Analysis of the distribution of time that patients spend in emergency departments
BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.38440.588449.AE (Published 19 May 2005) Cite this as: BMJ 2005;330:1188
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
The UK four-hour standard for the processing of all patients
attending type I emergency departments has been in place since 2004.
Initially the target was set at 90% of all patients being seen and treated
within the four-hour window and over a period of 2 years this increased to
98%. Performance across all acute trusts in England has improved such that
the majority report hitting the target each quarter, with those that fail
doing so by a few percentage points. We have previously shown that the
effect of the target is not uniform and those who are admitted to hospital
from the ED are being affected the most by a ‘spike’ in activity during
the last 20 minutes before four hours which affected 12.3% of admitted
patients and 3.6% of discharged patients in 2004 1.
We sought to find out whether this ‘spike’ was still occurring nationally.
We analysed 12.2 million new patient episodes at English emergency
departments from HES data for 2008/9. The figure shows that the ‘spike’ in
activity is still present and much larger than in 2004, with 30.7% of
patients who are admitted to hospital leaving the ED in the 20 minutes
before the four hour target is breached, and 10.5% of those patient who
are discharged (see figure 1).
Whilst many in the specialty of Emergency Medicine support the benefits
that the four hour target has produced, it is clear that they are not
being experienced by all patients, and that processes throughout the
hospital and wider healthcare system may not have improved to accommodate
it. If acute trusts are experiencing such apparently severe difficulties
in achieving this target, then perhaps a different approach such as the
one announced by the new government of a reduction to 95% combined with
phasing this out altogether and introducing other factors which measure
quality are more appropriate. However, good evidence based indicators of
quality in emergency medicine need development2. To date, we have no
evidence that the 98% four-hour target benefits patient care, and indeed
our findings would suggest that it has encouraged target-led care rather
than needs-led care.
Figure 1: Distribution of patient time in the emergency department by
admitted and discharged in 2008/9
References:
1. Locker T, Mason S. Analysis of the distribution of time that
patients spend in Emergency Departments. British Medical Journal,
2005;330:1188 – 1189
2.
http://www.dh.gov.uk/en/Publicationsandstatistics/Lettersandcirculars/De...
Competing interests:
None declared
Competing interests: No competing interests
Bell and Darling raise a number of points in their response to our
article which we feel may be misleading. As they state, the performance of
Trusts in relation to the four hour target is well known. However the
figures published by the department of health merely state the percentage
of patients treated within four hours, giving no information regarding the
actual length of time patients spend in the Emergency Department. This has
not previously been demonstrated.
Whilst not disputing that department times have improved, we feel that the
statement that the distributions demonstrated in our article would have
been shifted to the right if we had studied the period prior to the
inception of the four hour target is naïve. We believe that the peak in
the distribution we have demonstrated immediately prior to four hours
would have been absent prior to inception of the target. What is not known
is the whether improvements in performance against the target have been
achieved by shifting the whole distribution or whether it has been
achieved by the movement of subset of patients from times exceeding four
hours to times falling within the last twenty minutes of the target time,
thus generating the peak in the distribution. It is likely that both
factors have had some influence and we hope our future work in this area
will demonstrate the relative contributions of each.
Contrary to Bell and Darlings belief, we do not imply in our article
that patients are admitted unnecessarily, although there is some evidence
that this may be occuring [1].
The impact of the 4 hour target on patient outcome and discharge decisions
made is generally unknown, but of key importance in terms of safety and
quality of care provided. There are two other possible explanations for
the peak in the distribution immediately prior to four hours. Departments
may be closely monitoring the duration of time patients have been in the
Emergency Department and able to expedite transfer out of the department
in the last minutes of the target time with the patients either being
discharged or admitted to a ward or alternative assessment area
appropriate to their clinical condition. Secondly, it is possible that the
time a patient leaves the Emergency Department is being recorded
incorrectly, such that the patient’s total time in the Emergency
Department does not exceed four hours. Such selective alteration of
performance data has been previously demonstrated in relation to ambulance
service response times[2].
The data we have collected does not permit us to draw any firm
conclusions about the cause for the shape of the distribution. However,
what is clear is that one cannot entirely rely upon the government’s
current measure of Emergency Department performance in assessing how the
patient’s experience of the Emergency Department may have improved.
Thomas Locker
Suzanne Mason
1. A&E target and PbR hit Trusts with finance double whammy. HSJ
6 Jan 2005.
2. Commission for Health Improvement. What CHI has found in ambulance
trusts. London: CHI, 2003.
Competing interests:
None declared
Competing interests: No competing interests
The public presentation of health care performance information
remains greatly debated. Particularly, time spent by patients in emergency
departments (ED) is a topical challenge for both caregivers and patients.
Furthermore, time spent could influence patient satisfaction and quality
of health care (1). We read with great interest the article by Thomas E.
Locker and Suzanne E. Mason, which showed a peak of admission or discharge
just before «the four-hours target zone», defined by the NHS plan (2, 3).
However, the authors do not sufficiently stress that time spent indicator
is sometimes ambiguous and should be interpreted with caution.
With the aim of assessing the utilisation of a hospital emergency
department, in 2003-04 we conducted a prospective cohort study in the ED
of the Reference Hospital of Elbeuf (Upper Normandy region, France, 32,176
visits in the ED in 2004). Between July 2003 and February 2004, 169
patients aged 18 years and older were included in the study. For each
patient, the number of complementary tests (electrocardiography,
radiographic and laboratory) and discharge or hospitalisation were
collected. Time spent in the ED, categorised as waiting time (time between
patient arrival and first contact with a physician) and care time were
recorded.
Of these 169 patients, 125 patients (74.0%) resulted in discharge, 44
patients (26.0%) were hospitalised. No patient died. The total mean time
spent in the department was 148 minutes (standard deviation [SD]=91,
median [M]=130): 222 minutes (SD=110, M=194) for hospitalised patients
versus 124 minutes (SD=69, M=110) for discharged patients (p<0.001).
The mean care time was 197 minutes (SD=108, M=165) for hospitalised and 87
minutes (SD=70, M=65) in discharged patients (p<10-4) and the mean
waiting time was respectively, 28 (SD=34, M=15) and 36 minutes (SD=36,
M=26) (p=0.12). Eighty six per cent of the patients spent less than four
hours in the ED: 66% for admitted versus 93% for discharged patients
(p<10-4). A strong positive and significant correlation was found
between the numbers of complementary tests performed and the total time
spent (Spearman correlation rank=0.65, p<10-4). Near three-quarter of
patients (74.0%) had at least one of these complementary tests performed
in the ED. The number of complementary tests per patient was 5.4 (SD=2.2)
in hospitalised group and 1.7 (SD=2.0) in discharged group (p<10-4).
In France, to our knowledge, no threshold period regarding spent time
in the ED has been recommended by the national health authorities. Despite
a limitation of our study due to the small sample studied, our results
underline the fact that most care time is devoted to complementary
investigations performed in the ED and contribute directly to final
decision making: admission or discharge. These results also illustrate the
complexity and the shortcoming of using the time spent as a performance
indicator (4). Time spent data should certainly be used for internal
assessment, and could be a starting point for dialogue in the emergency
staffs, which is related to patient satisfaction. Using time spent as an
external indicator with the objective to inform the public and health
authorities requires exhaustive internal validation and cautious
interpretation prior to publication.
References
1 - Taylor C, Benger JR. Patient satisfaction in emergency medicine. Emerg
Med J 2004; 21:528-32.
2 - Locker TE, Mason SM. Analysis of the distribution of time that
patients spend in emergency departments. BMJ 2005; 330:1188-9.
3 - Department of Health. The NHS Plan. A plan for investment. A plan for
reform. London: Stationery Office, 2000.
4 - Stoop AP, Vrangbaek K, Berg M. Theory and practice of waiting time
data as a performance indicator in health care. A case study from The
Netherlands. Health Policy 2005; 73:41-51.
Competing interests:
None declared
Competing interests: No competing interests
Analysis of the distribution of time that patients spend in emergency
departments
Thomas E Locker, Suzanne M Mason
We were interested in this article but unclear as to the message.
The performance data for all acute hospitals against the 4 hour emergency
care target is well known(1). Significant improvements have occurred in
England over the last 3 years with fewer patients waiting and shorter
journey times for those requiring admission. If a comparative graph had
been drawn for a similar period 2 years previously both curves would have
been shifted (discharged and admitted patients) to the right with a long
tail of patients in both groups waiting longer than 4 hours.
We know from patient surveys that waits and delays in health care are
a major concern therefore this improvement is to be welcomed and public
perception of A/E has improved in England(2). It is known that older
patients wait longer which this study confirms however by definition fewer
must be waiting now than before. Clearly there is room for further
improvement.
Analysis undertaken during the Emergency Services Collaborative for
England would indicate that to achieve the 4 hour target the average time,
not surprisingly, would be just over 2 hours. The fact that 12.5% of
patients move in the last 20 minutes does not imply poor care. However,
to achieve a left shift in the time profiles will require more work by
organisations to pro-actively identify patients at an earlier stage in the
patient journey to smooth flows, whether for admission or discharge. The
data also shows a similar number of patients are discharged (11,161) or
admitted (11,563) between 220 and 239 minutes. It is important that
patients who require admission are admitted to the most appropriate care
setting as quickly and efficiently as possible. If the implied suggestion
is that patients are admitted unnecessarily then achieving the 98% target
would become increasingly difficult to achieve unless there was infinite
beds capacity. For example, if a move from 95% to 98% was achieved by an
increase in admissions of 3% then for an average Trust (500 beds) with 95%
bed occupancy capacity would be reached in 11 days.
1. Department of Health. Clinical exceptions to the 4 hour emergency
care target 2003.
2. Mori. Public perception of NHS – Winter 2003 Tracking Survey.
Research Study Conducted for the Department of Health
Competing interests:
National Clinical Lead Emergency Services Collaborative. Modernisation Agency
Competing interests: No competing interests
I must thank Locker and Mason for producing their data on the time
admitted patients spend in the A&E department prior to their
admission. They found that one in eight patients who are subsequently
admitted are moved out of A&E in the final 20 minutes before the four
hour target period is breached (seen on the graph as a rather unusual
spike just before the 4 hour limit).
This suggests what my colleagues and I in Acute Medicine have
suspected for some time: As the 4 hour limit approaches, A&E staff are
faced with a choice; either to keep a patient within the department and
appropriately deal with their problems before discharging them (but
thereby breaching the 4 hour limit), or to get rid of the patient by the
quickest means possible. They invariably choose the latter option, which
usually entails referral of the patient on to the acute medical team for
admission.
Competing interests:
None declared
Competing interests: No competing interests
The work of the authors is commendable. However there is further
analysis needed with regard to what actually causes delay and whether the
target that is actually declared as four hour is a sham or real.
Should the target of 4 hours actually be till the patient is reviewed and
referred or should it be till the patient is thrown out of the department?
Is patient care being affected because of moving the patient out of the
department prematurely? For e.g patient with fracture neck of femur seen
at the end of four hours is immediately moved out of the department to
prevent the breach of 4 hrs when the more morbid or major problem with the
patient is a major MI which has actually caused the fall and fracture.
The pressures on A & E are indeed enormous but should that be a reason
for inadequate assesment and prompt transfer of the patient out of the
department or of the hospital.
These areas need further assesment and auditing as it is the quality of
patient care and not quantity which is important.
Competing interests:
None declared
Competing interests: No competing interests
It is poor hospital processes that cause patients to wait not clinical need
The fact that trusts struggle to meet a target is not a good
reason to relax or change the target. It might be a good
reason to monitor a broader range of indicators to help
identify errant behaviour or poorly designed processes of
care across the trust. NHS London is launching a pilot of an
A&E dashboard next week (based on the national commissioning
dataset) that fulfils exactly this purpose.
One of the criticisms that has frequently been made of the
4hr target is that it leads to clinically inappropriate
decisions in that last quarter hour before the patients
reaches 4hrs in the department (Mason et. al.'s rapid
response puts it like this " [the target] has encouraged
target-led care rather than needs-led care"). It is surely
wrong to make rushed decisions. But critics rarely ask
whether the likely pathway of care without the target
(perhaps patients would wait 6-8hrs to be admitted) is
likely to be more clinically appropriate.
The evidence says not. Patients don't have long waits
because the waits are needed for clinical reasons. Very few
patients need either observation or treatment in A&E that
lasts longer than 4hrs (certainly fewer than the old 2%
tolerance in the operational standard.) The patients wait
for a number of reasons: because hospitals have no idea
where their free beds are; hospitals manage discharges so
badly that they are most short of beds at the time of day
demand is highest; hospitals completely fail to coordinate
their A&E demand with their inpatient processes; A&E
departments fail to achieve smooth handovers between
different medical staff. For many patients who are admitted,
the clinical need to admit them is obvious on first
examination, but they still often wait until close to 4hrs
to be admitted.
In short, bad things happen to patients because hospital
systems are badly managed and coordinated. Having the 4hr
target switches one sort of clinically inappropriate
behaviour (long waits) for another (rushed admission
decisions) and while rushed admission decisions are bad,
they are not obviously worse than very long waits for
admission.
We should stop blaming targets and start pointing the finger
where blame is due: the poorly coordinated ways hospitals
manage patients.
Competing interests:
Paid advisor to DH and SHAs
on A&E target.
Competing interests: No competing interests