Our planet was born approximately 4.8 billion years ago. A billion years later, the oxygen levels began to rise on the earth’s atmosphere. With the rise in oxygen levels, the first signs of life also appeared on the face of the earth. However, it took several billion years before Joseph Priestly (1774) identified oxygen as the life-giving, life-sustaining gas present on the atmosphere of the earth. A century later, oxygen therapy was used for the first time by George Holzapple as part of the treatment for pneumonia. A decade later, Lorraine Smith described acute lung toxicity due to oxygen; he noted that the toxic effects of oxygen were related to oxygen levels in the blood.
How do we reconcile with the adverse effects of oxygen therapy, first described 125 years ago by Lorraine Smith?
What is hyperoxia?
Hyperoxia represents the exposure of cells, tissues, and organs to higher than normal levels of partial pressure of oxygen. Conventionally, a liberal approach to oxygen therapy is followed by most clinicians, mainly as a precaution against avoiding low PaO2 levels. Besides, high oxygen levels are difficult to monitor, mainly because pulse oximetry does not help in recognizing hyperoxia. In an observational cohort study, the mean time-weighted PaO2 was 144 mm Hg in a mixed medical-surgical ICU during the first 7 days of mechanical ventilation and >80 mm Hg during 80% of the duration of mechanical ventilation (1). There is no clear definition of hyperoxia, although a PaO2 higher than 120 mm Hg has been considered to represent mild hyperoxia, and levels higher than 200 mm Hg as severe hyperoxia.
What are the physiological consequences of hyperoxia?
The first evidence of harm from excessive oxygen was described by Lorraine Smith, who described acute lung toxicity. The oxygen molecule has the unique property of acting as an “electron acceptor”. The newly accepted electrons, however, remain unpaired. These molecules with unpaired electrons are called oxygen free radicals, and when they participate in biological reactions, they are termed reactive oxygen species (ROS). ROS play an important physiological role in the removal of mutated or otherwise damaged cells through a process called apoptosis, to enable the generation of new cells. However, excessive oxygen levels result in increased levels of ROS. Anti-oxidants, including superoxide dismutase, catalase, vitamin C and E, which normally scavenge excessive ROS are rapidly overwhelmed. This results in the uncontrolled killing of normal cells by the ROS. Consequently, harmful effects become manifest in different organs.
One of the early effects of hyperoxia in the lung is the loss of ciliary activity, leading to tracheobronchitis. This is characterized by substernal distress, cough, and dyspnea. High FiO2 levels also lead to alveolar collapse, as the oxygen gets rapidly absorbed. However, the presence of nitrogen in the alveoli exerts a “splinting” effect and maintains them open. High FiO2 levels induce changes in the lung that are indistinguishable from changes that occur in acute respiratory distress syndrome.
Oxygen acts as an inhibitor of nitric oxide, which is a naturally occurring vasodilator. This results in a vasoconstrictor effect due to high levels of oxygen, compromising blood flow to the vital organs. Coronary vasoconstriction may ensure, leading to acute myocardial ischemia.
Control of breathing
The administration of high FiO2 is often associated with the development of hypercapnia, particularly in patients with acute exacerbation of chronic obstructive pulmonary disease (COPD). This effect has been conventionally attributed to the abolition of the respiratory drive. However, this is unlikely to be the mechanism for the hypercapnia as minute ventilation has been found to be unaffected with oxygen therapy among patients with acute exacerbation of COPD. A more likely explanation for the hypercapnia is the inhibition of hypoxic pulmonary vasoconstriction. The inhaled oxygen leads to vasodilatation of hypoxic regions of the lung, with the diversion of blood flow away from better-ventilated regions.
What does clinical evidence suggest in real-world practice?
We administer supplemental oxygen to most patients with acute myocardial infarction. In fact, oxygen is part of the MONA therapy (morphine, oxygen, nitrates, and aspirin) recommended by many guidelines. However, does the administration of supplemental oxygen to normoxic patients with ST-elevation myocardial infarction (STEMI) lead to any clinical benefit?
The AVOID study was conducted by the paramedic services in Melbourne, Australia. Patients with STEMI, who were normoxic with a baseline oxygen saturation of >94%, were included in this study. Supplemental oxygen at 8L/min was administered in the intervention group; the control group received no supplemental oxygen. In this randomized controlled study, the administration of supplemental oxygen to normoxic patients with STEMI revealed no beneficial effects. In contrast, an increase in infarct size at 6 months on magnetic resonance imaging, recurrent myocardial infarction, and an increase in the incidence of arrhythmias were noted with supplemental oxygen (2).
How about the effect of oxygen levels on mechanically ventilated patients in the intensive care unit? The OXYGEN-ICU study was a single-center randomized controlled trial conducted in a medical-surgical ICU in Italy, among patients who were expected to stay in the ICU for longer than 72 hours. In the restrictive oxygen arm, the target PaO2 level was between 70–100 mm Hg; in the liberal (conventional) arm, standard ICU practice was followed, with the PaO2 levels allowed to rise to a maximum of 150 mm Hg, and saturation levels were maintained between 97–100%. This study revealed significantly higher ICU mortality in the liberal oxygen arm; besides, there was a higher incidence of shock, liver failure, and bacteremia with liberal oxygen therapy (3).
Mechanically ventilated patients expected to be on ventilation beyond the day of recruitment were included in the multicentric, ICU-ROX study, conducted by the ANZICS-CTG. In the restrictive arm, the oxygen saturation was maintained between 91–96%. An alarm was triggered if the saturation touched 97%, and the FiO2 levels were turned down rapidly to the target level. The FiO2 levels could be reduced to 0.21 if the target saturation levels were met. In the more liberal arm (usual care), no specific measures were taken to limit FiO2 levels, with the saturation maintained between 91–100%. In this study, there was no significant difference between groups in the number of ventilator-free days at day 28; the 90- and 180-day mortality were also not significantly different. On subgroup analysis, possible harmful effects were observed with liberal oxygen therapy among patients with hypoxic-ischemic encephalopathy (HIE). Patients with HIE who received liberal oxygen therapy had fewer ventilator-free days, higher 180-day mortality, and a higher incidence of adverse outcomes on the Glasgow Outcome Scale at 180 days. A subsequent post-hoc analysis of septic patients revealed a trend to improved 90-day survival among septic patients who received a more liberal oxygen therapy (4). However, these findings can only be considered hypothesis-generating and needs to be addressed in future controlled studies.
A multicentric French study compared two levels of oxygen therapy among mechanically ventilated patients with septic shock. In the liberal group, an FiO2 of 1.0 was maintained for the first 24 hours, while the oxygen saturation was maintained 88–95% in the restrictive group. Predictably, this study was stopped prematurely due to concerns with safety with the use of 100% oxygen. The 28-day mortality was higher among patients who received 100% oxygen at the time of stopping. Besides, there was a significant increase in the overall incidence of serious adverse events with 100% oxygen administration. The incidence of ICU-acquired weakness was twice as high, and the incidence of atelectasis was also significantly higher (5).
The bottom line
- Robust clinical evidence suggests that excessive oxygen may be associated with adverse consequences, including increased mortality.
- The safe levels of oxygen are unclear; this may vary according to the underlying condition. Patients with hypoxic-ischemic encephalopathy may be particularly vulnerable to the adverse effects of excessive oxygen.
- Among mechanically ventilated patients, aiming for a saturation level of 91–96% may be appropriate under most clinical circumstances.
- It is appropriate to turn down the FiO2 levels to target saturation levels, in contrast to setting arbitrary lower limits of FiO2.
- Oxygen must be considered as a drug, with adverse effects associated with injudicious use.
- A target saturation level must be prescribed, and the FiO2 levels should be appropriately titrated among critically ill patients.
1. Panwar R, Capellier G, Schmutz N, Davies A, Cooper DJ, Bailey M, et al. Current Oxygenation Practice in Ventilated Patients—An Observational Cohort Study. Anaesth Intensive Care. 2013 Jul;41(4):505–14.
2. Stub D, Smith K, Bernard S, Nehme Z, Stephenson M, Bray JE, et al. Air Versus Oxygen in ST-Segment–Elevation Myocardial Infarction. :8.
3. Girardis M, Busani S, Damiani E, Donati A, Rinaldi L, Marudi A, et al. Effect of Conservative vs Conventional Oxygen Therapy on Mortality Among Patients in an Intensive Care Unit: The Oxygen-ICU Randomized Clinical Trial. JAMA. 2016 Oct 18;316(15):1583–9.
4. Young P, Mackle D, Bellomo R, Bailey M, Beasley R, Deane A, et al. Conservative oxygen therapy for mechanically ventilated adults with sepsis: a post hoc analysis of data from the intensive care unit randomized trial comparing two approaches to oxygen therapy (ICU-ROX). Intensive Care Med. 2020;46(1):17–26.
5. Asfar P, Schortgen F, Boisramé-Helms J, Charpentier J, Guérot E, Megarbane B, et al. Hyperoxia and hypertonic saline in patients with septic shock (HYPERS2S): a two-by-two factorial, multicentre, randomised, clinical trial. Lancet Respir Med. 2017 Mar;5(3):180–90.