Study shows Medicaid coverage has little clinical benefit

BMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f2878 (Published 02 May 2013) Cite this as: BMJ 2013;346:f2878
  1. Bob Roehr, journalist
  1. 1Washington, DC, USA
  1. BobRoehr{at}aol.com

Results of long awaited study have something to disappoint everybody, writes Bob Roehr

Access to healthcare through Medicaid generated no significant improvements in measured physical health outcomes in the first two years of coverage for a low income population in Oregon, according to a much anticipated paper in the New England Journal of Medicine.1

However, enrollment in Medicaid increased the use of healthcare services, raised rates of diabetes detection and management, lowered rates of depression, and reduced financial strain among beneficiaries.

Medicaid is most often thought of as a safety net insurance program that provides access to essential medical services for people on low incomes. The mixed message from this study on its effects on health adds fuel to the ongoing debate over expansion of the federal/state Medicaid program under provisions of the Affordable Care Act.

A unique opportunity

The unique opportunity for the study arose in 2008 when Oregon had limited funding to expand access to Medicaid and a waiting list of almost 90 000 people. The state used a lottery to draw the names of about a third of the people on the waiting list, who then went on to formally apply for Medicaid and enroll if they met the entry criteria.

Researchers took advantage of the randomization of Medicaid beneficiaries in their study. Analysis was based on 12 229 people in the Portland area who responded to the survey (73% response rate). Half had gained access to Medicaid under the lottery expansion and half had not; the groups shared a similar sociodemographic profile. There were no copayment barriers to accessing services.

“We found no significant effect of Medicaid coverage on the prevalence or diagnosis of hypertension or high cholesterol levels or on the use of medication for these conditions,” the researchers concluded.

“Medicaid coverage significantly increased the probability of a diagnosis of diabetes and the use of diabetes medication, but we observed no significant effect on average glycated hemoglobin levels or on the percentage of participants with levels of 6.5% [47.5 mmol/mol] or higher.”

“Medicaid coverage decreased the probability of a positive screening for depression (−9.15 percentage points; 95% confidence interval −16.70 to −1.60; P=0.02), increased the use of many preventive services, and nearly eliminated catastrophic out-of-pocket medical expenditures,” they continued.

These clinical measures were chosen because they are serious health conditions that are reasonably prevalent in the population, and treatments are available that could produce significant improvements in less than two years, lead author Katherine Baicker told the BMJ. “We looked where we thought the effect [of access to care] ought to be the largest, and still didn’t see any statistically significant improvement.”

“It is really a nuanced set of results,” the Harvard School of Public Health researcher continued. There were “substantial increases” in physician visits and use of prescription drugs. “We also saw substantial reductions in financial strain, something that is underappreciated in the debate about expanding public insurance. [These programs] are supposed to not only get access to care but also protect against financial ruin in case you have an expensive health condition.”

But critics say that if financial ruin is a principal reason for expanding Medicaid, why should aid not be extended to areas such as mortgages, student loans, and other financial burdens that also carry a broad social good?

The researchers will conduct further analyses, such as utilization of emergency services. But “in some sense the experiment is over,” Baicker said, because as of 2010, “there is no longer a treatment group versus a control group.”

Interpreting the results

Baicker believes the outcomes will disappoint both supporters and opponents of Medicaid expansion. Although the study showed some benefit to recipients, at a cost of about $1200 (£770; €911) per person covered, the benefits were not so dramatic that the reduction in hospital admissions would save money over time.

“And that leaves policy makers in the much less comfortable world of having to evaluate how much they value the benefits that enrollees receive against the cost of the program and the alternative uses of those public resources,” she said.

“It is a little sobering that you didn’t see more in two years in areas like hypertension and diabetes where we always assumed they were particularly sensitive to coverage,” said health economist Gail Wilensky, who works at Project HOPE. She served as administrator of Medicare and Medicaid in the George H W Bush administration.

“A lot of the effects that they found were in the right direction, even if they were not statistically significant,” noted Mark McClellan, former director of Medicare and Medicaid and a health policy fellow at the Brookings Institute.

“People need to keep in mind that healthcare is only one factor that influences chronic conditions like diabetes and heart disease. And health insurance only has some impact on healthcare,” he said. “We have to be reasonable about the kinds of impacts that are realistic to expect.”

McClellan said that just the expansion of coverage isn’t enough to improve health and avoid unnecessary medical costs. “Much more needs to be done about the impact and effective design of health insurance plans, and the actual delivery of care.”

“I wasn’t terribly surprised by the results,” said Diane Rowland, director of the Kaiser Family Foundation Commission on Medicaid and the Uninsured. It offers us a warning that there are many other factors that influence health,” such as work, housing, and other challenges of daily life.

Although insurance coverage can provide access to healthcare, it is only one step in the process; Rowland said, “The plans may have to do a better job in getting people to comply with their medications and diets so that they can improve some of these measures.”

Coverage “did improve access to preventive care” where benefits often accrue over a longer period of time, and Rowland said, “I think this is a good baseline but I’d want to see outcomes at five years” and beyond.

Sara Collins, vice president of health policy at the Commonwealth Fund, acknowledged that “the clinical benefits in this study are mixed.” However, she emphasized the issue of equality, saying that it was at least putting the poorest on an equal footing to the rest of the population with access to insurance coverage and care.

“There were fundamental changes in use of healthcare services,” Collins said. She said that people were “much more likely to have a regular doctor if they have health insurance, much more likely to get these critical preventive care services that everyone else with health insurance gets.”

The number of patients in the Oregon expansion of Medicaid coverage was relatively small and easily absorbed by existing providers. It may not be predictive of what occurs with the much greater expansion of coverage anticipated under the Affordable Care Act. Baicker said, “If you suddenly move millions of people into Medicaid, there may be system level effects that you don’t see if you only insure 10 000 people at once.”

Wilensky concurred. She believes the Oregon study population “certainly is among the better cases” in terms of motivated patients and challenge to service capacity. It may not be representative of what will occur with Medicaid expansion elsewhere, particularly “some of the large urban areas or the south, where there is potentially a huge increase in the number of people covered with very little change in terms of the supply of services available to them.”

Despite the study’s findings of limited clinical benefit, Wilensky said, “It is generally difficult to argue that individuals and communities are better off having large numbers of people without insurance.”

Baicker expects that advocates for and against expansion of Medicaid will interpret the study through their own values and expectations of what the program should accomplish.

“If your goal in expanding Medicaid is to improve chronic conditions like blood pressure and cholesterol, you might certainly think, if it hasn’t done it in the first two years, with a big enough effect that it is actually measurable, that gives me pause about the program,” she said.

“How do you value the benefits that we do see relative to the benefits that we don’t?” Baicker asks. “You can come to very different conclusions about what the study suggests one ought to do about expanding Medicaid or not.”


Cite this as: BMJ 2013;346:f2878


  • Competing interests: None declared

  • Provenance and peer review: Commissioned; not externally peer reviewed.