Manchester triage system in paediatric emergency care: prospective observational study
BMJ 2008; 337 doi: https://doi.org/10.1136/bmj.a1501 (Published 22 September 2008) Cite this as: BMJ 2008;337:a1501
All rapid responses
We read with great interest, the paper entitled Manchester Triage
System in Paediatric Emergency Care(1). We are surprised by the conclusion
drawn from the study both in the article and the editorial(2).
The Manchester Traige System has been validated in a number of
circumstances(3,4,5). The present authors compared an unvalidated triage
system (misleadingly called ‘Reference Standard’ by the authors) against
the Manchester Triage System and found little agreement between the two
systems. The inference could well be that the Reference standard triage is
not valid. However the authors conclude that Manchester Triage System is
at fault and is not valid.
Triage can be validated against mortality. It requires a large
sample. The WHO has developed Emergency Triage And Treatment (ETAT)
score(6). This was validated against the mortality in each score. At our
institution, we have recently developed a triage system utilising the
abnormal physical variables of the systemic inflamatory response
syndrome(7,8). This too has been validated against mortality in the UK and
India(9).
Thus, although mortality is low, a good triage system will exibit a
calibrated increase in mortality with increase in score if the sample size
is large enough. Mortality will perhaps be the best gold standard against
which traige scores can be evaluated. Morbidity may also be an objective
index against which such triage scores can be measured.
1.Veen MV, Steyerberg E W, Ruige M, van Meurs A,Roukema J, van der
Lei J, Moll HA. Manchester triage system in paediatric emergency
care:prospective observational study BMJ 2008;337:792-795
2.Mannconochie I, Dawood M. Manchester triage system in paediatric
emergency care BMJ 2008;337:767-768
3.Roukema J, Steyerberg EW, van Meurs A, Ruige M, van der Lei J, Moll HA.
Validity of Manchester triage system in paediatric emergency care
Emergency Med J 2006:23:906-910
4.Cronin JG. The introduction of Manchester triage scale to an emergency
department in Republic of Ireland Accid Emerg Nurse 2003;11:121-125
5.Cooke M W, Jinks S. Does the Manchester triage system detect the
critically ill? J Accid Emerg Med 1999;16:179-181
6.Robertson MA, Molyneux EM. Description of cause of serious illness and
outcome in patients identified using ETAT guidelines in urban Malawi Arch
Dis Child 2001;85:214-217
7.Kumar N, Thomas N, Singhal D, Puliyel J M, Sreenivas V.Triage score for
severity of illness Indian Pediatrics 2003;40:204-210
8.Bhal S, Tyagi V, Kumar N, Sreenivas V, Puliyel JM J Postgrad Med
2006;52:102-105
9.Gupta MA, Sahni M, Puliyel JM, Rangasami J, Chakrabarti A, Halstead R,
Green DA, Puliyel A,Sreenivas V.International collaboration validity SICK
score: a non-invasive severity of illness assesment Arch Dis Child 2008;93
supl 1:A10
Competing interests:
None declared
Competing interests: No competing interests
With interest we read the reaction by Dr. Sue Leraci ‘There’s more to
triage’ (1) on our paper on the validity of the Manchester Triage System
(MTS) in paediatric emergency care.(2)
Although, we are aware of the Australasian Triage Scale (ATS), we did
not discuss the system since our paper focused on the validity of triage
systems in paediatric emergency care. We did not find any publications on
the validity of the ATS for the specific paediatric population, in the
emergency care setting.
Our aim was to validate the MTS in paediatric emergency care. The
reference standard for urgency has to be a proxy for severity of outcome.
Therefore, our reference standard was based on vital signs and potentially
life threatening diseases for the highest urgency categories, according to
the literature. The reference standard for the lowest urgency categories
was based on hospitalization and resource use, as applied for validation
of other triage systems. (3-5) Experts combined these items to define our
reference standard for urgency level 3,4 and 5. The standard is described
in detail in the appendixes 1 and 2, which are published along with the
paper on BMJ.com.
(http://www.bmj.com/cgi/content/full/337/sep22_1/a1501/DC1)
We agree that a standard based on the literature and expert opinion
is a relative low grade of evidence-based medicine. However, our study is
the first, which validated all urgency categories of the MTS for
paediatric patients and did not only focus on a specific emergent
diagnosis or intensive care admission as outcome. The reference standard
defines five different urgency classes and can be applied in other
settings as well, which allow for comparison between different triage
systems in different settings.
Since a golden standard of urgency does not exist, the independent
reference standard is the best available proxy for severity of outcome.
We applied the methodology of diagnostic research on triage research
and are able to modify the MTS based on the individualized data. We agree
that the goal of triage is not to be 100% predictive for all urgency
outcome levels, but to identify high urgent cases that need immediate
attention and those low urgent patients who can safely wait. Over-triage
will be present in every triage system to allow a safe urgency
classification.
1. Leraci S. There's more to triage.. BMJ 2008; Rapid response 23
Okt.
2. van Veen M, Steyerberg EW, Ruige M, van Meurs AH, Roukema J, van
der Lei J, et al. Manchester triage system in paediatric emergency care:
prospective observational study. Bmj 2008;337:a1501.
3. Baumann MR, Strout TD. Evaluation of the Emergency Severity Index
(version 3) triage algorithm in pediatric patients. Acad Emerg Med
2005;12(3):219-24.
4. Gouin S, Gravel J, Amre DK, Bergeron S. Evaluation of the
Paediatric Canadian Triage and Acuity Scale in a pediatric ED. Am J Emerg
Med 2005;23(3):243-7.
5. Maningas PA, Hime DA, Parker DE. The use of the Soterion Rapid
Triage System in children presenting to the Emergency Department. J Emerg
Med 2006;31(4):353-9.
Competing interests:
None declared
Competing interests: No competing interests
Neither the Van Veen paper nor the accompanying editorial appeared to
be aware of the Australasian work that produced the first published five-
point triage scale in 1993 (Australasian College for Emergency Medicine).
Following a study of over 11,000 occasions of triage by Whitby et al, a
groupd of clinical descriptors was linked to the scale, and it was updated
and published as the Australasian Triage Scale (ATS) in 2000. The
descriptors can be found at
http://www.acem.org.au/media/policies_and_guidelines/G24_Implementation_...,
with an accompanying paper at
http://www.acem.org.au/media/policies_and_guidelines/G24_Implementation_....
This work formed the basis of subsequent five-point scales, including
both Manchester and CTAS. The clinical descriptors were found to be
significantly correlated with the categories. There are specific
descriptors for paediatric presentations. Although it has proven
impossible to show that triage categorisation affects clinical outcome
(because of the difficulty in controlling - or even knowing - the time of
onset), the categories have been shown to correlate with admission rates
to hospital.
It is puzzling to see a study that attempts to validate the
Manchester scale against the standard of a local, less-validated scale,
and puzzling that the standard was not precisely described in the
published study. Can the authors give us more information about their
standard and how it was derived? How has the local standard been
validated?
The main purpose of a triage scale is not to be 100% predictive of
outcome, but to ensure that more urgent cases are seen faster. This really
only becomes a problem when the scale starts to be used as a performance
measure, with waiting time targets. In Australia, the target of 30 mins
for ATS Category 3 patients is becoming very difficult to meet. It is yet
to be shown whether the 30 min target is clinically valid, with methods of
linking to clinical outcomes for far being elusive.
Competing interests:
None declared
Competing interests: No competing interests
With interest we read the reaction by James Plumb ‘But its trauma!
Can we measure the impact of over-triage?’(1) on our article on the
validity of the Manchester Triage System (MTS) in paediatric emergency
care. (2)
Mr Plumb wrote that in his experience urgencies are based on
observations and common sense, rather than using any of the 52 flowcharts.
Applying the system correctly, is a minimum requirement of the MTS to work
properly. In our study and in our current practice, we used and use a
digital application of the MTS. For nurses it provides all 52 flowcharts
and their discriminators, which is very useful since the MTS contains many
flowcharts and discriminators. Secondly, since triage characteristics
(chosen flowchart and discriminator) have to be documented, nurses are
forced to use the MTS correctly. Because all data is easily available,
compliance can be controlled and discussed with the nurses afterwards.
Since a triage system can never predict urgency 100% correctly, it
must be seen as a guideline and experienced triage nurses always have to
use their own expertise in their triage decisions as well.
As was illustrated by Plumb, if patients are over-triaged, it will be
favorable for the particular patient but real high urgency patients have
to wait longer. Under-triaged patients are placed in a too low urgency
category and will be seen too late by a physician.
We argue that triage must be based primarily on medical criteria
instead of social criteria, as Plumb showed in his example. For a patient
who presents with a minor injury and who needs diagnostics, it will be
pleasant for the patient if the X-ray is taken directly, so his total stay
at the ED will be shorter. Although, this patient will not be harmed when
he has to wait for more than one hour, he does take the place of a high
urgency patient who may be harmed.
Over-triage and under-triage will always be present in triage
systems. However, by validating triage systems with a prognostic reference
standard, we are able to develop modifications in order to reduce over-
and under-triage.
1.Plumb J. But its trauma! Can we measure the impact of over-triage?
Bmj. 2008; Rapid Response 4 okt.
2.van Veen M, Steyerberg EW, Ruige M, et al. Manchester triage system
in paediatric emergency care: prospective observational study. Bmj.
2008;337:a1501.
Competing interests:
None declared
Competing interests: No competing interests
I read with great interest the article by Van Veen et al.1 I commend
the study, which highlights some interesting points about this relevant
and important topic.
The authors recognise the considerable limitations of their study
particularly with regard to their devised reference standard but have
still highlighted a pertinent problem: the potential under-triage of
medical patients and over-triage of trauma patients.
As commented on in the editorial by Maconochie and Dawood the key
requirement of any triage system is that an experienced clinician
undertakes triage.2 The Manchester system works by the applying a pathway
approach based on structured questions. Does the system allow for the
experience of the nurse, the time of the day, the position of the shift
the nurse happens to be on or their current mental well being? I think
not. In my experience the numbers attributed are based on observations and
common sense rather than using any of the 52 flow pathways. On the whole
the system works rather well. Does it affect the time in which a patient
is seen? Sometimes it does, but in a busy department with limited numbers
of doctors a category 2 could often wait more than 10 minutes to be seen.
Mistakes will always be made with the appropriate categorisation of
patients presenting in the acute phase of illness or injury because human
error can never be eliminated. The point of systems such as the Manchester
triage system is to try to eliminate human error by creating a structured
approach. Any system designed in this way will be limited and will always
get things wrong. In the case of the aforementioned study it is suggested
that patients are more often over-triaged.
One only has to work for one week in a UK emergency department to see
the effects of over-triage on the telephone by NHS direct. The government
admit that walk in centres and NHS direct have failed to reduce emergency
department attendance.3
But is this actually problematic? Well I would argue that it is. In
the department in which I work any child who is called through on the ‘red
phone’ is automatically seen in the resuscitation room regardless of the
nature of the problem, not a bad idea I here you cry? In principle this is
a very safe way to approach the care of the potentially unwell paediatric
patient. However I have recently seem a small nail in a girl’s toe, and a
trivial abrasion to a boy’s leg, oh and not to mention the ‘anaphylaxis’
that comes in beaming at you and talking away happily.
The point is that 2 nurses and 1 doctor are taken away from the
paediatric emergency department to deal with the incoming call which at
night leaves one nurse to see and triage any one else coming through the
department. The under-triaged medical patient may then wait longer than
they should have done to be seen. This may be the fault of the ambulance
service but the principle is seen throughout the department.
The over-triage of the trauma patient is well recognised by this
observational study and in my experience this behaviour and culture
amongst emergency department nursing staff is a very real phenomenon. I
will often be asked to ‘check over’ or perform a ‘primary survey’ on a
well child who has fallen off their bike but has been put into category 2
due to the mechanism of injury. There appears to be a heightened awareness
to the importance of ‘that missed ruptured spleen'.
The old chestnut of the 4 hour wait becomes important in any
discussion about triage categories as children who have fallen off their
bike but come in collard and taped are often seen quickly to avoid time
wasted that they may have to spend in x-ray and to get them off an
uncomfortable spinal board, again good in principle but is it putting
patients in lower triage categories at risk?
This question is difficult to answer as noted by the authors, as it
is hard to relate morbidity to longer waiting times.
The problem will not disappear. Unwell children will unfortunately
continue to sometimes be missed. Whilst the Manchester system has value
modification appears to be needed.
1. Van Veen M et al, BMJ 2008;337:a1501
2. Maconochie I, Dawood M, BMJ 2008;337:a1507
3. Longwoods review Towards Faster Treatment: Reducing Attendance and
Waits at Emergency Departments. A Briefing Paper from the U.K. National
Coordinating Centre for NHS Service Delivery and Organization Research and
Development. Vol. 4 No. 1 2006
http://www.longwoods.com/product.php?productid=18130&cat=439&page=1
Competing interests:
None declared
Competing interests: No competing interests
Mortality in children is very rare at the emergency department in Western countries
In reaction to the rapid response of Pillai et al (1) on our paper on
the validity of the Manchester Triage System (MTS) in paediatric emergency
care (2) we have some important comments.
To validate the MTS for children, we used a reference standard for
urgency as outcome measure, which is a proxy for morbidity. This standard
is not a triage system. It consists of items, which are gathered at
presentation and at the end of ED consultation, independent of the triage
urgency level. These items were based on literature and combined by an
expert panel. The reference standard aimed to determine “real “ urgency.
Secondly, the authors argue that mortality can be used as an outcome
measure and refer to the study on the ETAT guidelines performed in Malawi.
Secondly they refer to a study evaluating the SICK score. (3,4)
The hospital in Malawi at which the ETAT guidelines were studied has
an in patient mortality in the paediatric population of 10%. The mortality
of our study population at the emergency department was 0.07% (n=17,600).
The congress abstract describing the evaluation of the SICK score used the
mortality of admitted patients as outcome measure. This score was not
applied to all emergency care patients. Furthermore, the mortality was not
provided in the abstract. (3) If mortality is used as an outcome measure
to study triage decisions in children, it must preferably concern
mortality at the emergency department and not the total in hospital
mortality. The total in hospital mortality is influenced by many other
factors besides the triage decision. Mortality can be used as outcome
measure to study triage systems, in settings with a high mortality.
Mortality in children is very rare at emergency departments in West
European countries.
1. Pillai A, Manchanda S. The term "reference standard" is
misleading. Triage can be validated against mortality/morbidity. BMJ
2008;Rapid Response 4 December 2008.
2. van Veen M, Steyerberg EW, Ruige M, van Meurs AH, Roukema J, van
der Lei J, et al. Manchester triage system in paediatric emergency care:
prospective observational study. Bmj 2008;337:a1501.
3. Gupta M, Sahni M, Rangasami J, Chakrabarti A, Halstead R, Green D,
et al. International Collaboration validating Sick score: a non-invasive
severity of illness assessment, RCPCH abstract. Arch Dis Child 2008;93
supl; A10.
4. Robertson MA, Molyneux EM. Description of cause of serious illness
and outcome in patients identified using ETAT guidelines in urban Malawi.
Arch Dis Child 2001;85(3):214-7.
Competing interests:
None declared
Competing interests: No competing interests