Topher Spiro is the Vice President for Health Policy at the Center for American Progress.
Imagine an academic study that is so well-timed, so relevant to a current policy debate, that it's widely cited to support a major policy decision. But now consider this: the study has severe limitations and doesn't turn out to support that policy decision after all.
Sound familiar? The study by Harvard's Carmen Reinhart and Ken Rogoff on the effects of debt on the economy (which was used to support fiscal austerity) fits this description – but now we also have the Oregon Medicaid study. Earlier this month, this study found that the Medicaid program – which provides health insurance to poor people – didn't improve certain measures of physical health. Many commentators took this as proof that Medicaid is fatally flawed and that states shouldn't expand their programs under the Affordable Care Act law.
The study has been heralded as the gold standard because it was "randomized." It compared people who got Medicaid through a lottery with people who did not – the only difference between the two groups was the randomness of the lottery. And random experiments are indeed the gold standard. But in key respects, this study is far from the gold standard on the question of Medicaid's effects on physical health.
The study didn't measure overall health status – or anything close to it – and the effects over time. Rather, it included only a few measures, such as blood pressure, cholesterol and blood sugar levels, over only two years. It's unclear why the researchers couldn't have waited a few years to provide a clearer picture.
But more important, it's now clear that the study wasn't even designed to detect an effect on physical health, even if one existed. Simply put, it didn't include enough people in poor health to measure effects that would be clinically meaningful. Doing the math, the economist Austin Frakt found that the sample sizes were as much as 23 times too small.
Before jumping to conclusions, it's important to have a plausible theory about why Medicaid might not improve physical health. One theory is that because Medicaid pays doctors so little, few doctors participate in the program, so the insurance card doesn't actually provide access to care. But the study showed that Medicaid substantially increased the use of health care services, including the use of preventive and screening services.
To believe that Medicaid doesn't improve physical health, we'd have to believe that clinical treatments don't work. The study showed that Medicaid improved the diagnosis of diabetes and increased the use of diabetes medication. Do we really believe that diabetes medication doesn't work, despite mountains of evidence from clinical trials?
It's possible that access to health care might not improve physical health if the quality of care is poor. We know that doctors don't always follow evidence-based guidelines for treatment, care isn't always coordinated and the payment system rewards more and more tests and procedures that may have no marginal benefit and may even be harmful.
Our health care system has a great deal of room for quality improvement, but this isn't a problem unique to Medicaid. Gutting Medicaid, as critics propose, isn't a rational or targeted policy response. Rather, look no further than Oregon – which is reforming its Medicaid program to provide more coordinated care.
To say that improvement is always possible doesn't mean that Medicaid doesn't improve physical health. The Oregon Medicaid Study certainly didn't prove that – the way it was designed, it couldn't have. But where the study did have large sample sizes, it showed that Medicaid does have sizable effects, on financial security and mental health.
The study showed that Medicaid nearly eliminated catastrophic medical costs. And people were about 60 percent less likely to borrow money or skip payments because of medical bills. The purpose of insurance, as a means of payment, is to provide financial security – and Medicaid clearly fulfills that purpose.