and increases the diagnosis of
    ADHD. Don’t get me wrong, both positive outcomes, but not the primary rationale
    for schooling, and reason enough to question some of the underpinnings of the

    If the goal of Medicaid is improved financial security and better mental health
    assessment, then we should provide a catastrophic insurance plan and improve
    screening and treatment services for mental health. This program would be
    significantly cheaper than the status quo, and free up additional funds to address
    some of the flaws in the current system. For example, we could guarantee better
    access to doctors for those on Medicaid by paying more like a “concierge medicine service” and which would better coordinate care.

    This is the debate that should be happening, not reframing the results to focus on the half-full glass, while the vulnerable linger in a $450 billion a year mixed results program. Don’t they deserve better?

    • Reasonable?

      We don’t need to add more enginerring to an overengineered system.
      Here’s my suggetion: let states experiment with patient directed incentive programs for Medicaid.

      Why not something like an “Earned Wellness Tax Credit”.
      Get your BMI below 30 and your HbA1c below 6 and get a $500 check at the end of the year.

      People would quickly figure out the diet ahd health habits that would allow them to achieve that goal.

      It could be that simple..

  • Dennis Byron

    Just to be accurate from a research point of view (and assuming the authors didn’t change the methodology after the study began), it is not accurate to say

    “the landmark Oregon study by researchers at Harvard and MIT offered a snapshot that compared Oregonians on Medicaid to those not on the public assistance program.”

    What the research appears to me to offer is a model — not a snapshot — of what being on Medicaid in Portland, OR might have been like in 2010 and 2011 compared to a model of what not having insurance in Portland, OR during those years might have been like. Models have the same problems as indexes such as the one described in another CommonHealth blog post recently.

    Apparently the authors had to model the results — rather than use the hard indisputable numbers — despite the large sample because many of the group of so-called “winners” had insurance before winning the lottery (as did many of the “losers”) and many of the losers who ended up in the control group kept their insurance or obtained other insurance after they “lost” the lottery.

    I say this because when the preliminary results came out two years ago, the lead author gave me a bootleg copy. I did not ask for a bootleg copy this time because these results are so highly modeled that in my opinion they don’t say much other than that some smart people in Portland OR saw a way to reduce their healthcare costs and rightly took advantage of it. (Again, I am assuming the authors didn’t change the methodology; I doubt they did.)

  • Reasonable?

    Having insurance is better than having no insurance, however health insurance itself is no panacea.
    To improve health outcomes, people/patients have to have an incentive to be healthy or well.
    Medicaid currently has no incentives for patient wellness, so the results above are not surprising.

    We have a “sick-care” system. When you get health insurance you get access to prescriptions & treatments that might slow the progress of the chronic diseases that are underway due to other behavioral and environmental factors.

    Our system does address the root causes.r