Without evidence of benefit, an intervention should not be presumed to be beneficial or safe.

- Rogue Medic

With Conflicting Evidence, What Should We Do? – Oxygen


Dr. Brooks Walsh writes about a recent paper on the amount of oxygen in the blood of resuscitated.[1] This paper appears to show benefit from high oxygen concentrations.[2] This is contrary to almost all the previous papers looking at the effects of supplemental oxygen.

How do we decide what to believe?

Does this paper mean that the people who claim that science constantly reverses itself are right?

Let me answer the second question first.

No. Science does not constantly reverse itself. The media does report oversimplified mistaken explanations of the results of research. When that media-reported inaccuracy is corrected, that does not mean that the original research made any of the claims the media reported. A reversal of a media misrepresentation is not a scientific reversal.

Science does continually modify previous understanding, but that is how we learn. We do not learn by making exaggerated claims on limited evidence.

So, how do we decide what to believe?

First, we need to understand what conclusions are actually being stated in the research. Reading a news article is not the same as understanding the original paper.

Second, we need to understand how the conclusions were arrived at and how the acknowledged limitations affect the results, as well as the effects of limitations that even the authors may not be aware of.

One limitation is the intervention and how the authors decide what the intervention is.

The intervention is the treatment that must be demonstrated to be safer and more effective than providing a placebo or even safer and more effective than doing nothing.

In these studies of supplemental oxygen, the intervention is supplemental oxygen. If we think of the intervention as withholding supplemental oxygen, we are looking at things backwards.

As with any drug, and oxygen certainly is a drug, the dose should also be taken into consideration. We should not administer unproven treatments in the highest dose possible, just because we can. We should have evidence that the intervention is beneficial and that the dose is safe and effective.

Where is that evidence?

The evidence against supplemental oxygen is the stronger evidence.

Stressing the hearts of cardiac patients leads to slower recovery and evidence of more ischemia with supplemental oxygen than with room air. Ischemia is supposed to be prevented by supplemental oxygen.[3]

This study was tiny, but it was prospective, double-blinded. The new study reviewed by Dr. Walsh has larger numbers, but everything else about the study is too poorly controlled to be considered good science. Over 60 years have gone by and we still do not have any studies that provide good evidence that supplemental oxygen is safe and improves outcomes for patients who are not hypoxic.

Dr. Walsh writes –


As the authors note, “Reasons for the benefit of higher oxygen tensions during CPR can more easily be hypothesized than explained.” Given the conflicting data, it might behoove us to proceed cautiously in modifying the targets for oxygen delivery in cardiac arrest.[1]


But which caution is appropriate?

Should we be cautious about giving too much of an unproven treatment?

Or –

Should we be cautious of evidence that we are harming patients with an unproven treatment?

Dr. Walsh is not telling us that there is clear evidence of benefit.

This is the history of a lot of medical treatment is that we harm patients with unproven traditional treatments.

Image credit.

How well do we understand the pathophysiology on which we base our treatment decisions?

We think we understand the way the body works. We start to use a treatment based on that misunderstanding. Somebody eventually studies the treatment. We admit that our misunderstanding misled us to harm patients. We still do not admit that we should have studied the treatment before making it a standard of care.

Until there is good evidence that supplemental oxygen improves outcomes, we should not be aggressive with this unproven traditional treatment.

Will we ever have good evidence of benefit? History continually provides more evidence that traditional treatments are harmful – not evidence that these traditional treatments have been neglected.

Oxygen is a drug. Drugs should be used when indicated by our assessment. Drugs should not be used out of a belief that drugs cannot hurt.

The intervention is the risk, the danger, the threat to the patient until there is evidence of safety and benefit.

It is rare that we ever have that evidence, because we keep unreasonably expecting each treatment to be magically beneficial.


[1] Hyperoxia during CPR associated with improved survival
Wednesday, April 3, 2013
Mill Hill Ave Command

[2] Increasing arterial oxygen partial pressure during cardiopulmonary resuscitation is associated with improved rates of hospital admission.
Spindelboeck W, Schindler O, Moser A, Hausler F, Wallner S, Strasser C, Haas J, Gemes G, Prause G.
Resuscitation. 2013 Jan 17. doi:pii: S0300-9572(13)00042-7. 10.1016/j.resuscitation.2013.01.012. [Epub ahead of print]
PMID: 23333452 [PubMed – as supplied by publisher]

[3] One hundred percent oxygen in the treatment of acute myocardial infarction and severe angina pectoris.
J Am Med Assoc. 1950 Sep 30;144(5):373-5. No abstract available.
PMID: 14774103 [PubMed – indexed for MEDLINE]

In five patients with angina pectoris the administration of 100 per cent oxygen did not favorably influence the onset or duration of pain or the electrocardiographic alterations induced by standard exercise. On the contrary, oxygen therapy actually appeared responsible for more pronounced and persistent electrocardiographic changes in several patients.



  1. It may turn out to be that timing is important as well, and that certain phases of medical emergencies require hyperoxia, while others require normoxia.

    Remember that the early studies looking at aggressive fluids and pressors for sepsis were negative at first. The interventions were based on strong physiologic reasoning, but were provided in the ICU, hours after admission. When they conducted these studies in the ED, however, in the first few hours (rather than hours later in the ICU), we saw a huge mortality benefit.

    For the moment, most of the studies showing evidence of harm of hyperoxia are retrospective, and are of uncertain clinical import. They are very suggestive, and some “common-sense” changes in practice are already being made, but let’s not push the pendulum too quickly in the other direction, until we have higher-quality data.

  2. Not sure I’m a fan of CAPTCHA – I hit “POST,” and I apparently just lost my long response.

    Ok, a briefer response then.

    The importance of this study is that it may suggest that hyperoxia may be beneficial at different points in critical illness, and that we need to be careful before acting too broadly.

    Perhaps there is an analog to the past studies that looked at the treatment of sepsis. Initial studies of sepsis therapy, using aggressive fluid resus and pressors, were conducted in the ICU, hours after admission. Although these studies were based on strong physiologic evidence, the clinical studies were negative. However, when these same interventions were started in the ED, minutes after arrival, we found a huge drop in mortality. Timing matters.

    And so it may be with oxygen. It doesn’t help that most of the studies out there are retrospective, and so are suggestive, but they require confirmation with prospective trials. Some “common-sense” changes in practice have already been implemented, but we need to be careful before extending this reasoning to all clinical situations.


  1. […] the comments to With Conflicting Evidence, What Should We Do? – Oxygen is the following from Dr. Brooks Walsh. […]