Turn the Oxygen DOWN!

by Bob Stuntz


One of the most common knee jerk reactions in medicine is to give oxygen.  Short of breath?  Oxygen.  Chest pain? O2.  Abdominal pain….you get the idea.  And why not, oxygen is harmless, right?  While we do require oxygen to live, and it is 21% of our atmospheric pressure at sea level, oxygen in the medical setting should be considered a drug.  It is a substance that you are administering to treat a medical condition, and it has been shown to have deleterious effects in the right setting.  

Situation 1

A 70 year old COPD patient is brought in by EMS with a complaint of worsening shortness of breath.  He was noted to have a pulse ox reading of 85% on EMS arrival, so a non-rebreather mask was placed.  You order a few nebulizer treatments and some steroids before you saw him, and now his oxygen is up to 98%.  You’re feeling much better about yourself, and you consult the hospitalist for admission.  An hour later, the hospitalist shows up raving about admitting an unconscious patient to the floor.  ”Why isn’t this guy going to the unit?”  You go back to check on your patient, and he is full-on gorked.  What happened?  In this case, oxygen has led to hypercapnic respiratory failure.  

The traditional teaching is that COPD patients depend on low oxygen to keep their reparatory rate up and blow off CO2.  Thus,  if you cause hyperoxia, you knock out their hypoxic reparatory drive, they retain CO2, and then they tank.  Truthfully, the hypoxic reparatory drive only accounts for about 10-15% of your respiratory drive, while CO2 levels contribute most of the rest.  So what does happen in these people when you give them high levels of O2?

A. The Haldane effect:  When we increase blood oxygen levels, hemoglobin binds to mostly oxygen (instead of CO2), leaving CO2 by itself in the blood.  As it builds up in the blood, the acute COPD patient cannot blow it off.  

B. Reversal of hypoxic pulmonary vasoconstriction: In COPD, hypoxic portions of lung are vasoconstricted to match ventilation and perfusion.  When you cause hyperoxia, this vasoconstriction eases, and you get vasodilation in these areas.  This leads to a V/Q mismatch, and increased dead space in the lung.  This again leads to poor ventilation and an increase in plasma CO2.    

There is probably some effect of decreased hypoxic respiratory drive, especially when you drive the paO2 incredibly high, but it’s true effect on hypercapnea is oven overstated (1).  This has been a contentious topic, especially in the prehospital setting.  However, there was a nice article in BMJ in 2010 that really answered the question in my opinion (2).  While it has its flaws, it showed that patients with true COPD exacerbation who received oxygen titrated to a sat of 88-92% had a significantly lower mortality rate than those who received high flow oxygen (2% versus 9%), and less incidence of hypercapnia.  

So what should you do?  First ask the patient if they know where they live normally in terms of a pulse ox. If they are always at 96% at their doctor’s office, 96% is fine.  If they live at 88%, that’s ok.  Get the non-rebreather off and titrate your oxygen to their level.  I have seen some resistance to this from nursing and prehospital providers, however, this is a great opportunity to educate and engage others in the patient’s care.  Get the mask off, put them on nasal cannula, and titrate their oxygen.  It is easy to overlook these patients, as they may initially look better, and if you do not think about this you will have someone crash on you.  

(Like COPD?  Look out for an upcoming EM Res Podcast episode…)

Situation 2 

You resuscitate a cardiac arrest patient.  You intubated them, but you never talked with the respiratory therapist about vent settings, and you never checked an ABG to titrate your oxygen.  You’re a busy man, “ain’t nobody got time for that.”  Your ICU is booked, and your patient stays in the ED for a few hours.  The intensivist has some choice words for you when he comes down and finds the vent to still be set to a FiO2 of 100%.  What’s his problem?

This is another case where oxygen has been shown to be bad for patients.  Hyperoxia following cardiac arrest has been shown to lead to increased free radical production and oxidative stress, alveolar injury and apoptosis, and increased in hospital mortality (3).  With patients across the country spending more time boarding in the ED waiting for an ICU bed, we have to realize that this is not something we can ignore.  Check ABG’s after intubation and titrate your vent settings.  

As you can see, oxygen is not always completely benign.  Titrate your oxygen in COPD patients.  Check for hyperoxia and alter your vent settings to correct it in your resuscitated cardiac arrest patients.  (This is where you need an ABG).  

On a side note, thanks for your patience with the lack of posts and podcasts in the last month.  I have been working on the Emergency Board Review Podcast series (On Twitter @EMBoardReview, and on iTunes as well!).  Hopefully you have been liking that.  If you don’t know about it, go check it out, 100% free Emergency Medicine US Board review.  Not taking the American EM Boards?  It’s still great review material for Emergency medicine no matter where you are.  Look out in the upcoming weeks and months for the EM Res Podcast and Blog to be back up at normal speed.  As always, I’d love to hear your feedback and suggestions!

1. New, A.  Oxygen: kill or cure? Prehospital hyperoxia in the COPD patient.  Emerg Med J. 2006 February; 23(2): 144–146.

2. Austin MA et al. Effect of high flow oxygen on mortality in chronic obstructive pulmonary disease patients in prehospital setting: Randomised controlled trial. BMJ 2010 Oct 18; 341:c5462. 

3. Kilgannon JH, et al. Association Between Arterial Hyperoxia Following Resuscitation From Cardiac Arrest and In-Hospital Mortality. JAMA Vol 303, No 21, June 2, 2010;2165-2171.