Evidence for oxygen use in the hospitalized patient: is more really the enemy of good?

Oxygen in arguably one of the most frequently utilized drugs in modern healthcare, but is often administered to patients at caregivers' discretion with scant evidence as to its efficacy or safety. Although oxygen is administered for varied medical conditions in the hospital setting, published literature supports the use of oxygen to reverse hypoxemia, for trauma victims with traumatic brain injury and hemorrhagic shock, for resuscitation during cardiac arrest, and for carbon monoxide poisoning. Oxygen should be titrated to target an S(pO2) of 94-98%, except with carbon monoxide poisoning (100% oxygen), ARDS (88-95%), those at risk for hypercapnia (S(pO2) 88-92%), and premature infants (S(pO2) 88-94%). Evidence for use with other conditions for which oxygen is administered relies on anecdotal experiences, case reports, or small, underpowered studies. Definitive conclusions for oxygen use in these conditions where efficacy and/or safety are uncertain will require large randomized controlled clinical trials.