The Oregon Medicaid Experiment: Good Study, Bad Politics

Examining the Oregon Medicaid Study

By now, most Disease Management Care Blog readers have probably read or heard about the release of the health and costs outcomes data from the Oregon Medicaid experiment. The purpose of the study was to ascertain whether one of the arguments still being used to support the Affordable Care Act is really true, i.e. that health insurance leads to better health leads to lower health care costs.

It's not a minor argument. When the important provisions of the ACA go "live" in 2014, liberal-progressive supporters of the ACA will be tenaciously seeking to validate the merits of health care reform, while conservative critics will be looking for any excuse to strangle Obamacare in its crib.

So naturally, if a well performed study supported or contradicted the health insurance hypothesis, it would make little political difference.

How was the study done?

Unable to afford universal enrollment of approximately 30,000 eligible persons into Medicaid, Oregon figured the only fair way to administer the program was to have a lottery. Because participating individuals were allocated to one of two options (with or without Medicaid) by chance, Oregon's approach had all the makings of a prospective randomized clinical trial.

The lottery was conducted in 2008 and the health status and health care costs for the winners and losers were compared using face-to-face interviews an average of 25 months later. While the lottery was state-wide, the study itself was limited to the Portland area. The researchers planned to compare the health status of 10,405 individuals who had won the lottery to 10,340 individuals who had lost and were not enrolled in Medicaid.

Not all persons signed up for Medicaid and others couldn't be tracked down for the one-on-one interviews, leaving a final number of 6487 (62% of the eligible) lottery "winners" versus 5842 (57% of the eligible) "losers."

What were the results?

All in all, the results were disappointing for persons believing the health insurance hypothesis. Having Medicaid didn't lead to better control of high blood pressure, diabetes, blood cholesterol levels or overall cardiac risk. There was also no impact on the likelihood of being admitted to a hospital or having to go to an emergency room.

The good news was that having Medicaid seemed to lessen the likelihood of battling untreated clinical depression and, if pap smears and prostate specific antigen tests are markers for access to primary care services, Medicaid increased that also. Last but not least, Medicaid participants were less likely to have catastrophic medical expenses or be in debt to cover medical bills.

Was the study perfect?

Nope. The study's generalizability was limited by being restricted to persons age 19-65 years in Medicaid and in an urban environment. Two years may not be long enough to truly gauge the benefits of having insurance. Many persons were lost to follow-up. Statistics limited the ability of the researchers to spot smaller improvements in health status.

And how did supporters and detractors of Obamacare react?

As expected. If bloggers are a window into the soul of the body politic, this KHN article amply demonstrates that no one's mind has been changed. Between inflating or minimizing the study's imperfections, cherry picking the outcomes and spinning them, we are no closer to achieving any consensus.

The Disease Management Care Blog's take:

Studies like the Oregon Medicaid study are not only rare, they're as good as we're going to get. Since no study is perfect, the likelihood that Washington DC will be able to use any nuanced research insights to inform the next steps in pursuit of the Triple Aim is not good.

The battle lines have hardened.