Why is ED Holding Still an Issue?

1 Comment

altBoarding admitted patients in the ED is as bad for patient care as it is for the hospital’s bottom line. So why aren’t more CEOs bringing this pervasive problem to an end?

Boarding admitted patients in the ED is as bad for patient care as it is for the hospital’s bottom line. So why aren’t more CEOs bringing this pervasive problem to an end?


I think if you ask most emergency physicians who work in dysfunctional emergency departments (many) what is the greatest source of their angst, they would say it is the holding of admitted patients.

The literature on this subject is very extensive (reflecting how serious a problem holding is). All manner of solutions have been suggested yet holding continues to cripple many EDs.

Holding of ED patients is a widespread problem and is likely to get worse when the EDs are flooded by newly insured individuals as the result of the Affordable Care Act coupled with the graying of America and the transition of the baby boomers into large consumers of healthcare.
Why does this problem still exist? Is holding an insoluble problem? Hard to conceive that it is. We have solved all sorts of more difficult problems than ED holding. How about smart phones, air travel, robotic surgery, cars that drive themselves, solar power airplanes, man on the moon. So what’s the big deal about ED holding?


The case to fix ED holding is compelling. Holding patient in the ED consumes nurse and physician time and precludes the ability to see more patients due to the blocking of an ED bed. Decreasing the ability to see and treat patients is costly (assuming there are patients waiting to be seen).

Conservatively, every new patient who is discharged (representing about 80% of patients in most EDs), generates about $600 — $100 for the physician and $500 for the hospital. See 2.5 patients per hour and it is $1500.  Hold a patient for six hours and it’s $9,000 in hard cash (not billings but actual collections). Hold multiple patients and there’s a lot more patients who can’t be efficiently treated in the ED (better hope there is no urgent care center in the area). If you can’t see the patients someone else will be happy to. There are about 9,000 urgent care centers in the country and the number is rising rapidly.

Patients held in the ED generate essentially no additional money after the work-up and initial treatment are over and the patient is just tying up an ED bed – but they do take up nursing and to a lesser extent, physician time. That is assuming the patient is stable. If the patient being held is an ICU patient there is likely substantially more nursing and physician work.

Here’s a paper focusing on the costs associated with holding admitted patients in the ED. They specifically focused on the lost opportunity cost associated with the inability to see other patients when admitted patients are held more than two hours. The bottom line – they calculated that at the study hospital boarding prevented the treatment of another 8.7 patients per day and that cost the hospital $4 million a year. Clearly there are a lot of challengeable assumptions in this study but even if the real loss is $2 million, many hospitals don’t even make this much in a year in total.


Falvo, T., et al, Acad Emerg Med 14(4):332, April 2007

BACKGROUND: Increased crowding in the ED has been reported to be attributable to the practice of boarding admitted patients until an in-patient bed can be assigned. These occupied ED beds represent a loss of opportunity to treat additional ED patients. The extent to which this practice affects a hospital’s profitability might not be fully appreciated.

METHODS: The authors, from WellSpan Health System in York, PA, and Johns Hopkins University, evaluated data from 62,588 ED visits to a 450-bed nonprofit community teaching hospital (York Hospital) during fiscal year 2005 and calculated the total number of hours in which admitted patients were boarded in the ED. This information was used to determine ED treatment bed occupancy and the revenue potential that were lost to inpatient holding.

RESULTS: The 120-minute benchmark for the interval between the decision to admit and the time of transfer out of the ED was exceeded by 30% of all the patients who were admitted, accounting for a total of 10,397 hours in the ED during the fiscal year. If these patients had not been “boarded” in the ED for this length of time, it was estimated that an additional 3,175 patients could have been seen (averaging 8.7 additional patients per day), generating a median net revenue of nearly $4 million (about $3 million from additionally admitted patients [$5,432 per patient] and about $1 million from discharged patients [$384 per patient including professional fees]), which would have represented a 13% increase in the ED net revenue during the fiscal year.

CONCLUSIONS: This study illustrates the financial consequences of the practice of “boarding” patients in the ED pending the availability of inpatient beds. The authors suggest that, as in any other industry, healthcare facilities cannot prosper if resources are used inefficiently. 58 ref. ([email protected]) 8/07 – #11

I’m trying to make the case that holding patients in the ED is a costly proposition when held patients preclude the opportunity to see new patients. There is a more compelling reason to move patients out of the ED. Studies have shown patients are likely to do better when they are admitted sooner. This is particularly true for ICU patients. Here are a few papers that make the case. We’ll not even bother with the raft of papers saying that ED care suffers on multiple fronts when the ED is crowded (which is usually / often the result of holding ED admits) – decreased time to pain management, increased adverse outcomes in acute coronary syndrome patients, delays in provision of antibiotics, decreased quality of care for children, increase numbers of patients who leave without being seen, etc.

The following paper makes the point that hospital length of stay increased by 0.8 days when patients were held in the ED more than eight hours. Four other studies support a one day increase in the average length of stay when admitted patients are held in the ED (if the patients are Medicare or case rate patients, this represents a loss to the hospital).

Richardson, D.B., Med J Australia 177:492, November 4, 2002

BACKGROUND: “Access block” is the situation in which ED patients who require hospital admission spend an unreasonable amount of time in the ED because of inability to gain access to an inpatient bed. It has been linked to overcrowding in the ED but its effects on inpatient outcomes have not been evaluated.

METHODS: This retrospective cohort study examined the relationship between access block in the ED and the inpatient length of stay (LOS) for 11,906 admissions through the ED to an Australian hospital in 1999. Access block was defined as a total duration of the ED stay in excess of eight hours. LOS was truncated at a maximum of ten days for this analysis.

RESULTS: Access block was identified for 7.7% of the admissions. The mean total time in the ED was 3 hours, 58 minutes in the no-block group compared with 10 hours, 30 minutes in the access block group. The mean LOS for the hospital stay was 4.1 days for the no-block group compared with 4.9 days in the access block group. A LOS of one day was less frequent for the access block group than the no-block group (21% vs. 29%) while the opposite was true for a LOS of four or more days (54% vs. 49%). A prolonged LOS for the access block group was occurred across different diagnoses and severity of illness. The difference in LOS between the access block and no-block group was particularly pronounced for patients admitted to the hospital outside of “office hours” (5.2 days vs. 4.1 days).

CONCLUSIONS: This study demonstrates an effect of “access block” in the ED on use of hospital resources outside the ED setting. 12 references ([email protected]) 6/03 – #13

But the real wake-up call associated with the holding of admitted patients in the ED is a documented increase in patient mortality. The compelling study by Chalfin, et al, using a database of the Society of Critical Care Medicine found a six or more hour delay in admission of ICU patients was associated with an absolute 2.3% increase in ICU mortality (10.7% vs 8.4%) and an overall inpatient mortality increase of 4.5% (17.4% vs 12.9%).  These are huge differences and are far more important to fix than the giving of antibiotics within four hours or the following of sepsis protocols (both of which have received far more press and visability).

Chalfin, D.B., et al, Crit Care Med 35(6):1477, June 2007

BACKGROUND: “Boarding” of critically ill patients in the ED (i.e., holding the patient in the ED pending ICU bed availability) is increasing in frequency and might delay the provision of essential elements of care. The effect of this practice on outcomes is uncertain.

METHODS: These multicentered authors examined the effect of delayed transfer from the ED to the ICU on selected outcomes in 50,322 critically ill patients treated from 2000 through 2003 at one of 120 adult ICUs included in the Project IMPACT database developed by the Society of Critical Care Medicine.

RESULTS: Delayed transfer to the ICU (six or more hours after the decision to admit) was documented for 2.1% of the patients. There were no significant differences between these patients and those without delayed transfer in measures of illness severity. The median hospital length of stay in surviving patients was 7 days in patients with delayed transfer to the ICU vs. 6 days in those without a delay in transfer (p<0.001). ICU mortality was 10.7% in the group with delayed transfer vs. 8.4% in the comparison group (p<0.01), and corresponding inpatient mortality rates were 17.4% vs. 12.9% (p<0.001). On logistic regression analysis, delayed transfer to the ICU from the ED was an independent risk factor for lower inpatient survival (odds ratio, 0.71). Other independent predictors included advanced age, higher APACHE II scores, male gender and selected diagnoses (trauma, intracerebral hemorrhage and neurologic disease).

CONCLUSIONS: This large study documents an adverse effect of “boarding” of critically ill patients in the ED for six or more hours prior to transfer to the ICU. 33 references ([email protected])
Copyright 2007 by Emergency Medical Abstracts – All Rights Reserved 11/07 – #16

Fisher and colleagues note that even a period of boarding of as little as two hours was associated with an increased inpatient mortality (2% at 2-5hrs; 2.3% at 6-11hrs; 3.1% at 12-24 hrs and 3.8% if greater than 24 hours — and like other studies, length of stay increased by one day).

Singer, A.J., et al, 54, 3 suppl, pS134, September 2009

STUDY OBJECTIVES: Emergency department (ED) crowding results in prolonged boarding of admitted patients in the ED. Concerns that admitting patients to inpatient hallways is unsafe have led to widespread resistance to such practices increasing patient boarding in the ED. The current study determines the association between length of ED boarding and patient outcomes. We hypothesized that prolonged ED boarding of admitted patients would be associated with worse outcomes.

METHODS: Study Design Retrospective cohort study. Setting: Suburban, academic ED, annual census 80,000. Subjects: Consecutive patients visiting our ED between 1/04–9/08. Measures and Outcomes: An electronic medical record system was used to extract patient demographics, ED disposition (discharge, admit to floor, admit to hallway), ED and hospital length of stay (LOS), in-hospital mortality. Boarding was defined as ED LOS >2 hrs after calling in admission. Data Analysis: descriptive statistics were used to evaluate the association between length of ED boarding and hospital LOS, subsequent transfer to an ICU, and mortality.

RESULTS: There were 310,094 ED patient visits, of which 65,940 patients (21.3%) were admitted; there were 2,081 deaths (0.7%). Of all admissions, 2,420 patients went to an inpatient hallway, 5,081 went to an ICU, and the rest went to standard inpatient beds. In-hospital mortality increased with increasing length of ED boarding (2.0% [2–5 h], 2.3% [6–11 h], 3.1% [12–24 h], and 3.8% [>24 h]; P<0.001). Total hospital LOS increased by one day from 4 to 7 across the four ED boarding groups (P 24 h]; P<0.001).

CONCLUSION: There is a direct association between length of ED boarding and increasing mortality and hospital LOS. Efforts to reduce ED boarding of admitted patients are necessary.

Here are two more studies also indicating an increase in mortality associated with ED crowding and length of ED holding.  We’re talking mortality – the ultimate marker of a very dangerous ED problem.

Richardson, D.B., Med J Australia 184(5):213, March 6, 2006

METHODS: The author of this Australian study retrospectively compared ten-day inpatient mortality among 34,377 patients managed during 736 ED shifts with ED overcrowding (the 8% of shifts with the greatest number of patients in the department throughout the shift) and 32,231 patients presenting during an equal number of time-matched shifts in which the ED was not overcrowded.

RESULTS: Patients managed during ED overcrowding were more likely to receive “inferior” care in terms of “standard performance measures” and less likely to receive treatment within recommended time frames – although neither of these parameters is defined in the study’s Methods section. Furthermore, they were more likely to experience “access block” (delay while waiting for an inpatient bed assignment) and to leave without being seen. The ten-day inpatient mortality rate was 0.42% among patients managed during ED overcrowding compared with 0.31% in those managed when the ED was not overcrowded. The author acknowledges that it is impossible to exclude the possibility that these differences were due to higher triage acuity during the overcrowded shifts.

CONCLUSIONS: The author feels that ED overcrowding may be associated with increased short-term inpatient mortality. 14 references ([email protected]) 8/06 – #21

Sprivulis, P.C., et al, Med J Australia 184(5):208, March 6, 2006

BACKGROUND: It has been reported that a hospital occupancy of 85% represents a good balance between unused beds and inpatient flow, and that an occupancy of 90% is associated with ED overcrowding and “access block” (a total ED time in excess of eight hours while awaiting an inpatient bed).

METHODS: This Australian study examined the relationship between hospital occupancy, ED access block and patient mortality among 62,495 first emergency admissions to three tertiary care hospitals. An Overcrowding Hazard Scale (OHS) was developed based on hospital occupancy (with scores of 1, 2 or 3 for occupancy levels below 90%, 90-99% or 100% or higher) and access block occupancy (with scores of 1, 2 or 3 corresponding to less than 10%, 10-19% or 20% or higher ED occupancy), multiplying the hospital occupancy score by the ED access block score to produce an OHS scores of 1-9.

RESULTS: After adjustment for potential confounders, and when compared with a hospital occupancy below 90%, the hazard ratio (HR) for 7-day mortality was 1.2 when occupancy was 90-99% and 1.3 with higher levels of occupancy. Corresponding rates of ED access block at these levels of hospital occupancy were 4.6%, 6.8% and 9.7%, respectively. An OHS score above 2 (defined as “overcrowding”) was associated with increased mortality; the HRs for mortality at 2, 7 and 30 days were 1.3, 1.3 and 1.2, respectively, and the number of deaths at 30 days attributed to overcrowding was 2.3 per 1000 admissions.

CONCLUSIONS: These findings suggest that hospital and ED overcrowding is a source of increased mortality and represents a patient safety issue. 24 references ([email protected]) 8/06 – #22

So why don’t the CEOs fix the ED boarding (crowding) problem? We’ve demonstrated the money lost is substantial and the poorer care that results. A litany of potentially effective solutions have been document over and over – so what’s the problem?

The problem may be in priorities. Who are the customers of the hospital? The customers of the hospital are the members of the medical staff. They are the ones who admit patients to the hospital (even though it is often unappreciated that about 50% of hospital admissions come through the ED). CEOs must befriend the medical staff — their careers depend on it. Are you going to befriend the medical staff if you push them to use discharge lounges and abide by the noon discharge times, if you ask them to monitor outliers whose lengths of stay are excessive, if you have a closed ICU that prevents unqualified physicians from caring for really sick patients, if you take away their ability to admit patients to the hospital by bringing in hospitalists (who may incidentally be more expert at inpatient care), and if you ask that surgery be scheduled evenly throughout the week and not be bunched up on Monday and Tuesday, and if you ask the medical staff to adopt standardized, evidence-based order sets for the treatment of common problems like pneumonia, COPD and CHF.  All of these can be opposed by the medical staff – every one. I know having been the director of a community ED for 25 years.

Yes, there are lots of opportunities to make EDs more efficient, but until the CEOs are willing to press nursing, the lab, imaging, house keeping, and most importantly, the medical staff, it will be just easier (and safer) to not rock the boat too much. If CEOs bonuses were based on not holding admitted patients in the ED for more than two hours after the admit decision was made the problems would be fixed pronto. But it requires leadership and the making of some unpopular and potentially dangerous decisions for the CEO.

Richard Bukata, MD
Editor of Emergency Medical Abstracts (www.ccme.org)


1 Comment

  1. Great article, Richard!

    One question I have, though, is in regard to this sentence:

    ” Hold a patient for six hours and it’s $9,000 in hard cash”

    I don’t understand what you mean by that.

Leave A Reply