Published on

Joshua Russell, MD, MSc, FCUCM, FACEP

I didn’t want to send her to the ER, but what else could I do? Mrs. C was a gaunt 72-year-old Chinese-American woman who had never been to our clinic before. Her son carried her in at 7:56 pm like a bride across the threshold.

              “She cannot walk,” he said, startling the MA sitting at the front desk.

              “Has she been seen here before?” our MA asked with widened eyes.

They both stared back blankly. She’d already run into the limits of the pair’s English. Astutely, she quickly stood and ushered them into the nearest exam room. I followed behind, mentally running through a differential as I hurried across the clinic. I was able to come up with stroke, ischemic limb, septic joint, and fracture before greeting them.

The MA began to collect Mrs. C’s vital signs as I attempted a history. I was able to glean that she’d had pain in her lower leg and ankle that started sometime that morning. Why had they waited until 4 minutes before closing to come in? My wondering wasn’t as much due to frustration about staying late, but rather because our x-ray technician had left 15 minutes earlier to relieve his wife who’d been solo with sick-kid duty at home all day. The clinic was empty at the time and it was less than a half hour before closing. It seemed reasonable in the moment.

My MA shot me a look of desperation belying her end-of-shift fatigue as she held the language tablet up between the patient and her son as they jawed over one another and the interpreter. I soon found out that they spoke a Chinese dialect that was not similar enough to Mandarin or Cantonese for the language operator to discern her story, either. The barriers to providing definitive, or even adequate, care to Mrs. C continued to mount.

I decided we’d reached the tipping point. We couldn’t get an x-ray or even communicate with her and the clinic was now closed. Reluctantly, I explained, through a combination of pantomime, sentence fragments, and Google maps, that her son would have to take her to the emergency department.

In the ED, the patient was given a dose of oral pain medication. An x-ray of her lower leg didn’t show any cause for her pain, and she was discharged with a provisional diagnosis of gout (which it probably was). Nothing heroic was done for her and she ended up being fine.

Was this an “unnecessary” ED referral? Maybe. It depends on who you ask.

A handful of studies examining patients referred from UC to the ED have been published in the emergency medicine literature in recent years. In each of these studies, the authors have developed their own criteria for “unnecessary” or “nonurgent transfers.” The investigators defined aspects of care such as specialist consultation, performance of lab testing or advanced imaging, and hospital admission among appropriate justifications for the ED referrals.

Conversely, patients not receiving such testing or care were deemed inappropriate for referral.1 At first glance, these seem reasonable. The definition is certainly arbitrary, but for research purposes, a line needs to be drawn somewhere. The issue, however, lies in that these criteria only take into account what happens to the patient in the ED, while failing to consider what the circumstances were in the UC leading to the referral.

Consider a bystander calling 911 after witnessing a man collapse on the street. Imagine then that the would-be patient comes to before the paramedics arrive and subsequently reveals that he has vasovagal syncope weekly.

In this scenario, was this an “unnecessary” activation of EMS? Perhaps if you asked the paramedics, they’d be of that opinion. But this is only because the situation had become apparent by the time they’d arrived and it was clear their assistance wasn’t needed after all. But in the absence having the advantage of hindsight, we can all agree the safest decision when witnessing a stranger faint is to call 911. Yes, there will be occasional “false alarms,” but it’s preferable to an alternative society where medics are only dispatched when it’s certain there’s a disaster unfolding.

Similarly, the existing literature regarding UC-to-ED referrals examines necessity and urgency exclusively based on the outcome of care in the ED. According to the definition outlined by Zitek, et al, referring Mrs. C to the ED that evening was “unnecessary.”1 Perhaps it would be better to consider “appropriateness” rather than labeling visits as “unnecessary” or “nonurgent” based on the circumstances at the time of the UC referral. This is the medical equivalent of “Monday morning quarterbacking” and fails to consider the situation faced by the referring UC clinician. Patients, in general, present to UC centers because that’s where they want to be seen; neither UC providers nor patients commonly seek rationale to transition care to the ED. Escalating care is merely a recommendation. Moreover, patients who disagree, for whatever reason, will simply not go to the ED (which, as an aside, is why the label of “referral” is more precise than “transfer” when this occurs).

These studies are problematic, additionally, because they are not only entirely ED-centric, but are also retrospective. The researchers are forced to choose an arbitrary and rigid definition of necessity/urgency which universally fails to capture the complexity of the decision and the myriad of other reasons we refer our patients to EDs. Common justifications for ED referrals which aren’t captured within this definition, for example, include acute mental health concerns, limited UC staffing, patient demands for specialist care, and complex presentations near closing time, as was the case with Mrs. C.

An additional concern with the asymmetric perspective of these publications involves an exaggeration of the scope of this “problem.” When I work shifts in the ED these days, I certainly see more patients who’ve been sent in from UC, but this is expected given the dramatic rise in UC visits over the past few decades.2 It is spurious logic to presume that the reason more patients have been referred from UC centers is that UC providers are referring an excessive proportion of patients to the ED. It is simply an expected occurrence based on escalating UC volumes.

In several recent studies examining UC-to-ED referrals, the authors were nominally examining ED care for referred patients and whether pediatric urgent care centers would refer fewer “nonurgent” patients to the pediatric ED than would full-scope UCs (ie, those that also see adult patients).

While their data did support this to be the case, the more interesting finding was the total number of patients referred from any UC to their pediatric ED over the 11-month study period between the two studies: 349.3,4 Regardless of the perceived necessity for the referrals, this amounts to less than one extra patient per day. In a pediatric ED with roughly 50,000 annual visits,5 even if some UC referrals weren’t captured, it’s difficult to imagine additional volume of this scale is highly consequential, or even perceptible.

As the number of patients seen in UC centers has expanded, it stands to reason that there would be a commensurate increase in the total number of ED referrals. What isn’t clear from these studies examining ED referrals is whether or not the increase in ED referrals is disproportionate—and this is the more important question.

The authors of these papers fail to address, or even speculate, however, on this alternative possibility: UC centers, on average, may very well prevent ED visits. In other words, these researchers have focused only on the numerator and disregarded the denominator—the millions of patients who are seen in UC and not referred to an ED.

Prevention of a certain outcome is inherently difficult to study and requires considerable assumptions. Thankfully, a series of studies in recent years have attempted to empirically quantify urgent care’s effects on ED volume. This body of research offers a distinctly different assessment of the situation. These researchers took a different approach by looking at the changes in ED presentations for low-acuity conditions in local EDs before and after UC openings in the area. The consensus of findings from authors using this approach: nearby UC openings were associated with significant reductions in low-acuity ED visits.6-8 Moreover, other investigators have shown that alternate sites for care, such as primary care offices, are more likely than urgent care centers to refer patients to EDs.9

In other words, the alleged problem of “unnecessary” or “nonurgent” referrals, based on these findings, would likely be much worse in a world without UC.

Where the truth lies is an unsettled debate. The studies which counter the narrative of UC as a contributor to the problem of ED overcrowding are refreshing.  However, there remains a considerable and undeniable bias in much of the emergency medicine community towards inflating the frequency and frivolousness of UC-to-ED referrals. It’s the same bias that paramedics may have against “false alarm” situations like the one previous mentioned, when they arrive on scene and their services are not needed. In both cases, frustrations are understandable. ED providers are already overtaxed with current patient volumes. Consequently, the tendency to seek a worthy scapegoat for this untenable reality is unsurprising. Whether UC deserves blame remains unclear.

We find ourselves faced with competing narratives regarding our utility. Does UC prevent or promote additional ED visits, and what proportion of these referrals are appropriate? These are vital questions for us to answer correctly. So why are we allowing them to be addressed exclusively by EM researchers? To settle the score and defend the value of our services, it is incumbent upon the UC community to take action and conduct UC-based studies on this practice.

The most effective way to do this is by defining who our patients are, who gets escalated to an ED level of care, and how often we are referring patients in appropriate situations. This will offer the added benefit of demonstrating our virtue to the larger medical community through self-policing and striving to ensure we are providing the high-value, resource conscious care we aspire to.

Until we perform UC-based studies evaluating such questions, these referrals will continue to be judged based solely on ED outcomes. This is simply unfair. Just as it only makes sense to assess activation of EMS based on the circumstances that led onlookers to calling 911, ED referrals are most appropriately judged by reviewing the care in our clinic, not the ED chart. This is, after all, the context within which the referral decision is made.

Until such UC-based research is performed, however, our portrayal in the medical literature will remain at the mercy of those focused on their own challenges and who have likely never faced a 7:56 pm “walk-in” like Mrs. C.


  1. Zitek T, Tanone I, Ramos A, et al. Most transfers from urgent care centers to emergency departments are discharged and many are unnecessary. J Emerg Med. 2018;54(6):882-888.
  2. Poon SJ, Schuur JD, Mehrotra A. Trends in visits to acute care venues for treatment of low-acuity conditions in the United States from 2008 to 2015. JAMA Intern Med. 2018;178(10):1342–1349.
  3. Eason M, Clingenpeel J, Vazifedan T, et al. Transfers From a pediatric urgent care to an academic pediatric emergency department. Pediatr Emerg Care. 2022;38(2):e507-e510.
  4. McCarthy JL, Clingenpeel JM, Perkins AM, Eason MK. Urgent care transfers to an academic pediatric emergency department. Pediatr Emerg Care. 2018;34(3):185-188.
  5. Children’s Hospital of the King’s Daughters. Accessed November 6, 2022.
  6. Allen L, Cummings JR, Hockenberry JM. The impact of urgent care centers on nonemergent emergency department visits. Health Serv Res. 2021;56(4):721-730.
  7. Carlson LC, Raja AS, Dworkis DA, et al. Impact of urgent care openings on emergency department visits to two academic medical centers within an integrated health care system. Ann Emerg Med. 2020;75(3):382-391.
  8. Llovera I, Loscalzo K, Gao J, et al. Increased access to urgent care centers decreases low acuity diagnoses in a nearby hospital emergency department. Am J Emerg Med. 2019;37(3):486-488.
  9. Gorodetzer R, Alpert EA, Orr Z, et al. Lessons learned from an evaluation of referrals to the emergency department. Isr J Health Policy Res. 2020;9(1):18.
Which ED Referrals Are Appropriate? The Problem of Perspective
Joshua Russell, MD

Joshua Russell, MD, MSc, FAAEM, FACEP

Quality and Provider Education, Legacy - GoHealth Urgent Care, is affiliated with the University of Chicago Medical Center in Vancouver, WA, and is Editor-in-Chief of JUCM.