Uninsured are not the problem in ED overcrowding

Whenever you are having a debate -- particularly a policy debate -- it is always important to check your premises. That is why I found this article in the Journal of the American Medical Association refreshing.

Emergency Department utilization is clearly on the rise in the US, and this rise in use is leading to longer wait times and diminished quality of care. One assumption that the cause of this problem are the uninsured, i.e. the uninsured are using the ED as an alternative to primary care and causing overcrowding.

However, Newton et al., in what will likely be a provocative article, reject this assumption. After surveying the literature with respect to ED and the uninsured, they found the uninsured are not using the ED at a disproportionate rate to the insured.

Newton et al. scoured the literature with respect to ED utilization in order to identify the assumptions -- which they define as assertions taken as fact with out citation or supporting data. They identified the 10 most prominent assumptions and then went back and check whether these assertions were supported by the data. In some cases, they were and in some cases they weren't.

Here are the two most interesting findings:

Assumption 1

Assumption. Uninsured patients use the ED for nonurgent/nonemergent/primary care-type/"inappropriate" care ("[uninsured] patients realize no matter what may be their complaint, even if it is not an emergency, they can receive care at any local ED for free").

Assessment: Not Clearly Supported by Current Data. While this is the most common assumption, occurring in more than 20% of all articles reviewed, it also is the most difficult to define. What does "nonemergent" mean? Who decides what is or is not an emergency? Emergency departments triage patients based on the immediacy with which patients should be seen. Patients in the lowest triage category (ie, those who should be treated within 2-24 hours) are often classified by insurers and researchers as requiring nonurgent care, even though many nonurgent complaints (eg, sprains, fractures, lacerations) may be most appropriately cared for in the ED.

National evidence suggests that uninsured patients are minimally more likely to make nonurgent visits, based on the immediacy-of-care definition from ED triage practice. The National Hospital Ambulatory Medical Care Survey found that ED visits classified as nonurgent increased from 10% to 14% of visits from 1997 to 2005 overall, and from 11% to 16.7% for uninsured patients. A 2002 analysis of the Medical Expenditure Panel Survey, with urgency defined by whether the patient considered their visit to the ED an emergency, found no relationship between insurance status and urgency of need on presentation to the ED for patients who had a primary care physician. Several studies have found that uninsured patients are no more likely to make a nonurgent visit than those with private insurance.

The assumption that uninsured patients present for less urgent care comes largely from a single 2003 study that examined billing and insurance data from more than 150 000 visits to a single urban, academic ED and that found that uninsured patients were half as likely to have received the highest-acuity care while in the ED. The authors of that article cautioned, however, that "the magnitude of most differences noted was not large and may not reflect important differences in health care need or ED use based on insurance." The Science Citation Index/ISI Web of Science records only 24 direct citations to that article, but when we followed the chains of citation in other articles, that article was the common source document for this assumption when any citations were given.

As opposed to seeking care primarily for nonurgent or primary care visits, evidence exists that uninsured patients are underrepresented in the ED for primary care -- type visits compared with their percentage in the population; this may be owing to unwillingness to seek ED care, given its cost. (Citations removed. Emphasis mine.)

And...

Assumption 4

Assumption. Uninsured patients are a leading cause of ED crowding ("The ED is used as a primary care provider for the uninsured, which adds to overcrowding").

Assessment: Not Clearly Supported by Current Data. Emergency departments across the United States have been dealing with increasing crowding for almost 2 decades and increasingly are struggling with overcrowded conditions. However, Europe, Canada, and Australia are also struggling with ED crowding, despite having universal health care systems.

The etiology of crowded EDs is multifactorial and includes a lack of staffed inpatient beds, hospital and ED closings, increased ED use by all patients, and an aging population with increasing prevalence of chronic illnesses. On a national level, 75% of the increase in ED use over the last decade is attributed to increased use per person, mostly by insured patients (from 35 visits/100 population per year to 39 visits/100 population per year); the remaining amount is predominantly due to an increase in population size.

Weber and Showstack showed that insured patients accounted for 84.8% of all ED visits, a rate that remained stable from 1996 to 2004. These rates have been supported by national data; other literature exploring demographics of ED patients report similar percentages. These rates mirror the proportion of insured and uninsured patients in the nation and suggest that neither group uses the ED disproportionately.

While uninsured patients are not a major source of ED crowding on a national level, some hospitals most likely to be crowded are safety-net hospitals in low-income or low-access areas where a large percentage of the population depends on the ED for care. In safety-net hospitals serving vulnerable populations, inadequate access to primary care for patients with public insurance as well as for those with no insurance contributes to increased ED use. For these hospitals, a small increase in the number of ED visits by uninsured patients can greatly increase crowded ED conditions. (Emphasis mine. Citations removed.)

To be fair, the authors also confirm that the uninsured use the ED due to lack of primary care access and that it is more expensive to treat them than elsewhere. What the authors dispute is that the uninsured use the ED disproportionately for non-emergent reasons and that they are the cause of overcrowding (overall...they acknowledge that they may be the cause in some low-income, urban settings).

Attacking these two assumptions is indeed provocative because you hear them all the time in debates about covering the uninsured. An argument often made in these debates is that by failing to provide adequate primary care in the form of health insurance, we are shifting the burden of care onto the ED. Whereas an insured person might call up their primary care doc about a cold, the uninsured person will show up at the ED. If we provided more health insurance, people would be less inclined to go to the ED for often spurious reasons.

There are many compelling reasons why we need wider health insurance coverage. EDs are indeed paying for the uninsured, and this is one reason that hospitals are closing. But it would appear that giving these patients insurance in and of itself will not solve the problem of ED overcrowding. (This point is emphasized by the observation that ED overcrowding is becoming an issue in nations that have universal health care systems as well.)

Instead, I would argue that you can have all the insurance you want, but it won't make a lick of difference to ED crowding if there are no primary care alternatives are available. Likewise, policies attempting to limit the use of the ED by uninsured patients or attempting to divert them to non-emergency settings will only work if there is care to be had. This article comes to a similar conclusion:

Policies designed to address ED crowding by blocking or creating barriers to ED access for uninsured patients are unlikely to be effective, because little evidence exists that uninsured patients are a large proportion of the problem. Policies that attempt to redirect patients requiring nonurgent (by whatever criteria are used to define nonurgent) care to primary care sources are unlikely to succeed unless those sites are readily accessible. If patients --including privately insured, publicly insured, and uninsured patients -- are unable to find primary care clinicians who accept new patients or accept insurance or cash payments; if patients are forced to wait weeks for an appointment; if the hours or location of primary care make it inaccessible; or if patients perceive the care to be substandard compared with care received in the ED, they will continue to come to the ED. (Emphasis mine. Citations removed.)

This article is a compelling evidence for the idea that insurance coverage and health care access -- while similar -- are not the same thing. You cannot simply will access into existence by providing everyone with insurance.

Rather, I think the solution to ED crowding is that we need more -- many more -- primary care providers. Only an excess in the supply of primary care will bring the prices down and increase access for both the insured and uninsured public. How to do that? Programs to forgive the loans of medical students entering these fields would be a good start. Also, trying to bring reimbursement to parity with specialist fields would make this field more attractive.

But the emphasis in the debate about ED overcrowding should be more on the supply of health care rather than the supply of insurance.

Hat-tip: USA Today and SEA