Medicaid managed care may not be all it’s cracked up to be

by Academyhealth on June 6, 2012 · 8 comments

Just over a month ago, Christopher Flavelle  published a study on the growth of Medicaid managed care, and how many states are turning to these types of plans in order to address “out of control” Medicaid growth. His study showed that growth in Medicaid was much less that one might expect, given the rhetoric many of us often hear.

Recently, he published the second of his three pieces on this subject, “Evidence Is Limited That Medicaid Managed Care Reduces ER Visits.” In it, he examines reports of managed Medicaid performance in the five largest states. His findings:

The largest Medicaid managed-care plans in Florida and Illinois, and some for-profit plans in California and New York, provide health-care access that is significantly and consistently worse than the national median for such plans. Studies suggest that failing to provide some types of care, such as childhood immunizations, may lead to higher medical costs in the future.

Since pediatrics is always of special interest to me, here’s their chart describing how these plans do with respect to vaccinations for kids:

In California, two of the four plans are performing slightly better than the state average with respect to immunization rates. However, Health Net fell below the state average in 2009. Anthem has been well below the average consistently, and falling steadily since 2007. Florida looks better. In Illinois, one plan did better than average, and one plan did worse. And in New York, three of the four plans consistently underperformed the state average. Overall, there’s plenty of reason to be concerned.

Other findings:

The evidence is limited that managed care sustainably reduces emergency-room visits in the five largest states. California had shown some reductions in ER visits among managed-care plan patients, but the gap between them and fee-for-service beneficiaries is shrinking. In New York that gap has almost disappeared, and in Illinois, ER visits are now slightly more common among managed-care beneficiaries. Only in Texas does managed care consistently produce significantly lower ER visits than in fee-for-service, though the cost of those visits is much higher.

It’s surprisingly hard to do this kind of work. Evidently, Flavelle had to get a fair amount of this data through Freedom of Information Act requests. But this finding may be the most significant. This is, as the study contends, the “litmus test” for managed care. After all, actively “managing” care is specifically supposed to reduce the need to emergency room visits. It’s not clear that these plans are providing that expected benefit. Even in Texas, where managed care is succeeding in reducing the number of emergency room visits, the cost for those visits are, on average, more than twice as expensive as emergency visits under fee-for-service plans. Why that is so isn’t clear.

To be fair, this study doesn’t prove that managed care can’t improve or maintain outcomes that are important over fee-for-service plans. It also doesn’t prove that managed care can’t reduce emergency department utilization in its beneficiary population. However, it does show that many plans are performing worse than we’d like and many plans are not reducing ED utilization. The Medicaid Health Plans of America responded to the study here, without ever really questioning any of the findings.

This isn’t concerning just from a quality standpoint. Providing care that is sub-optimal can also lead to increased spending in the long term, especially with respect to care such as childhood vaccinations. That is a problem, as we see more and more states turning towards these plans in an effort to reduce spending. The worst thing we could do in our efforts to reduce Medicaid spending would be to make outcomes worse while also increasing costs. We will need to keep a close eye on this, and, perhaps, find a way to make it easier to do so.



Dr. Aaron E. Carroll is an associate professor and vice chair of health policy and outcomes research in the department of pediatrics at the Indiana University School of Medicine. He blogs about health policy at The Incidental Economist and tweets at @aaronecarroll.

As part of our ongoing effort to raise awareness of health services research and increase its application in policy and practice, AcademyHealth has partnered with Austin Frakt, Ph.D., and Aaron Carroll, M.D., M.S., to contribute posts on the subjects of health care costs, delivery system transformation, and public and population health – areas AcademyHealth has identified as a priority in the current policy environment. As regular contributors, they’ll be discussing current events with an eye toward how new and existing research informs the issues.



Rick Lundgren, RN June 6, 2012 at 9:42 am

I have been leery of the Medicaid for all how cannot afford health insurance one there own. There was a study a few years ago which correlated Medicaid ED visits and subsequent hospitalizations with the lack of specialists that take Medicaid.

I think more data is needed to identify the root causes. If is Medicaid expanded to all will people use it? If not then why not? One of the major issues with the healthcare “system” is access. People may very well have insurance, but if they cannot still not access the “system” easily, then it will fail.

Weiwen June 6, 2012 at 4:54 pm

This is certainly concerning. Focusing specifically on dual eligibles with disabilities, I found some evidence that MCOs which coordinate both acute and long-term care did sometimes produce superior outcomes and lower spending than FFS. However, there are only a handful of extant fully integrated plans. Those guys have a sort of survivorship bias going, in that they likely only survived because they’re good (good management, good relationships with providers, good provider networks, good relationship with state, or some combination).

And there was some inconsistent evidence as well – Arizona’s managed LTC system may have produced poorer outcomes in long-stay nursing facility residents, Minnesota Senior Health Options may not have reduced costs, etc. These questions need to be answered.

I’m of the opinion that good managed care plans can be better than the status quo, which is quite clearly broken for dual eligibles. But states need to have strong enough oversight to select the best plans, and those plans are still going to have to work hard to improve outcomes and contain costs. And with the wave of state budget cuts, I’m not certain states will be able to attract and retain the quality personnel necessary to properlyh oversee managed care plans.

Sarah Thomas June 8, 2012 at 8:47 am

The author indicates that it is hard to get the data for these types of analyses but we would like other researchers to know that NCQA makes available quality results for a whole host of measures for more than 100 Medicaid plans. These can be obtained through Quality Compass or through special data extracts.

Lisa Lines June 13, 2012 at 5:07 pm

The FFS average ED visit cost could be lower because of more low-acuity visits, which are more expensive to treat in the ED than in other settings, but are still less costly than visits for high-acuity complaints (which might need more imaging and other procedures).

In other words, if the number of ED visits for less-acute complaints went down in Texas Medicaid enrollees, that would make their average costs for ED visits go up relative to the average costs for those enrolled in FFS.

southwest fragrances September 16, 2013 at 5:34 am

It’s an awesome piece of writing in support of all the online
people; they will take benefit from it I am sure.

pcb assembler August 25, 2014 at 12:14 am

These tiny components can damage easily if a large amount of current passes through them.
They also sustain enough space between the elements to allow warm dissipation. The screen paste file for surface mount applications will be
required to order the stencils for the solder paste application to the board prior to the part

Comments on this entry are closed.

{ 2 trackbacks }