Make this page my home page
  1. Drag the home icon in this panel and drop it onto the "house icon" in the tool bar for the browser

  2. Select "Yes" from the popup window and you're done!

Can oxygen hurt? News

July 01, 2012

by Mike McEvoy

Can oxygen hurt?

Drug we use most often can cause harm if we give it without good reason

By Mike McEvoy

Bound Tree University

EMS providers began giving oxygen not because it had medically or scientifically demonstrated benefits for patients, but because they could. Yet, inarguably, hypoxia is bad.

John Scott Haldane, who formulated much of our understanding of gas physiology, said in 1917, “Hypoxia not only stops the motor, it wrecks the engine.”

Patients begin to suffer impaired mental function at oxygen saturations below 64 percent. People typically lose consciousness at saturations less than 56 percent, giving airplane passengers no more than 60 seconds to breathe supplemental oxygen when an airplane flying at 30,000 feet suddenly depressurizes1-3.

More recent studies suggest that hyperoxia, or too much oxygen, can be equally dangerous. Hence the drug EMS providers administer most often may not be as safe as originally thought.

Studies on benefits and dangers of oxygen therapy are not new; intensive care practitioners have long recognized the adverse effects of using high concentration oxygen4.

The Guidelines for Emergency Cardiac Care (ECC) in 2000 and 2005 recommended against supplemental oxygen for patients with saturations above 90 percent. The current 2010 ECC Guidelines call for supplemental oxygen only when saturations are less than 94 percent, perhaps in an effort to soften the impact of change5.

What is new are prehospital research studies comparing outcomes of patients treated without oxygen or with oxygen titrated to saturations versus patients routinely given high flow oxygen. These data are frightening; they invariably show impressive patient harm from even short periods of hyperoxia. 

We’ve known since 1999 that oxygen worsened survival in patients with minor to moderate strokes and made no difference for patients with severe stroke6. In fact, the American Heart Association recommended in 1994 against supplemental oxygen for non-hypoxemic stroke patients.

The dangers from giving oxygen to neonates have also been long appreciated7. The most compelling outcome studies of neonates published in 2004 and repeated in 2007 showed a significant increase in mortality of depressed newborns resuscitated with oxygen (13 percent) versus room air (8 percent)9. This led to the current neonatal resuscitation recommendations for use of room air positive pressure ventilation.

In 2002, a study of 5,549 trauma patients in Texas showed prehospital supplemental oxygen administration nearly doubled mortality9. A Tasmanian study of prehospital difficulty breathing patients published in 2010 compared patients treated with oxygen titrated to saturations of 88 to 92 percent to patients treated with non-rebreather oxygen masks.

It showed a reduction in deaths during subsequent hospitalization of 78 percent in COPD patients and 58 percent in all patients10. New studies are showing a troubling pattern of worse outcomes associated with hyperoxia post cardiac arrest11.

Why would oxygen worsen patient outcomes? One mechanism may be absorption atelectasis. Gas laws mandate that increases in the concentration of one gas will displace or lower the concentration of others. Room air normally contains 21 percent oxygen, 78 percent nitrogen, and less than 1 percent carbon dioxide and other gases.

Nitrogen, the most abundant room air gas, is responsible for secretion of surfactant, the chemical that prevents collapse of the alveoli at end expiration. Premature infants often are not developed sufficiently to produce surfactant and require endotracheal administration of animal surfactant.

“Washout” of nitrogen in adult lungs occurs when high concentration oxygen is administered. Lower concentrations of nitrogen can lead to decreased surfactant production with subsequent atelectasis and collapse of alveoli, significantly impeding oxygen exchange.

Oxygen is also a free radical, meaning that it is a highly reactive species owing to its two unpaired electrons. From a physics perspective, free radicals have potential to do harm in the body.

The sun, chemicals in the atmosphere, radiation, drugs, viruses and bacteria, dietary fats, and stress all produce free radicals. Cells in the body endure thousands of hits from free radicals daily.

Normally, the body fends off free radical attacks using antioxidants. With aging and in cases of trauma, stroke, heart attack or other tissue injury, the balance of free radicals to antioxidants shifts.

Cell damage occurs when free radicals outnumber antioxidants, a condition called oxidative stress. Many disease processes including arthritis, cancer, diabetes, Alzheimer’s and Parkinson’s result from oxidative stress.

The concept of free radical damage suggests the old EMS notion that, “high flow oxygen won’t hurt anyone in the initial period of resuscitation” may be dead wrong.

Tissue damage is directly proportionate to the quantity of free radicals present at the site of injury. Supplemental oxygen administration during the initial moments of a stroke, myocardial infarct (MI) or major trauma may well increase tissue injury by flooding the injury site with free radicals.

Finally, consider this: five minutes of supplemental oxygen by non-rebreather decreases coronary blood flow by 30 percent, increases coronary resistance by 40 percent due to coronary artery constriction, and blunts the effect of vasodilator medications like nitroglycerine12. These effects were demonstrated dramatically in cath lab studies13 published in 2005.

Wonder why the 2010 ECC Guidelines recommended against supplemental oxygen for chest pain patients without hypoxia? Now you know: supplemental oxygen reduces coronary blood flow and renders the vasodilators ALS providers use to treat chest pain ineffective.

Where do we go from here? Knowing that both hypoxia and hyperoxia are bad, EMS providers must stop giving oxygen routinely. Oxygen saturations should be measured on every patient.

Protocols need to be aligned to reflect the 2010 ECC guidelines: administer oxygen to keep saturations between 94 and 96 percent. No patient needs oxygen saturations above 97 percent and in truth, there is little to no evidence suggesting any clinical benefit of oxygen saturations above 90 percent in any patient.

Modifications in prehospital equipment will be inherent in controlling oxygen doses administered to patients. In all likelihood, the venturi mask will make a comeback, allowing EMS providers to deliver varied concentrations of oxygen as needed to keep oxygen saturations between 94 and 96 percent.

Few patients will require non-rebreather masks which are prone to deliver too much oxygen (hyperoxia). CPAP (Continuous Positive Airway Pressure) devices will also need redesign as most conventional EMS CPAP delivers 100 percent oxygen. A study conducted by Bledsoe, et al in Las Vegas found that prehospital CPAP using low oxygen levels (28 to 30 percent) was highly effective and safe14.

Bottom line: the drug we use most often can cause harm if we give it without good reason. In the absence of low saturations, oxygen will not help patients with shortness of breath and it may actually hurt them. The same holds true for neonates and virtually any patient with ongoing tissue injury from stroke, MI or trauma. Indeed, oxygen can be bad.



Akero A, Christensen CC, Edvardsen A, et al. Hypoxaemia in chronic obstructive pulmonary disease patients during a commercial flight. Eur Respir J 2005;25:725–30.

Cottrell JJ, Lebovitz BL, Fennell RG, et al. Inflight arterial saturation: continuous monitoring by pulse oximetry. Aviat Space Environ Med 1995;66:126–30.

Hoffman CE, Clark RT, Brown EB. Blood oxygen saturations and duration of consciousness in anoxia at high altitudes. Am J Physiol 1946;145:685–92.

Alteiemer WA, Sinclair SE. Hyperoxia in the intensive care unit: why more is not always better. Curr Opin Crit Care 2007;13:73-78.

O'Connor RE, Brady W, Brooks SC, Diercks D, Egan J, Ghaemmaghami C, Menon V, O'Neil BJ, Travers AH and Yannopoulos D. 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science Part 10: Acute Coronary Syndromes. Circulation 2010; 122: S787-S817.

Ronning OM, Guldvog B. Should Stroke Victims Routinely Receive Supplemental Oxygen? A Quasi-Randomized Controlled Trial. Stroke 1999;30:2033-2037.

Rabi Y, Rabi D, Yee W: Room air resuscitation of the depressed newborn: a systematic review and meta-analysis. Resuscitation 2007;72:353-363.

Davis PG, Tan A, O’Donnell CP, et al: Resuscitation of newborn infants with 100% oxygen or air: a systematic review and meta-analysis. Lancet 2004;364:1329-1333.

Stockinger ZT, McSwain NE. Prehospital Supplemental Oxygen in Trauma Patients: Its Efficacy and Implications for Military Medical Care. Mil Med. 2004;169:609-612.

Austin MA, Wills KE, Blizzard L, Walters EH, Wood-Baker R. Effect of high flow oxygen on mortality in chronic obstructive pulmonary disease patients in prehospital setting: randomised controlled trial. BMJ 2010;341:c5462.

Kilgannon JH, Jones AE, Parillo JE, at al. Emergency Medicine Shock Research Network (EMShockNet) Investigators. Relationship between supranormal oxygen tension and outcome after resuscitation from cardiac arrest. Circulation 2011;14:2717-2722.

Harten JM, Anderson KJ, Kinsella J, et al. Normobaric hyperoxia reduces cardiac index in patients after coronary artery bypass surgery. J Cardiothorac Vasc Anesth 2005;19:173–5.

McNulty PH, et al. Effects of supplemental oxygen administration on coronary blood flow in patients undergoing cardiac catheterization. Am J Physiol Heart Circ Physiol 2005; 288: H1057-H1062.

Bledsoe BE, Anderson E, Hodnick R, Johnson S, Dievendorf E. Low-Fractional Oxygen Concentration Continuous Positive Airway Pressure Is Effective In The Prehospital Setting. Prehosp Emerg Care 2012;16:217-221.

About the author

Mike McEvoy, PhD, NRP, RN, CCRN is the EMS Coordinator for Saratoga County, New York and a paramedic supervisor with Clifton Park & Halfmoon Ambulance. He is a nurse clinician in cardiothoracic surgical intensive care at Albany Medical Center where he also Chairs the Resuscitation Committee and teaches critical care medicine. He is a lead author of the “Critical Care Transport” textbook and Informed® Emergency & Critical Care guides published by Jones & Bartlett Learning. Mike is a frequent contributor to and a popular speaker at EMS, Fire, and medical conferences worldwide.Contact Mike at
The comments below are member-generated and do not necessarily reflect the opinions of or its staff. If you cannot see comments, try disabling privacy and ad blocking plugins in your browser. All comments must comply with our Member Commenting Policy.
Bob Sullivan Bob Sullivan Saturday, July 14, 2012 8:18:40 PM How ironic that we've moved from "don't treat the pulse ox." What about patients in hypovolemic shock? I was taught that oxygen can also be carried by plasma, and extra oxgyen helps shocky patients with a normal pulse-ox. Is that still true?
Mike Leming Mike Leming Thursday, August 02, 2012 6:25:22 PM Do patients dietary intake of antioxidants play into any measurable effect against free radical damage in these studies?
James L. Jones James L. Jones Thursday, August 02, 2012 7:54:30 PM Dear Mike, I totally respect your perspective. I have always enjoyed reading everything you write. But introducing Nuance into the administration of oxygen really could kill somebody. With all due respect, we need to keep the oxygen algorithm very simple. If a patient is breathing too fast or too hard, they need oxygen, until proven otherwise.We should never teach EMS providers to treat the reading on a machine. Keep in mind in many rural areas, Nuance is not a word well understood. Most of American EMS is rural, served by well intentioned and dedicated individuals who could easily misconstrue the meaning of your article. Until EMS is a real profession, which employs college educated folk, we really need to keep this stuff very simple. Nuance is deadly. Don't treat any machine, treat every patient.
Greg Friese Greg Friese Friday, August 03, 2012 5:45:03 AM James, Mike has provided an easily digestible and well referenced article. I am don't see this to be particularly confusing. While I agree that EMS needs to be a profession of paramedics with at least an associates degree high quality articles like this one are excellent education tools for providers of all levels and education backgrounds.
Grant Johnson Grant Johnson Friday, August 03, 2012 8:48:18 AM I think it's important to remember to use the pulse ox as one tool of many to help treat the patient. The studies are showing that we will have better outcomes if we use all available tools, our patients will have better outcomes. I remember when it was awesome for patients to have pAO2 levels of 300+ and obscenely low CO2 levels. Medicine is ever changing and, even with me being an "older dog", we need to adjust with it.
Jake Stein Jake Stein Monday, August 06, 2012 7:29:23 AM (@Grant)A PAO2 level is not a good indicator since it is calculated. A PaO2 level is the determinant of hyperoxygenation. If the A-a gradient is 300 then there is a problem and it is not from "too much oxygen". The A Tasmanian study is also a joke if anyone bothered to actually read more than the title. The published study itself even states all the limitations of the study. This article also fails to mention just what the guidelines are in NRP for neonates. Some information presented here is vague without an explanation of the "numbers" which can lead to dangerous assumptions just like "never give more than 2 liters of O2 to COPD" or every COPD patient lives at 88 - 92%. Most COPD patients are hypoxic at those levels. If anyone was to look at the oxyhemoglobin dissociation curve to see what the PaO2 is these SpO2 levels and the factors that drive it even lower, you might have a better understanding of the actual oxygenation. Maybe it is time to learn more about oxygen delivery, oxygenation, ventilation and disease processes rather than just more paint by the numbers recipes. But then there is another article here about "Passing gas and PEEP" which suggest PEEP "pushes" oxygen into the circulation.
Jake Stein Jake Stein Monday, August 06, 2012 7:36:31 AM This article just gives a few numbers for a recipe. He mentions "examples" but fails to provide the whole story such as the NRP "guidelines" for oxygen and resusucitation and what the A Tasmanian study of prehospital . Maybe more education about the differences between oxygenation and ventilation is required for prehospital. If you anticipate the rise in PaCO2 then treatment is no mystery. You can still give the patient the needed oxygen to relieve work of breathing and provide ventilatory support for the deadspace or V/Q mismatching that occurs. The problem with EMS is that shortcuts in a way of a recipe is always being sought rather than education. Even the AHA goes into extensive research and has produced some excellent references which also includes the exceptions to the "guidelines" they promote.
Robin Thornburg Robin Thornburg Monday, August 06, 2012 6:15:53 PM I really enjoyed reading this and I had this question. Something to think about and study more!
Jon Schirripa Jon Schirripa Tuesday, December 04, 2012 10:44:21 AM Even in med school the information on this topic is not directly explained. Nonetheless, patients with any chronic obstructive disease have adapted to the high levels of PCO2 and the low levels of PO2, but their peripheral chemoreceptors (PO2 below 60) is whats keeping them alive and breathing via central control. Giving the unconscious patient with chronic obstructive disease supplemental O2 will raise the PO2 to the point where the peripheral chemoreceptors will not fire, thus eliminating central control/respiratory drive (hyperventilation). You will see the patient start to suffocate and die right in front of you. I agree with not teaching the reading on a pulse ox. 99% of the time supplemental O2 is never contraindicated, but you need to make the distinction in this one case.
Lynne Curtis Gudes Lynne Curtis Gudes Saturday, March 09, 2013 8:29:54 AM Our volunteer service doesn't have many "college educated folk" but we are smart enough to learn anything that will help our patients. As I tell my students, "Your first EMT class is black and white. Always do this, never do that. Now that you know something, you have to learn nuances. Know all the rules and WHY they are the rule, so you know when to break them." I think rural EMS can handle learning about when not to use 02. If we expect them to be smart enough to give 02 to a pt who is turning blue when the pulse oximeter "says" he is satting at 99%, then we can surely teach them when 02 might not be a good idea.
Kelly Grayson Kelly Grayson Monday, April 01, 2013 10:36:51 AM "Until EMS is a real profession, which employs college educated folk, we really need to keep this stuff very simple. Nuance is deadly." And as long as EMS personnel think nuance is deadly, we will never be a real profession.
Daniel Katzenstein Daniel Katzenstein Saturday, May 11, 2013 3:05:11 PM I have a couple of concerns - 1- Pulse ox readings in the field can inaccurate and therefore misguiding for numerous reasons a)low quality pocket oximeters b)improper fitting of the device due to size of finger and/or field circumstances of application c) hypothermic extremities d) insufficient time spent on getting a proper reading (I left out the nail polish issue as research indicates that its effect is not significant). I see a problem in relying on a device for administering oxygen. Each agency needs to follow the protocols that guide its operation with the hope that those protocols were written with field realities in mind and not just utilizing those from hospital/research facilities.
Jake Stein Jake Stein Sunday, June 02, 2013 8:03:24 AM I can not believe you are theorizing about N2 washout. That is not immediate even in the best PFT labs without special equipment. Your RN education should have covered that. With just a NRB mask it takes a much longer period of time. This is not to say people should be kept on NRB masks for no reason for several days. You are grasping at bits and pieces from articles which have very little sampling data and statistical relevance. A couple of the other articles relate to ICU stays. This is after a few hours in the ED and then at least 24 hours in the ICU on a ventilator with high concentrations of oxygen, PEEP and various pressors. The damage will also depend on the type of ventilator and protocol used to spare the lungs. Sometimes 100% O2 and high PEEP is not enough. Nitric Oxide is used to reduce the PVR with the pulmonary vasoconstriction resulting from hypoxia. Also to give recipe SpO2 numbers is irresponsible unless you go into PVR, pulmonary hypertension, BP, MAP and sepsis. But, there is also another article on this website about "passing gas and PEEP" which it is said PEEP "pushes" oxygen into the circulation.
Bruce Draper Bruce Draper Monday, September 09, 2013 2:44:46 PM All great info which is slow coming to ACLS classes and the ER where I work in San Jose. Thanks for the quality research, Mike! Possible inaccuracies with pocket oximeters is a. concern, but remember, always trust your eyes and ears along with the equipment! I've had my equipment check out with ABG results and it IS a ballpark figure....Problem is when the ballpark is right at the 90% level.
Jake Stein Jake Stein Thursday, November 14, 2013 5:43:03 AM The thing is after a cardiac arrest you will have poor EF (Ejection Fraction from the heart). The lactate will also be very high. Hypothermia will be initiated which shifts the curve and the BG lab will correct for temperature once hypothermia is initiated. Look at the oxyhemoglobin curve to see how different shifts affect the PaO2 (which is what counts) and the measured (not just pulse ox) oxygen saturation relationship. It is rare to actually see an exceptionally high PaO2 post arrest to fit the definition of hyperoxia. It is much, much more than just a "pulse ox" reading and it is not covered in EMT or Paramedic school.

Today's Top Stories

Saturday, April 19, 2014

Line-Of-Duty Deaths

Submit information on fallen EMS providers in your area.

Featured Columnist

Mike McEvoy

Drug Whys

    Featured Product Categories