Much has already been written about the Oregon Medicaid study that came out in the New England Journal of Medicine. Unfortunately, the vast majority is reflex, rather than reflection.  The study seems to serve as a Rorschach test of sorts, confirming people’s biases about whether Medicaid is “good” or “bad”.  The proponents of Medicaid point to all the ways in which Medicaid seems to help those who were enrolled – and the critics point to all the ways in which it didn’t.  But, if we take a step back to read the study carefully and think about what it teaches us, there is a lot to learn.

Here is a brief, and inadequate, summary (you should really read the study):  In 2008, Oregon used a lottery system to give a set of uninsured people access to Medicaid.  This essentially gave Kate Baicker and her colleagues a natural experiment to study the effects of being on Medicaid. Those who won the lottery and gained access were compared to a control group who participated in the lottery but weren’t selected.  Opportunities to conduct such an experiment are rare and represent the gold standard for studying the effect of anything (e.g. Medicaid) on anything (like health outcomes).  Two years after enrollment, Baicker and colleagues examined what happened to people who got Medicaid versus those who remained uninsured.  There are six main findings from the study.  Compared to people who did not receive Medicaid coverage:

1. People with Medicaid used more healthcare services – more doctor visits, more medications and even a few more ER visits and hospitalizations, though these last two were not statistically significant.
2. People with Medicaid were more likely to get lots of tests – some of them probably good (cholesterol screening, Pap smears, mammograms) and some of them, probably bad (PSA tests).
3. People with Medicaid, therefore, not surprisingly, spent more money on healthcare overall.
4. People with Medicaid were less likely to go bankrupt due to healthcare expenditures.
5. People with Medicaid had less depression and overall, had better health-related quality of life.
6. People with Medicaid did not have meaningful improvements in their hypertension, cholesterol, diabetes, or other measures of overall health.

It is first worth taking a moment to dispense with those who will try to nitpick the methodology.  Read through the paper carefully and spend time going through the 62 page single-spaced supplementary appendix and you’ll find that this is about as good of a study as will be done on this topic for the next generation.  Kate Baicker, who led the study, is the smartest person I know and whenever I disagree with her, it’s because she’s right and I’m wrong.  These are the gold standard of folks using the gold standard of methodology to answer an incredibly important question: what is the effect of Medicaid on financial, mental, and physical health.  So, let’s get to the lessons.

1. Insurance works.  The goal of my homeowner’s insurance is that if I have a fire, I won’t become bankrupt.  The goal of health insurance should be to ensure that if you get hit by a bus, you won’t go bankrupt.  Medicaid, as insurance, worked.
2. Insurance gives you peace of mind.  I never lie awake worrying that if I get sick, my family will go bankrupt.  Medicaid may therefore be giving people reassurance, and making them feel better, which may be why there was less depression in the Medicaid group (they certainly weren’t taking more anti-depressants).
3. Insurance improves access to healthcare services.  Although people without health insurance still got healthcare (they were spending $3,257 per year on healthcare, seeing 5.5 doctors a year, getting medications, outpatient surgery, etc.), people on Medicaid got more.

But this was all predictable and none of it should be a surprise.  What has been fundamentally misunderstood is why it didn’t lead to better health.  And that is the biggest lesson from Oregon:

Healthcare isn’t health, and the missing link is quality.

Let’s unpack this.  To date, the notion for improving health has been simple: if we give people access to health insurance, they will get more care, and therefore will have better health.  Oregon tells us that’s not quite right.  The lesson from Oregon is:  if we give people access to health insurance, they will use more healthcare, and they will feel better for it.  But, their health may not be that much better off.  How could this be?

The explanation is simple.  It’s not about access to healthcare; it’s about access to high quality healthcare.  Americans fail to consistently receive high-quality care and the people on Medicaid in Oregon were no exception.  There is some evidence that providers who disproportionately care for Medicaid patients deliver lower quality care, but the problem is much bigger than Medicaid.  In fact, most Americans get pretty mediocre quality healthcare.  Therefore, not surprisingly, healthcare often fails to improve health.

In the Oregon study, we see that many people, especially in the high risk groups, have poorly controlled blood pressure, diabetes, hypercholesterolemia, and depression.  Yes, Medicaid seemed to help a little, but not enough.  People with Medicaid had, on average, 7.2 office visits over the past 12 months.  That’s more than once every 2 months.  Yet, their blood pressure, cholesterol, and blood sugars barely budged.  This is not an access problem.  This is not about “Medicaid is bad” or “insurance is bad”.  This is about what happened (or did not happen) in those visits – namely, evidence-based care that we know improves health.

Most healthcare policymakers talk about the three legs of the stool of the healthcare system: cost, access, and quality.  The Affordable Care Act makes a big effort to improve access, but does less on cost and little on quality.  That’s unfortunate.  Oregon reminds us that if we want to improve the health of the population, we will have to make real and concerted efforts to ensure that people are receiving high-quality care.  We can’t just improve access and think that our job is done – in fact, its just the beginning.

Ultimately, while many factors affect health (such as education, income, neighborhoods where we live, etc.), healthcare matters too.  And it better – we’re spending $2.8 trillion on it.  Oregon tells us that insurance has its benefits – it gives us peace of mind and improves access to health services like office visits and preventive screening.  But it doesn’t do that much for health, because it’s not about access to healthcare.  It’s about access to high quality healthcare.  Quality is the link between healthcare services and better health outcomes.  And we need to spend more time working on that.

Ashish Jha is an Associate Professor of Health Policy and Management, Harvard School of Public Health.  He blogs at An Ounce of Evidence and can be found on Twitter @ashishkjha.