Lee A. Resnick, MD, FAAFP
Much has been written of late about use of emergency services by patients covered by Medicaid. For some time, consensus has been that Medicaid patients overuse emergency services for non-emergencies. The emergency department (ED), it was thought, served as the de facto primary care physician for this because of problems with access and lack of pricing pressures to deter use. Until recently, supportive data were lacking and the notion of overuse was based primarily on physician and hospital experience.
To better define actual usage rates, the Center for Studying Health System Change, with funding from the Robert Wood Johnson Foundation, set out to provide better definition for the data. Unfortunately, the study relied on pre-existing data from 2008 National Hospital Ambulatory Medical Care Survey Emergency Department (NHAMCS-ED), leaving us with the same questions as before the study. The key trouble with these data is that they are far too interpretive to be of any use from a policy perspective.
In the report, ED visits are divided into five categories as determined by a triage nurse: (1) emergent – patient needs to be seen immediately or within 15 minutes of arrival; (2) urgent – 15 to 60 minutes; (3) semi-urgent – 1 to 2 hours; (4) nonurgent – 2 to 24 hours; and (5) no triage or unknown. A full 75% of Medicaid patients received triage scores of 1 to 3. However, very little is written about how such a subjective yet definitive categorization will affect the interpretive impact of the findings. In fact, despite the utter lack of visibility behind the triage score and its misleading terminology, several policy groups and media outlets have carelessly presented the study findings as a statement of fact.
So, I have to ask, “Is this what the study really showed?” The authors are all really smart people with deep knowledge of statistics and knowledge, so they must be on to something … right? Just to be sure, I did some of my own investigative work and what it reveals is less than encouraging.
- The study uses common lay terms like “urgent” and “emergent” to define a very specific category with meaning only for purposes of this study. It is tempting but incorrect to assume that anyone who needed to be seen in 15 to 60 minutes had a “proper” ED visit. Much the same could be said about the 1- to 2-hour category. By that definition, at least 75% of all ED visits are “proper.” However, closer inspection would reveal that a large percentage of such patients could be seen elsewhere. One also might make the reverse error and draw the conclusion that all the urgent, semi-urgent, and non-urgent visits are, by definition, “non-emergent” and therefore, could be interpreted as “appropriate for alternative care” settings such as urgent care clinics. By that definition, only 12% of all ED visits are true emergencies (a very different story indeed).
- The study’s main defining data point is almost entirely subjective, dependent on the triage nurse’s interpretation with very few objective components. Nowhere is it defined “how” the triage nurse determines the amount of time within which a patient needs medical attention.
Policy groups and the media made quick work of drawing broad conclusions about the definition of “routine” care. I’m not sure what that means, either. Regardless, the vast majority of complaints seen in the study were for non-emergencies like fever, sore throat, respiratory infections, headaches, rashes, minor injuries, and minor pain. The vast majority of these patients were “stable,” and while their conditions required some level of urgent or semi-urgent evaluation, they were by no means emergencies.
Perhaps, then, the study’s most glaring flaw is that it failed to ask the right question. That is, how many of the 75% to 90% so-visits called “non-emergent” could and/or should have taken place in an alternative setting such as an urgent care clinic? Until we focus our inquiry on the right question we will be left with nothing but literary fodder for pundits and interest groups.
In my next column, I will propose a method to quantify the potential impact of urgent care services on emergency department utilization.
