Applying What Works to Reduce Non-Urgent Emergency Department Use


(The RAND Blog)

emergency department sign

by Lori Uscher-Pines

May 22, 2013

The ER is for emergencies — or at least it should be. Nearly everyone agrees on that point. But when it comes to implementing policies to reduce non-urgent use of ERs, opinions quickly diverge.

For the past few years, Washington State has been ground zero for this fight. In 2011, the state's health care authority announced its intention to stop paying for emergency department (ED) visits by Medicaid beneficiaries “when those visits are not necessary for that place of service.” Because the plan was based on a retrospective determination of what was — or was not — an emergency, the proposal faced strong opposition from a range of health care stakeholders. Ultimately, a compromise was worked out.

It appears that this alternative plan is bearing fruit. According to a recently issued report by the Washington State Health Care Authority, a combination of interventions including patient education on what constitutes an emergency (i.e., a sore throat is probably not) and alternatives for after-hours care, are beginning to produce results, including a 23 percent reduction in ED visits by frequent users enrolled in Medicaid.

Washington State is not the only entity interested in redirecting non-urgent ED patients to less costly settings. Other states and private payers are too. Several months ago, Anthem Blue Cross/Blue Shield of Virginia reported the results of a multi-pronged strategy that included educational outreach about appropriate ED use and higher patient copays for ED visits. Preliminary data suggest that such interventions can change care-seeking behavior. How much money these efforts actually save, and whether redirecting this care is good or bad for patients, is less clear.

One of the most challenging aspects of this issue is how little we know about why patients come to EDs with apparently non-urgent problems. At least one factor is clear — when a community's primary care system is inadequate and patients simply cannot get timely care elsewhere, many will turn to EDs. This relationship is so strong, Billings and colleagues at NYU demonstrated that the percentage of ED visits that are “primary care treatable” or “primary care preventable” is a useful metric for assessing a community's access to primary care.

But sometimes patients ignore less costly options even when they are readily available. What we don't know is why they do, and how we can prevent this. Recent RAND research suggests that because the evidence base on what drives non-urgent ED use is so limited, we can't actually predict which interventions will work best over the long run.

When my RAND colleagues and I recently conducted a systematic literature review of reasons why people seek care in EDs for non-urgent conditions, we uncovered major gaps in the existing evidence base. Our goal in doing this work was to inform the design of future interventions to reduce the number of non-urgent ED visits, thereby decreasing costs and curbing unnecessary testing and treatment. What we found was that the existing literature provides little guidance.

First, across the 26 studies we reviewed, the average fraction of all ED visits that were judged to be non-urgent (whether prospectively at triage or retrospectively following ED evaluation) was 37 percent. However, before we rename the ED the “Non-Emergency Department,” it should also be noted that no two studies defined non-urgent visits the same way. This lack of standardization poses a problem because it means researchers do not have a common understanding of a common problem.

Second, because of the heterogeneity and limitations of published research, there is only weak evidence regarding exactly what is driving non-urgent ED use. For example, many articles compared non-urgent and urgent ED visits to identify characteristics that predict inappropriate use; however, my colleagues and I believe that a more valid approach would be to compare patients who use EDs for non-urgent conditions to those who see primary care providers for the same set of conditions. Based on the limited evidence compiled to date, younger patient age, the greater convenience of EDs compared with other ambulatory care options, referral to the ED by a health care provider, and negative perceptions of non-ED care sites play a role in patients' decisions. Other factors, such as cost, were studied too infrequently to support conclusions about their role in either driving or deterring use.

Third, efforts to deter non-urgent ED use can produce unintended consequences. It is well documented that cost-sharing impacts care-seeking behavior. For this reason imposition of steep copayments and deductibles may deter some patients from using the ED for serious or even life-threatening symptoms. Another problem is that even seemingly minor symptoms, such as sore throat or back pain, can be the harbinger of a serious or even life-threatening problem. A third unintended consequence is that advertisements for low-cost alternatives, such as retail clinics, may induce patients who would otherwise recover at home to seek care they don't really need. Finally, efforts to steer patients away from the ED triage desk may do more harm than good if patients cannot get care elsewhere.

Our systematic review suggests that with the current availability of data and information about the drivers of non-urgent ED use, it is challenging to select and implement strategies to reduce such visits. However, innovative interventions that are piloted to address this problem can serve as natural experiments and should capture data on a range of unintended consequences. The results of such interventions can in turn help to improve the evidence base. While we lack clarity as to the precise definition of a “non-urgent” or “unnecessary” ED visit, the concept is a constructive one that should be standardized and explored.

Billings and colleagues' work indicates that non-urgent ED use is a useful barometer of access to primary care. If so, rates of non-urgent ED visits and other conditions that should be amenable to good outpatient care could be used to monitor the impact of ACOs and other measures to improve population health. It is likely that communities with low rates of non-urgent ED use not only have better access to primary care, but patients who are educated about appropriate care seeking and convenient alternatives for acute care.

This week RAND released a major report on the evolving roles of EDs in the nation's health care system. The report asserts that policymakers are disproportionately focused on non-urgent ED use (which it argues is best addressed by improving access to primary care) and not enough on the growing role of EDs as a major entry point for enabling or preventing costly hospital admissions.

Despite the merits of this argument, the Washington State Health Authority and other groups aren't likely to shift their focus from non-urgent ED use anytime soon. Big bills for minor illness care stimulate ongoing interest. Use of the ED for non-urgent conditions that can and should be handled elsewhere is not only costly, but the lack of continuity may not be good for patients.

If we hope to solve this problem, we must do the necessary work to understand, measure, and reduce non-urgent ED use in ways that improve outcomes and enhance value rather than simply turning people away. This is the best way to ensure that patients to get the right care in the right place at the right time.

Lori Uscher-Pines is an associate policy researcher at the nonprofit, nonpartisan RAND Corporation.

Commentary gives RAND researchers a platform to convey insights based on their professional expertise and often on their peer-reviewed research and analysis.