Exercise ECG useful in finding coronary artery disease
BMJ 2010; 340 doi: https://doi.org/10.1136/bmj.c1971 (Published 13 April 2010) Cite this as: BMJ 2010;340:c1971
All rapid responses
The labelling of the axes of the figure with this letter are
transposed on bmj.com but not in the print journal.
The y axis should be post-exercise probability (%) and the x axis pre-exercise probability (%).
I apologise for this error.
Competing interests:
None declared
Competing interests: No competing interests
Dr Bourdillon prompts a valuable discussion in questioning NICE’s
statement that the exercise ECG has no role in the diagnosis of
obstructive coronary disease causing the symptom of chest pain (or “the
diagnosis of angina”, to use NICE’s technically incorrect phraseology)
(1,2). He makes the point that the exercise ECG can change the likelihood
category of approximately one third of patients presenting with stable
chest pain. To be fair, NICE’s methodology takes this into account in a
model of cost-effectiveness that shows that non-invasive imaging
strategies can achieve the same but more cost-effectively than those
starting with the exercise ECG. This finding is supported by earlier
clinical studies of diagnostic strategies (3,4).
Perhaps a more cogent reason for questioning NICE’s conclusion with
regard to the exercise ECG is that they considered only the diagnosis of
angina (again, using their phraseology). They took no account of tests
that might be required for risk assessment after diagnosis and their cost-
effectiveness model cannot therefore determine best practice in overall
management. For instance, if a diagnosis were achieved using an
anatomical test, such as CT coronary angiography, most cardiologists would
want to know, among other things, the exercise time in order to consider
the need for revascularisation in addition to medical therapy on
prognostic grounds alone. A treadmill ECG or functional imaging test
would then be valuable and a diagnostic strategy that included these in
the first place would become more cost effective.
Additional difficulties with NICE’s model are that some assumptions
do not reflect real-life, such as the perfect accuracy of coronary
angiography (an anatomical test) for diagnosing angina (a functional
problem) and the assumption that all patients with coronary disease
require an angiogram. These assumptions lead directly to the relatively
low likelihood threshold (60%) above which initial invasive angiography is
cost-effective, but more realistic models conclude differently (5).
Furthermore, in considering the different imaging strategies, NICE assumed
that all are similar in terms of diagnostic accuracy and cost. In fact,
myocardial perfusion scintigraphy is the only one that is readily combined
with treadmill testing and equivalence between the tests cannot be taken
for granted without a more detailed analysis.
In general terms, I support NICE’s conclusions but I regret that they
did not consider the cost and practicality of implementation and they did
not extend beyond diagnosis. I am concerned that what might be termed “a
good start” may suffer a fate similar to the technology appraisal of
myocardial perfusion scintigraphy (6) – which was largely ignored by UK
cardiology.
References
1 Bourdillon P. NICE and chest pain diagnosis. Exercise ECG useful
in finding coronary artery disease. Br Med J 2010; 340: c1971
2 National Institute for Health and Clinical Excellence. Chest pain
of recent onset: assessment and diagnosis of recent onset chest pain or
discomfort of suspected cardiac origin. http://www.nice.org.uk/cg95.
Accessed 14 April 2010
3 Underwood SR, Goodman B, Salyani S, et al. Economics of
myocardial perfusion imaging in Europe: the EMPIRE study. Eur Heart J
1999: 20: 157-66
4 Shaw LJ, Hachamovitch R, Berman DS, et al. The economic
consequences of available diagnostic and prognostic strategies for the
evaluation of stable angina patients: an observational assessment of the
value of of precatheterisation ischaemia. J Am Coll Cardiol 1999; 33: 661-
9
5 Underwood SR, Barwell A, Basu S, Chambers M. Probabilistic
modelling of the cost-effectiveness of myocardial perfusion scintigraphy
in coronary artery disease (abstract). Eur Heart J 2008; 29: 528
6 National Institute for Health and Clinical Excellence. Coronary
imaging: myocardial perfusion scintigraphy for the diagnosis and
management of angina and myocardial infarction.
http://www.nice.org.uk/ta073. Accessed 14 April 2010
Competing interests:
Noninvasive cardiologist specialising in cardiac imaging
Competing interests: No competing interests
Why exercise ECG is not recommended for the diagnosis of stable angina
Dr Bourdillon is correct that the recent NICE guideline on the assessment and diagnosis of patients with recent onset chest pain does not recommend exercise ECG for the diagnosis of stable angina. The lengthy and detailed justification for this decision is given in the full version of the clinical guideline, but it can be summarised in a few words: other testing strategies have better diagnostic accuracy and are more cost-effective. The data presented by Dr Bourdillon demonstrate that among some patients exercise ECG will correctly identify CAD, i.e. having an exercise ECG is better than tossing a coin. It was not the Guideline Development Group’s contention that exercise treadmill testing provides no useful diagnostic information in patients with chest pain, only that other tests provide more accurate and more cost-effective information. Dr Bourdillon does not present data comparing the full strategy recommended in the guideline with continuing use of exercise ECG.
The letter from Dr Bourdillon reports that the exercise ECG would have changed the management of about 1/3 of patients attending a rapid access chest pain clinic. However, the calculations used seem to apply to all-comers and our recommendation is for no testing in patients with nonanginal pain and patients with probabilities of CAD >90% or <_10. this="this" excludes="excludes" about="about" _46="_46" of="of" all="all" patients="patients" from="from" testing.="testing." among="among" those="those" for="for" whom="whom" diagnostic="diagnostic" testing="testing" is="is" recommended="recommended" dr="dr" bourdillons="bourdillons" attempt="attempt" to="to" amplify="amplify" value="value" through="through" consideration="consideration" the="the" degree="degree" st="st" depression="depression" merely="merely" buys="buys" specificity="specificity" at="at" expense="expense" sensitivity="sensitivity" increasing="increasing" false="false" negative="negative" diagnoses="diagnoses" and="and" further="further" diminishing="diminishing" in="in" with="with" a="a" low="low" pre-test="pre-test" probability="probability" disease.="disease." br="br"/>
Turning to the letter from Professor Underwood, the remit of the chest pain guideline was to look specifically at the assessment and diagnosis of patients with recent onset chest pain/discomfort that may be of cardiac origin. The guideline did not consider the role of exercise ECG in assessing prognosis among people with an established diagnosis of CAD. This issue will be considered further in the forthcoming stable angina guideline. While the chest pain clinical guideline did not cover costs or implementation, these issues are the focus of considerable ongoing work being undertaken by NICE; implementation support materials are available on the NICE website at http://guidance.nice.org.uk/CG95.
Competing interests:
The author is Chair of the Guideline Development Group
Competing interests: No competing interests