Is Exposure to Arterial Hyperoxia During Critical Illness Dangerous?
By Richard Kallet, MS, RRT, FCCM
Director of Quality Assurance, Respiratory Care Services, San Francisco General Hospital
Mr. Kallet reports no financial relationships relevant to this field of study.
SYNOPSIS: Exposure to severe hyperoxia during critical illness is associated positively with increased ICU and hospital mortality
and associated negatively with ventilator-free days.
SOURCE: Helmerhorst HJF, Arts DL, Schultz MJ, et al. Metrics of arterial hyperoxia and associated outcomes in critical care. Crit Care Med 2017;45:187-195.
This retrospective, multicenter, observational study examined approximately 295,000 arterial blood gases (ABGs) in more than 14,000 critically ill subjects throughout the course of their ICU stay. The primary variable of interest was exposure to hyperoxia, which was defined as either mild or severe as judged by the arterial oxygen tension (PaO2) (mild: 120-200 mmHg, severe: > 200 mmHg). These groups were compared to those who were defined as normoxic (PaO2 60-119 mmHg). Fourteen oxygenation metrics (ranging from first PaO2 to PaO2 area under the curve during the first 96 hours in the ICU) were entered into a multivariate logistic model, along with demographic information and illness severity scores (i.e., APACHE IV and SAPS II).
The outcomes of interest were ICU and hospital mortality, as well as ventilator-free days. Subjects tended to be older (mean 65 years) and male (65%), and 50% were classified as a “planned hospital admission.” Although not controlled by protocol, the standard practice in these institutions emphasized “conservative oxygenation,” which was undefined.
The main finding was the existence of a dose-response relationship between exposure to arterial hyperoxia and increased ICU and hospital mortality, as well as decreased ventilator-free days. Metrics that signified greater exposure over time (e.g., mean PaO2 over the ICU length of stay, PaO2 area under the curve) had the strongest associations with mortality.
These results generally, but not universally, were restricted to those experiencing severe hyperoxia. Depending on the metric used, the odds ratio for mortality in severe hyperoxia ranged from 1.74-5.93.
Historically, most hyperoxia experiments focused on pulmonary oxygen toxicity rather than systemic effects. In these classic studies, most animals died from severe hypoxemia and cardiac failure because of severe lung inflammation after several days of exposure to hyperoxia. The degree of systemic organ damage found at pathological exam was relatively minor, and its attribution to either initial hyperoxia or subsequent terminal hypoxia could not be determined. Nonetheless, hepatic and renal congestion were noted and observed to require more than 72 hours of exposure.1
With the advent of electron microscopy, hepatic and renal damage with marked disruption in mitochondrial structure were discovered even in the absence of pulmonary damage. Systemic mitochondrial damage occurred within 24 hours of exposure to fraction of inspired oxygen (FiO2) of 1, whereas it took a week to manifest in animals exposed to a FiO2 of approximately 0.33.2
There is a growing body of evidence that exposure to arterial hyperoxia is more harmful than previously supposed, and allowing patients to have more than a transient exposure to an elevated PaO2 must be re-examined. Indeed, some aspects of our approach to oxygen therapy reflect practices that pre-date pulse oximetry. At that time, the predominant concern was the risk of undetected hypoxemia when ABGs were the only quantitative assessment tool available. During that time, it was common to manage patients with a higher-than-necessary PaO2 as a safety buffer, “just in case” of subsequent pulmonary deterioration.
Despite the ubiquitous presence of pulse oximetry, this tendency seems to persist, although probably on a smaller scale (e.g., during periods of recurrent cardiorespiratory instability). However, this practice likely comes at a cost, and similar to ventilator-induced lung injury, it’s a phenomenon that only can be observed retrospectively using very large databases.
Prolonged exposure to hyperoxia causes considerable tissue damage by generating reactive oxygen and nitrogen species that overwhelm local anti-oxidative defense mechanisms. As shown recently, excessive generation of reactive oxygen species during hyperoxia in patients already in a profound hyperinflammatory state, such as sepsis (or in those at risk for reperfusion injury), likely drives poorer outcomes.3
A potential blind spot may be our tacit assumption that the degree of hyperoxia sufficient to induce pulmonary toxicity is likely to be similar in other visceral organs (i.e., environmental PaO2 > 450 mmHg). Yet, non-pulmonary tissues exist in an oxygen-protected environment compared to the lungs. From an evolutionary standpoint, the lungs of most land-dwelling animals must have robust anti-oxidative defenses to survive, something unnecessary in other organs. Moreover, arterial vasoconstriction resulting in a 20% maximal reduction in organ perfusion occurs when PaO2 exceeds 150 mmHg (the level found in inspired air), which may account for that value being offered as an upper boundary for clinical tolerance.4
This study reminds us that sustained exposure to arterial hyperoxia in clinical practice likely is detrimental to our patients. Although retrospective and observational in nature, the results are consistent with the emerging implications of both preclinical and clinical studies.
Given the ubiquitous presence of pulse oximetry, it behooves us to reassess our assumptions about the risk-benefit ratio of allowing even moderate levels of arterial hyperoxia in the critically ill. Moreover, it provides justification for incorporating closed-loop control of oxygen therapy as this technology becomes available.
- Bean JW. Effects of oxygen at increased pressure. Physiol Rev 1945;25:1-109.
- Felig P. Oxygen toxicity: Ultrastructural and metabolic components. Aerospace Med 1965;36:658-662.
- Rodriguez-Gonzales R, Martin-Barrasa JL, Ramos-Nuez A, et al. Multiple system organ response induced by hyperoxia in a clinically relevant animal model of sepsis. Shock 2014;42:148-153.
- Kallet RH, Branson RD. Should oxygen therapy be tightly regulated to minimize hyperoxia in critically-ill patients? Respir Care 2016;61:801-817.
Exposure to severe hyperoxia during critical illness is associated positively with increased ICU and hospital mortality and associated negatively with ventilator-free days.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.