Oxygen is harmful - you just didn't know it

Changing practice is plain hard. Many of us cling to what we've been taught in medical school - what the professors said and what the textbooks teach must be true. Just like fluid therapy, which we all learnt somewhere, all ill patients should get a drip - right? According to Professor John Myburgh at SMACC 2013, fluids are given by junior staff with brownian motion like randomness, and they have all the potential to cause morbidity and mortality downstream. Similarly, supplemental oxygen, which is slapped onto almost all patients in the resus room , has the potential to cause harm, and we don't even think twice about it.
Now, hypoxia is bad, so oxygen therapy is a given in that situation. But, HYPERoxia may just be as harmful. The article below gives a great review of the historical evidence on oxygen therapy in critically ill patients:

The evidence is rather overwhelming against hyperoxia. Did you know, giving too much oxygen may be detrimental in:
  • AMI patients
  • acute stroke patients
  • septic patients
  • cardiac arrest patients
  • COPD patients
Which is pretty much the majority of the resus cases we see day to day, of whom all would get some form of supplemental oxygen. Do NOTE that while the totality of evidence suggests harm, there is no high quality evidence (i.e. RCTs) to prove that hyperoxia is really harmful. Conversely, there is also no high quality evidence to suggest the opposite; that hyperoxia will result in improved outcomes. The reviewers in the above article note this, and point out that guidelines advocating oxygen use acknowledge the paucity of data to guide oxygen recommendations, and most of it is based on expert opinion.
How do we then rationalise oxygen use? IMO, titrate oxygen use to SpO2, targeting 88-92% for patients with COPD, 94-98% for everyone else, and not letting the SpO2 reach 100% (which may mean PaO2 in the hundreds). The BTS guidelines (see below) lays it out pretty nicely:

In summary, next time, when you're on shift looking after that critical patient, remember to check the oxygen requirements for each patient, and use supplemental O2 wisely.