Winning the Health Care Lottery

Yesterday, another installment was published from the research team of the Oregon Health Insurance Experiment. The major finding – Medicaid coverage results in a 40% increase in emergency department (ER) use. Many of the health care pundits quickly sifted through the scientific results to support their opinions. You can read some of them here:

Sarah Kliff reports the facts: Expanding Medicaid doesn’t reduce ER trips. It increases them. Scott Gottlieb claims that Medicaid fails the poor and Obamacare will fail the middle class while Avik Roy thinks that the Oregon study undermines the rationale behind the Affordable Care Act. Aaron Carroll reminds us that more emergency room use isn’t necessarily bad.  Art Kellermann conjures up a 20-year old paper explaining why Medicaid patients likely have to go to the ER in the first place instead of visiting a primary care doctor.

Let’s assimilate all this information – in the context of the pre-existing Oregon results – into what it all actually means for low-income Americans, ERs, and everyone else. As a quick reminder, the was a randomized trial of uninsured patients. Some were offered Medicaid via a lottery. Others were not. The researchers reported the marginal effect of Medicaid. Here are 5 important findings from the most recent publication (and my interpretation in italics).

• Medicaid does not have an effect on emergent, non-preventable conditions (e.g. heart attacks, trauma, etc.) EMTALA and ERs are doing a fine job of providing equal access to emergency care.
• People without prior visits to the ER are most likely to increase their utilization once getting Medicaid. Medicaid won’t exacerbate the ER frequent-flyer issue; it will allow people to enter the health care system for the first time.
• ER visits resulting in hospital admissions were unchanged. ER visits resulting in the patient being sent home were increased. Medicaid prompts treatment for lower acuity conditions in the ER setting.
• Medicaid increases the number of primary care treatable ER visits by over 50%. It appears that the primary care infrastructure is largely inadequate in Portland, as I suspect it is elsewhere in America.
• The following conditions were most likely to see increased visits to the ER after Medicaid enrollment – chronic conditions, ambulatory sensitive conditions, headaches, and injuries. People are deferring treatment for acute injuries and many primary care treatable issues due to lack of coverage.
• ER visits went up by 0.41 from 1.02 per person (up 40%) while expenses went up $120 per person  (up 28%). Individually that marginal ER visit doesn’t seem to be that expensive (about $292), unless you compare it to a Medicaid PCP visit in Portland ($148 CPT 99205).

Prior results from the Oregon Experiment showed that overall hospital admissions went up 2.1% (so if ER admissions stayed the same, these are presumably elective admissions or maybe the unicorn-like ‘direct admission’), prescriptions drug use increased 15%, outpatient office visits increased 35%, screenings increased, all while financial strain and depression decreased by about 9%.

On the other hand, clinical outcomes like HbA1C, cholesterol levels, blood pressure, and mortality were not significantly changed. However, since patients in the Oregon study tended to be fairly healthy and the follow up time relatively short, the power to detect short-term differences in clinical outcomes was extremely low.

What does all this mean?

People receiving Medicaid have pent-up health needs for which they will seek treatment. Primary care visits might go up 35%, ER visits might go up 40%, and health care costs will certainly go up. But low-income people will have less financial difficulties resulting from medical bills. Whether or not actual health outcomes will change will likely take far longer than 18 months to figure out.

ERs stand to benefit if visits go up and the cost of providing that increased care is adequately covered by Medicaid reimbursements.

As for everybody else, this study doesn’t really provide a direct answer on whether securing your neighbor’s health coverage alters your own health at all. But the IOM would suggest that it does.

Using the Oregon Studies as a guide, how will we gauge the success (or failure) of the Affordable Care Act’s Medicaid expansion? First of all, enrollment numbers. Only 30% of those who won the lottery actually received Medicaid. The ACA has to beat those uptake numbers if we are to consider it successful. Secondly, increased access to services. I would recommend we check process measures like recommended screening tests to determine if new Medicaid enrollees are getting the services we want them to receive. Looking at ER visits, hospitalizations, and primary care visits don’t actually tell us about what happens when the patient and doctor meet. Lastly, debt collections. If Medicaid is successful, fewer people will have collections due to medical bills. It would be nice to see fewer bankruptcies too (but the science does not yet support that).

That’s what the Oregon Health Study has told us so far. No more, no less.

However, many political pundits will inappropriately use certain statistics to grade the Affordable Care Act in the coming months.  Nobody should expect Medicaid to fix clinical outcomes by the time the 2016 presidential campaigns are in full swing. Health care costs won’t magically go down. Science says neither of those things will happen. So, political pundits, don’t even go there. Instead, let’s just see if Medicaid is doing what Medicaid is supposed to do – providing coverage, increasing access, and protecting poor people from further financial ruin.

by

Cedric Dark, MD, MPH