The Dismal Science

Does Expanding Health Care Coverage Reduce ER Visits?

On the contrary: A new study suggests it increases visits to the emergency room—significantly.

Coney Island Hospital October 5, 2002
When prices go down, demand goes up, so insurance naturally leads to more health care use overall, including emergency care.

Photo by Mario Tama/Getty Images

A few weeks ago, I was asked to write a comment to accompany a study released Thursday in Science, which shows that Medicaid access increases emergency room visits by more than 40 percent. To get a sense of how medical practitioners would respond to the findings, I asked my wife, a primary-care physician at Bellevue Hospital, whether she thought Medicaid would increase or decrease ER use by the kinds of low-income individuals who constitute her patient population. Her reaction: “I refer people to the emergency room all the time. Of course it’ll go up.”

Even if my wife and her fellow primary-care providers aren’t shocked to find their everyday experience validated by a large-scale experiment, many others will be. Public officials from Health and Human Services Secretary Kathleen Sebelius to state governors in Michigan and Ohio have cited a reduction in traffic to overstressed emergency departments as a rationale for insurance expansion. They’d do well to change their talking points.

The clear and compelling evidence that ER usage increases with Medicaid coverage is the latest finding to come out of the landmark Oregon Health Insurance Experiment, which randomly selected low-income uninsured adults in Oregon to receive health insurance through the state’s expansion of Medicaid. This randomization is critical to understanding the effects of insurance, since the insured and uninsured are different in so many ways that it’s otherwise nearly impossible to make an apples-to-apples comparison.

Thursday’s Science study has the potential to be the most influential of the trio of papers that have come out of the OHIE so far, both because of the magnitude of the impact (a 40 percent increase is a very large effect) and also because reasonable people (my wife and Kathleen Sebelius, for example) disagree on the likely effect of Medicaid expansion.

To understand why this disagreement exists, it’s useful to think of health coverage as an across-the-board price cut for all medical services. When prices go down, demand goes up, so insurance naturally leads to more health care use overall, including emergency care. A single visit to the emergency room for an uninsured patient can easily generate a bill that runs into the thousands of dollars, which is why the uninsured typically try self-medicating first, or just wait and hope the chest pain is heartburn and not cardiac arrest.

Advocates for health care expansion reason that the newly insured will seek out health care more often, but will choose a primary-care physician over the ER. One reason the uninsured tend to go to the ER when they do finally seek treatment is that it’s their only option. You need to present an insurance card to get service at most physicians’ offices, whereas the ER is legally bound to attend to any case that comes through the door. Once insured, a patient does have every reason to see a primary-care physician and skip the ER: As every American knows, you don’t go to the ER for a good time. You go there expecting to sit under a glaring fluorescent light on a hard-backed plastic seat for four hours (if you’re lucky), surrounded by a combination of crying babies and hacking coughs. If you’ve got the option of making an appointment at your doctor’s office instead, you take it.

But according to the Oregon study, the newly insured still choose the ER over the doctor’s office. What’s more, the increase in ER use documented by the study comes in large part from patients with ailments like cuts and sprains—problems that could have been managed through a primary-care physician or by an urgent care clinic (both of which are covered by Medicaid). Patients aren’t, in fact, substituting primary care for the ER to the extent that many insurance advocates have hoped they would.

This doesn’t come as a surprise to my wife, whose overscheduled practice is booked up with regular, predictable appointments, which doesn’t leave much slack to handle these urgent (but nonemergency) cases. She’s not alone in this: a sizeable fraction of ER visitors (including those with nonemergency conditions) go there because they were told to do so by a health care professional. Some of these referrals are for things like chest pain that might just be heartburn, but need to be checked out in a timely fashion. For other cases—the deep cuts and bad sprains—the ER has become the de facto site of treatment for many Americans. They’re going to the ER not because they’re uninsured, not because they don’t know any better, but because an overstressed primary health care system is sending them there. (Hospital administrators have long since accepted this fact. At Bellevue Hospital, where my wife works, the ER is set up to handle urgent, non-emergency cases, and has a fast-track to route them to appropriate care.)

In my policy comment accompanying the current study, I express hope that these results will not simply lead to further bickering and greater divisiveness on the question of whether we should expand insurance coverage. Both sides of the Obamacare debate took the first two studies from the Oregon experiment as evidence in support of their opposing causes, and it’s easy to imagine the same response to the results reported in Thursday’s study. Health care proponents will point to the findings as evidence that ERs are now serving the suppressed medical needs of the uninsured, people who were previously scared away from getting any care for fear of financial ruin. Opponents will see yet more evidence that when you make health care too cheap (i.e., free), people use too much of it.

More constructively, we could take the study as a compelling indication that ER usage—which we can all agree isn’t the most efficient way of caring for scrapes and sprains—is likely to go up, not down, as the Affordable Care Act leads to wider insurance coverage, and prepare for that outcome. Regardless of how you feel about the ACA, this study should demand a greater sense of urgency for such preparations.

Thankfully, the problem of providing too much health care through emergency rooms is already well-appreciated by the medical community, and many health care entrepreneurs are already experimenting with ways of caring for patients more efficiently and at lower cost (to the patient and the system): whether through a neighborhood walk-in clinic, an urgent care setting, or a doctor’s office. Better still, we could aspire to provide preventive care with an eye toward individuals’ longer-term best health interests, hopefully keeping more cases out of the ER and hospital inpatient wards in the future. All sides should be able to agree that it’s time to double down on our support for these efforts.