Intended for healthcare professionals

Rapid response to:


Should oxygen be given in myocardial infarction?

BMJ 2010; 340 doi: (Published 17 June 2010) Cite this as: BMJ 2010;340:c3287

Rapid Response:

Maintain a normal oxygen level until more evidence is available. Hyperoxaemia reduces coronary artery blood flow.

We read with some surprise the editorial by Professor Atar, although
had it been peer-reviewed it would have been more alarming. The question
of whether to administer oxygen in non-hypoxaemic acute coronary syndrome
was give considerable thought in preparing the British Thoracic Society
Emergency Oxygen Guidelines published in 2008 of which we were co-

In considering areas of medicine where evidence is weak, the
physician is faced with a choice. The first principle he or she will
apply is that of "primum non nocere" or first do no harm. He or she will
ask if there is any evidence that giving oxygen is beneficial, the answer
to which is “no”. He or she will then ask if there is evidence of harm to
which the answer is “maybe” based on the recent Cochrane review that
Professor Atar criticises. In response to this alone the answer should be
to limit oxygen to those who are hypoxaemic.

The next area a physician can turn to is expert opinion. Experts too
will consider the principle of doing no harm, but may look to extrapolate
from other knowledge, in this case oxygen physiology. Professor Atar
recommends oxygen on these grounds, citing that the principal problem in
myocardial ischaemia is a lack of oxygen where in fact it is a lack of
blood flow. A systematic review by Farquhar and colleagues reported a
mean reduction in coronary artery blood flow of 8-29% in five studies
where oxygen was given to patients with heart disease and there was a 21-
41% increase in coronary vascular resistance and a mean reduction of 15-
27% in myocardial oxygen consumption.2 Giving oxygen to a patient with
normal haemoglobin oxygen saturation (96-98%) can increase the blood
oxygen content by only 2-4% because very little oxygen is carried in the
blood independently of haemoglobin. This small rise in blood oxygen
content cannot compensate for the reduced coronary blood flow found in all
published studies. Oxygen has similar effects on cerebral blood flow and
oxygen was given routinely to normoxaemic patients with strokes until it
was shown to increase mortality significantly in patients with mild and
moderate strokes with no mortality effect in severe strokes.3 For these
reasons and because published studies have shown a non-significant
increase in death and a significant increase in aspartate aminotransferase
when oxygen was given to non-hypoxaemic patients with myocardial
infarction, the British Thoracic Society Guideline for emergency oxygen
use recommends that oxygen should be given to heart attack patients (and
angina patients) only if they are hypoxaemic (oxygen saturation below
94%).3 This guideline was endorsed by the British Cardiovascular Society
and by 21 other Societies and also by the recent NICE guideline for the
management of acute chest pain of suspected cardiac origin.4

Professor Atar appears to be unaware of any of these publications. He
states that the use of oxygen for patients with angina is “a cornerstone
of treatment” and “undisputed” whereas he is directly contradicting the
BTS and NICE Guidelines for oxygen use in ischaemic heart disease. He has
offered no evidence other than “custom and practice” to support the
continuation of a form of therapy where there is strong physiological
evidence and weak clinical evidence of harm but no proven benefit.

Professor Atar uses the argument that hyperbaric oxygen may reduce
complications in myocardial infarction. It is conceivable that such high
partial pressures may be beneficial in overcoming the reduction in flow,
but this cannot be extrapolated back to using normobaric oxygen for the
reasons given above. It is clear that further trials of normobaric
hyperoxaemia and hyperbaric hyperoxaemia for patients with acute
myocardial ischaemia will be required but clinicians should aim to
maintain normoxaemia until such trials have been concluded.

It is regrettable that the BMJ failed to have this editorial peer
reviewed because the views expressed are so far out of line with current
UK guidance and the author appears to be unaware of the overwhelming
physiological evidence that hyperoxia causes a reduction, not an increase
in oxygen delivery to ischaemic myocardium. We would suggest publishing a
prominent editorial comment or a statement of retraction.

1. O'Driscoll BR, Howard LS, Davison AG; British Thoracic Society. BTS
guideline for emergency oxygen use in adult patients. Thorax. 2008 Oct;63
Suppl 6:vi1-68

2. Farquhar H, Weatherall M, Wijesinghe M, Perrin K, Ranchord A,
Simmonds M,
Beasley R. Systematic review of studies of the effect of hyperoxia on
blood flow. Am Heart J. 2009 Sep;158(3):371-7

3. Rønning OM, Guldvog B. Should stroke victims routinely receive
oxygen? A quasi-randomized controlled trial. Stroke. 1999 Oct;30(10):2033-

4 National Clinical Guideline Centre for Acute and Chronic
Conditions. Chest pain of recent onset: Assessment and diagnosis of
recent onset chest pain or discomfort of suspected cardiac origin

Competing interests:
We are the lead authors of the British Thoracic Society Guideline for emergency oxygen use in adult patients (2008)

Competing interests: No competing interests

28 June 2010
B Ronan O'Driscoll
Consultant Respiratory Physician
Anthony G Davison, Luke S Howard
Salford Royal University Hospital. Salford M6 8HD