Left Without Being Seen: What Emergency Physicians Can Actually Fix

The triage tech calls the name a third time. Nobody stands up. The chair is empty, and the triage note says abdominal pain since last night. That patient is now somewhere in the parking garage with no exam, no labs, and no one who owns the next step.

Left without being seen (LWBS) is a clinical event, and the front half of it sits inside the emergency physician's reach. Speed up the first clinician contact, put a physician at the front door during the hours people leave, and tell the waiting room what is happening.

Who leaves, and what happens to them

The patients who leave are not a random slice of the waiting room. A 2026 scoping review of 94 publications found that LWBS mostly affects low-acuity, socioeconomically vulnerable adults, and that rates track waiting time, crowding, and time of day [1]. In a 10-ED integrated system that saw 1,474,395 adult visits from 2019 through 2022, 1.2% left without being seen. Those patients were younger, more often Hispanic, and more often needed an interpreter [2].

Low acuity at triage is a screening result, not a diagnosis. That is why the outcome data matter.

The outcome evidence is mixed. Here it is plainly. In that same 10-ED cohort, compared with patients who were seen and discharged, LWBS patients had higher adjusted odds of a return visit within 72 hours (aOR 2.56), ICU admission within 30 days (aOR 1.35), and death within 30 days (aOR 2.59) [2]. A prospective study of 107,119 adult patients at Iceland's capital-area adult ED found a higher 30-day return rate after leaving without being seen (HR 4.84) but no increase in hospitalization (HR 1.09) or death (HR 0.84) [3]. A single health system can count only the returns it sees. The honest summary: these patients come back more often, and the mortality signal depends on the setting and the comparator. A department that never evaluated them cannot claim they were all minor.

Why patients walk out

Ask them. In a Canadian telephone survey of 1,508 LWBS patients matched to 1,504 who stayed, 79% named the long wait as the reason they left. Shorter waits (66%) and more information about the wait (41%) were the top things that would have kept them [4]. LWBS patients were also more likely to seek further care (63% vs 28%), and they sought it sooner [4].

A French ED surveyed its LWBS patients within 7 days. Of those who answered, 74% suggested a change that would have helped them wait longer, and those changes sorted into two groups: better communication and a more comfortable waiting room [5]. In a Pennsylvania ED survey, patients said they wanted information, early tests, and pain relief while they waited [6].

The clock matters. What the patient is told about the clock matters on its own. A waiting room in silence reads as abandonment, and patients with a shift to work, a child to collect, or a ride that will not wait act on that reading.

Front-end fixes with evidence behind them

Physician in triage. This has the best evidence of any front-end change. A meta-analysis of 12 US studies covering 329,340 patients found that a triage liaison model staffed by attending physicians lowered the risk of LWBS (RR 0.62, 95% CI 0.54 to 0.71). Its effect on length of stay was too inconsistent to pool [7]. A UK systematic review of senior doctor triage pooled two Canadian randomized trials and found lower LWBS (RR 0.79, 95% CI 0.66 to 0.94) with no change in adverse events, although 12 of its 25 included studies were weak before-and-after designs [8].

The randomized data are more modest than the before-and-after reports. In an Ontario cluster randomized trial of a physician-nurse team at triage on weekday shifts from 08:00 to 14:30, LWBS was 1.5% vs 2.2% (p = 0.06), while time to physician assessment fell from a median of 81 to 55 minutes [9].

Nurse-initiated protocols. A systematic review of 26 studies (9,144 patients) found that nurse-initiated interventions shortened time to analgesia and time to treatment for acute respiratory distress, and improved pain relief [10]. The review did not show an LWBS effect. It does answer what patients told the surveys they wanted during the wait: tests started and pain treated. Write the protocols narrowly and audit them.

Split flow and vertical care. Keep upright patients upright and out of beds they do not need. We covered the guardrails for AI-assisted versions of this in vertical ED flow with clinicians in charge. Put a named owner on reassessment for every patient waiting outside a bed.

Boarding sets the ceiling

Front-end design has a ceiling, and boarding sets it. In a single urban ED, a machine-learning review of four years of daily data found that triage-to-bed time, boarding time, and door-to-triage time were the strongest drivers of daily LWBS. After the department added rapid triage, direct bedding, and boarding reduction, the mean daily LWBS rate fell from 4.82% to 1.93% [11]. It was an observational before-and-after study at one site. The direction is still the one every ED director recognizes.

The boarding levers sit mostly outside the department, on the inpatient side of the house. An observation unit is one lever the ED can help design, and we laid out how in fixing boarding with an ED observation unit. Software does not create inpatient beds, which is the argument in AI won't fix your boarding crisis. When you take LWBS numbers to the C-suite, bring boarding hours on the same slide.

When a patient leaves anyway

LWBS and leaving against medical advice are different events. A patient who leaves against medical advice has been evaluated and counseled. An LWBS patient has had neither an evaluation nor a risk discussion. Do not chart one as the other. Record what is true: arrival time, triage acuity and vital signs, chief complaint, each attempt to locate the patient, and the time the encounter closed.

Consider a nurse callback for a defined high-risk group, such as chest pain, shortness of breath, abnormal triage vital signs, or a febrile infant. We found no trial that measures what a callback does to outcomes, so treat it as outreach with unproven benefit. Set the criteria, the script, the time window, and the documentation standard with your institution's risk management team.

Key Takeaways

• Pull last quarter's LWBS cases by arrival hour and triage acuity. Staff a physician in triage for that window first; the pooled US data show a relative risk of 0.62 for LWBS.
• Report LWBS next to boarding hours. In one ED, a package that included boarding reduction cut daily LWBS from 4.82% to 1.93%.
• Give the waiting room information about the wait. In a matched survey, 41% of patients who left said more information would have helped them stay.
• Define LWBS documentation and any callback criteria in writing with risk management before you need them.

FAQ

What is a good LWBS rate for an emergency department?

There is no single safe number. A 10-ED US system reported 1.2% from 2019 through 2022 [2], and published rates vary widely by setting [1]. Compare your department against itself by hour, acuity, and day of week, and against peers with a similar volume and boarding burden.

Does a physician in triage reduce left without being seen rates?

Yes, in most studies. A meta-analysis of 12 US studies found a relative risk of 0.62 with attending physicians at triage [7]. One cluster randomized trial showed a smaller drop that did not reach significance [9].

Are patients who leave without being seen at higher risk of death?

The data conflict. A 10-ED US cohort found higher 30-day mortality than in patients who were seen and discharged [2]. An Icelandic study found more return visits but no increase in 30-day death [3].

Dr. Chet's Take

I have worked more than 25 years as an emergency department attending, and LWBS is the one metric that counts patients we never met. Every other quality measure describes care we delivered. This one describes care we did not. The article gets the core argument right. The front end belongs to us, and the evidence for putting a physician at the door is better than most of the operational evidence we act on every week. I expected the outcome data to settle the argument about how sick these patients are. It does not, and I would rather say so than pretend otherwise. The case for fixing LWBS does not need a mortality number. It needs the plain fact that a person asked for an emergency physician and left without one.

That being said, the article treats LWBS mostly as a flow problem, and it is also an access problem. The patients who walk out are the ones with the least slack in their lives: the hourly worker who loses the shift if he stays, the mother with two kids in the car, the retiree who lives alone and drove herself in. A front-end model that runs only on weekday mornings will improve the dashboard and miss them. The honest answer is that I have not seen a callback program shown to change outcomes, and I would not present one to a hospital board as a safety intervention. It is good outreach and good documentation. Seeing the patient is still the intervention.

If you are leading an emergency department, ask for your LWBS data broken out by language need, insurance status, and the overnight hours, not only by acuity. Then ask your night-shift nurses what the waiting room hears in the second hour of a wait, because that is the part of this problem no dashboard shows you. Staff the hours when people leave, measure again in ninety days, and put both numbers in front of the people who own the inpatient beds.

— Chester Shermer, MD, FACEP | Emergency Medicine, 25+ Years Clinical Experience | State Surgeon

Continue Your Training

If you're an emergency physician (or any clinician treating patients daily) trying to understand how AI will actually impact your clinical practice — not just the hype — I put together a free practical guide. You can download the AI in EM Survival Guide here.

Browse All Courses at Global MedOps Command

Relevant Reading on Global MedOps Command:

How to Avoid Becoming an AI Casualty — Dr. Shermer's guide to navigating AI tools in clinical and operational settings without compromising judgment or patient outcomes.

Emergency Department Efficiency Playbook — Practical systems for throughput, triage optimization, and operational efficiency.

Read more from Dr. Shermer on Medium

Connect with Dr. Shermer: LinkedIn — Chester "Chet" Shermer, MD, FACEP.

References

1. Ardila Jara AR, Barbosa Malagon C, Castillo Moreno LD, et al. Left Without Being Seen: A Scoping Review of a Silent Crisis in Emergency Departments. Open Access Emerg Med. 2026;18:604937. doi:10.2147/OAEM.S604937
2. Joseph JW, Bukhman AK, Baymon DE, et al. Mortality and Return Visit Frequency Among Emergency Department Patients Who Leave Without Being Seen at a Regional Health Care System. Jt Comm J Qual Patient Saf. 2025;51(7-8):466-473. PubMed 40180867
3. Geirsson OP, Gunnarsdottir OS, Baldursson J, et al. Risk of repeat visits, hospitalisation and death after uncompleted and completed visits to the emergency department: a prospective observation study. Emerg Med J. 2013;30(8):662-668. PubMed 22983976
4. Fraser J, Atkinson P, Gedmintas A, et al. A comparative study of patient characteristics, opinions, and outcomes, for patients who leave the emergency department before medical assessment. CJEM. 2017;19(5):347-354. PubMed 27692013
5. Ibanez G, Guerin L, Simon N. Which improvements could prevent the departure of the left-without-being-seen patients? Emerg Med J. 2011;28(11):945-947. PubMed 20961927
6. Johnson M, Myers S, Wineholt J, et al. Patients who leave the emergency department without being seen. J Emerg Nurs. 2009;35(2):105-108. PubMed 19285171
7. Benabbas R, Shah R, Zonnoor B, et al. Impact of triage liaison provider on emergency department throughput: A systematic review and meta-analysis. Am J Emerg Med. 2020;38(8):1662-1670. PubMed 32505473
8. Abdulwahid MA, Booth A, Kuczawski M, et al. The impact of senior doctor assessment at triage on emergency department performance measures: systematic review and meta-analysis of comparative studies. Emerg Med J. 2016;33(7):504-513. PubMed 26183598
9. Cheng I, Lee J, Mittmann N, et al. Implementing wait-time reductions under Ontario government benchmarks (Pay-for-Results): a cluster randomized trial of the effect of a Physician-Nurse Supplementary Triage Assistance team (MDRNSTAT) on emergency department patient wait times. BMC Emerg Med. 2013;13:17. PubMed 24207160
10. Burgess L, Kynoch K, Theobald K, et al. The effectiveness of nurse-initiated interventions in the Emergency Department: A systematic review. Australas Emerg Care. 2021;24(4):248-254. PubMed 33727062
11. Kirby JJ, Knowles HC, Asad S, et al. Maximizing efficiency in emergency care: triple interventions to minimize left without being seen: An observational study. Medicine (Baltimore). 2024;103(52):e40763. PubMed 39969359

Related simulation training: EM-Sim · Books by Dr. Shermer