Is it possible to exclude a diagnosis of myocardial damage within six hours of admission to an emergency department? Diagnostic cohort study
BMJ 2001; 323 doi: https://doi.org/10.1136/bmj.323.7309.372 (Published 18 August 2001) Cite this as: BMJ 2001;323:372All 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.
Acute chest pain is an important, but neglected, problem in the
United Kingdom.[1] Emerging diagnostic approaches, such as the use of ST
segment monitoring in the emergency department,[2] new cardiac markers,[3]
and chest pain units,[4] have been extensively investigated in the United
States. Yet evaluation in the United Kingdom has progressed little beyond
audit. Herren et al should therefore be congratulated for embarking upon
rigorous evaluation of this problem.[5] The protocol they describe has
impressive diagnostic performance for myocardial infarction, yet there are
several reasons why we cannot assume that this will lead to improved
patient care and cost-effectiveness.
Assessment of acute chest pain requires more than simply ruling out
myocardial infarction. Chest pain units in the United States typically
provide provocative cardiac testing to further risk stratify their
patients. Immediate exercise stress testing is eminently feasible in
British emergency departments and is provided to patients within six hours
of attendance at the Northern General Hospital in Sheffield.
The Manchester study enrolled 383 patients over the course of one
year. This represents approximately one patient per day and accounts for
only a small proportion of attendances with chest pain to an urban
emergency department. In these circumstances the selection process may be
as important as the diagnostic protocol itself. A substantial proportion
of patients have known coronary heart disease and present with
characteristic anginal-type pain, but have no diagnostic ECG changes. Were
these patients included in the study? If not, how were they excluded?
Most importantly, without a control group it is impossible to know
how the cohort described would be managed if there were no chest pain
unit. American studies of chest pain units have demonstrated cost savings
when compared to a control group that is routinely admitted, and
demonstrated improved effectiveness when compared to control groups with
substantial discharge rates.[4] However, a meaningful comparison should
reflect current routine practice, i.e. patients admitted or discharged
according to the clinicians' judgement.
A randomised controlled trial incorporating such a control group is
currently in progress at the Northern General Hospital in Sheffield. An
identical gold standard to that used in Manchester (troponin T) is being
used to compare diagnostic accuracy. Evaluation also includes cardiac
events over six months, quality of life, health utility, patient
satisfaction and cost-effectiveness. Until such data are available chest
pain units should be considered to be of unproven value in the United
Kingdom.
Competing interests: No competing interests
In their study, Herren et al make a case for the use of the short, 6-
hour stay protocol for exclusion of acute myocardial infarction in the
emergency department.1 The rate of missed diagnosis of acute myocardial
infarction in their study is less than 1%, which is startling compared to
the accepted average of 3%-6%. This rate clearly leads us to underestimate
the risks as well as the public health issues involved in the early
discharge of the patients who are misdiagnosed. Although the data is
motivating but further extensive studies are clearly needed before it can
be regarded as acceptable.
The percentage of patients that present to the emergency department
with complaints of chest pain or other symptoms suggestive of acute
myocardial infarction and are not hospitalized are low, but the discharge
of these patients is associated with increased mortality.2 In the United
States, over 1 million patients suffer from myocardial infarction
annually3 of whom about half visit the emergency department. Therefore, a
3.5% missed diagnosis rate represents over 17,000 cases of acute
myocardial infarction missed each year in U.S. alone. This is in addition
to the misdiagnosed cases of unstable angina.
These statistics represent that there is a tremendous task ahead in
effectively screening for acute myocardial infarction and unstable angina.
As the cardiac troponin T rises in less than 4 hours of a cardiac muscle
injury, it is not only considered a non-surgical gold standard in the
diagnosis but can also help improve the clinical outcome and lower
mortality in addition to fewer missed diagnoses.
Therefore, we believe that a comprehensive approach to all chest pain
patients should be utilized including chest pain management algorithm4,
risk factor determination as well as the use of cardiac troponin T in
addition to the CKMB & EKG monitoring in the setting of an emergency
department. Although the present environment constantly demands for
shorter patient stays in the emergency department, the increased mortality
associated with the out of hospital myocardial infarction clearly
outweighs the benefit. Therefore, the high-risk patients should be
hospitalized whereas the intermediate risk patients as well as atypical
presentations may be observed at the short stay units for conventional
periods as the unstable coronary plaque rupture is a dynamic process which
may sometimes be evident later in the course of observation.5
I have no competing interests
Reference:
1. Herren KR, Mackway-Jones K, Richards CR, Seneviratne CJ, France
MW, Cotter L. Is it possible to exclude a diagnosis of myocardial damage
within six hours of admission to an emergency department? Diagnostic
cohort study. BMJ 2001; 323: 372.
2. Pope JH, Aufderheide TP, Ruthazer R, et al. Missed diagnoses of acute
cardiac ischemia in the emergency department. N Engl J Med 2000;342:1163-
1170.
3. Cardiovascular disease statistics, heart and stroke A to Z guide.
Dallas: American Heart Association, 1995:1-2.
4. Stomel R, Grant R, Eagle KA. Lessons learned from a community hospital
chest pain center. Am J Cardiol 1999;83:1033-1037.
5. Mehta RH, Eagle KA. Missed diagnoses of acute coronary syndromes in the
emergency room -- continuing challenges. N Engl J Med 2000;342:1207-1210.
Competing interests: No competing interests
The major flaw of the article is that it proposes that ruling out
myocardial infarction allows the physician to assume that the patient is
"low risk" and may be discharged from the emergency center. Perhaps these
patients are low risk for MI, but not CAD. Another study, using a stress
test or minimally invasive angiography could elucidate this point.
In our unit, we typically use a 9 to 12 hour period of serial cardiac
enzymes (CPK-mb or troponin i), in conjunction with ST segment monitoring,
to rule out patients with myocardial infarction. The patient then
undergoes a stress thallium test or, at the preference of the
cardiologist, a heart catheterization/coronary angiogram. After a
negative test for ischemia, the patient can then be safely discharged. It
would be interesting to see if a 6 hour period of observation could be
substitued for our current 9-12 hour period. However, in order to
confidently assert that a patient is "low risk" for coronary disease and
poor outcome, this study would have to include a larger patient
population, the gold standard of coronary imaging, and careful follow-up
of patients for adverse outcomes.
Competing interests: No competing interests
Thank you for your reply
As your will note in the paper we discussed this point, the results
of a diagnostic test study are totally dependent on the prevalence of
disease in the group you conduct the study in. As such we could have, in
common with many investigators of selected an easier trial group to follow
up. However, this would have changed the prevalence of cardiac and other
disease in this group, different to the group we are all interested in
i.e. low to moderate risk chest pain presenting to emergency departments.
By nature this group is very difficult to follow up,many of the missing
individuals were homeless or had no telephones and fixed addresses.
We felt that a study on a "real group" of patients most of whom would
be discharged would add the greatest value to the body of evidence
available on diagnosis of emergent chest pain.
Finally, we did not just check the patient was alive and well at four
weeks, rather reviewed them in a chest pain clinic with access to ETT and
angioplasty.
Competing interests: No competing interests
Thank you very much for you response to our paper
Your point is very true that it is important to identify those
patients with Unstable angina features. We excluded patients with
evidence of a medical cause of the pain- this explicity included patients
with symptoms of unstable angina.
Although not covered in this paper, we only discharged patients if
they no longer had pain and patients were counselled and given paper
information to return immediately if the pain returned.
What is important about this piece of work is we followed up
discharged patients with a gold standard predictive of increased risk of
morbidity and mortality (reference 19)at 48 hours. Patients with normal
troponin T at 48 hours have no significant increase in risk.
Competing interests: No competing interests
Another problem with this paper is the small number of actual MI's by
the gold standard (in the people that actually got the gold standard).
With only 36 positives (a relatively low event rate, sensitivity can look
quite good just by chance. If there were one more false negative, the
numbers would not look so good. A much larger trial would be needed to
confirm this.
YS
Competing interests: No competing interests
There is one fatal methodological flaw in this paper. If the authors
think that diagnosing MI's acutely is so important and should be done via
the gold standards they have mentioned, how can the test be sure to be
accurate if 65 patients (That is about one in five patients) did not have
the gold standard? The fact that they were still alive at four weeks is
not evidence that they did not have an MI, and how could we know they "did
not have an MI" if the gold standard was not done? The majority of MI's
would still do reasonably well without treatment, especially in the group
of patients in this trial, and might not have evidence of it at four
weeks...
YS
Competing interests: No competing interests
The primary task of the emergency physician is to determine which patient is suffering an acute coronary syndrome, of which acute myocardial infarction is but a relatively small subset. Many studies, including this one, demonstrate the efficacy of markers in identifying those patients with myocardial damage. However, markers do not identify those patients with other variants of the acute coronary syndrome, such as crescendo angina, which can be as dangerous to the patient if discharged from the hospital. The literature does not support the premise that no infarction means no admission, but rather that such patients do not require monitoring in an intensive care unit.
Competing interests: No competing interests
Swapping one admission headache for another
Dear Editor,
Herren et al suggest a useful algorithm for sending chest pain
patients home from an A&E chest pain unit (1), however there are some
practical reservations in their findings, and an important piece of
information missing which determines the usefulness of their findings.
In selecting a group of chest pain patients to enter the 6 hour chest
pain assessment unit (CPAU) they make no mention of how many of the 292
patients admitted to the CPAU would have been admitted or discharged
before the introduction of the CPAU. The true value of picking up 35 gold
standard positive patients is thus dimished if all 35 would have been
admitted for an inpatient MI screen before the unit's introduction due to
other criteria such as the quality of the cardiac chest pain history or a
previous history of cardiac ischemic events.
This also applies to the 238 gold standard negative patients. If
most of these patients would have been discharged directly from A&E
safely before the unit's introduction, then all the unit achieves is to
substantially increase their department's workload by prolonging these
patients time in A&E, and swapping one admission decision headache for
another ('Do I admit to CPAU' vs 'Do I admit as inpatient') . However if
all these 238 patients would have been admitted for an MI screen under the
previous system then it may be of some value.
The study results are extremely important to an acute medical
assessment unit or inpatient ward where A&E patients are admitted for
their MI screen's as they could enable an earlier and safer discharge from
these for a more select group of patients, however setting up a CPAU in
A&E increases A&E workload, increases pressure on space in the
department, and could prolong waiting times for young patients, with no
cardiac history, a non-cardiac history of chest pain and an extremely low
risk of infarction.
Matthew Reed,
Senior House Officer,
Accident & Emergency Department,
Royal Infimary of Edinburgh,
Lauriston Place,
Edinburgh EH3 9YW.
1 KR Herren et al. Is it possible to exclude a diagnosis of
myocardial damage within six hours of admission to an emergency
department? Diagnostic cohort study. BMJ 2001;323:373-375. (18 August.)
No competing interests
Competing interests: No competing interests