The advent of pulse oximetry as a widely adopted technology has transformed the overall clinical assessment of children with acute respiratory illness. One result of this transformation is a stronger clinical focus on treating hypoxemia in young children hospitalized with common respiratory illnesses, such as bronchiolitis, asthma, and pneumonia. For example, in acute viral bronchiolitis, international guidelines provide lower-limit oxygen saturation thresholds despite a weak evidence base and lack of consensus on the specific values. Nonetheless, advice to keep oxygen saturations above a precise threshold can be found in almost every bronchiolitis order set or hospital pathway.

On the other hand, there has not been much debate about the opposite end of the spectrum: an upper saturation boundary once supplemental oxygen is administered for these common pediatric conditions. Pulse oximetry readings of 100% are not generally treated as abnormal, but they can represent a wide variety of Pao2 values and tissue oxygen delivery; thus, some readings of 100% while on supplemental oxygen will likely constitute tissue hyperoxia. Furthermore, nebulized therapies given in the hospital setting are typically driven by the flow of 100% oxygen from centralized supply lines, thus providing another source of exposure, which can be substantial for patients on continuously nebulized treatments. Nevertheless, there is a dearth of pediatric research on the scope of the exposure and the potential for harm from excess oxygen outside of the neonatal period.

Recently, the risk of excessive supplemental oxygen has garnered significant attention in adult patients. In a comprehensive systematic review involving 16 000 patients in 25 randomized controlled trials, researchers compared liberal and conservative oxygen strategies in adults with sepsis, critical illness, stroke, trauma, myocardial infarction, cardiac arrest, or emergency surgery.1  Liberal oxygen use increased both in-hospital and out-of-hospital mortality. The largest increase in risk was seen with in-hospital mortality (relative risk, 1.21; 95% confidence interval, 1.03–1.43), translating to a number needed to harm of 71. Findings were consistent across populations and interventions (nasal cannula, face masks, invasive ventilation). These findings have already begun to inform adult guidelines for many different conditions in several countries, with “goalposts” (eg, saturation targets between 90% and 96%) suggested rather than simply a lower threshold for initiation.1 

We propose that similar research on supplemental oxygen use warrants consideration in pediatric acute respiratory illness. For years, the potential harms of excessive supplemental oxygen have been recognized in neonatology, in which there has been a sustained effort to achieve a balance between preventing the negative effects of both hypoxia and hyperoxia in preterm neonates. Hyperoxia is associated with increased risk of retinopathy of prematurity, bronchopulmonary dysplasia, and intraventricular hemorrhage, whereas hypoxemia is associated with poor growth and higher mortality rates. Theoretically, the premature infant’s limited defense mechanisms against oxidative stress underlie the negative sequelae associated with hyperoxia.2  In term infants, there is now an understanding that hyperoxia contributes to negative outcomes in neonatal resuscitation, resulting in recommendations for room air resuscitation of depressed term neonates and tighter oxygen saturation targets by hour of life in the most recent American Heart Association resuscitation guidelines.3 

There is evidence of emerging pediatric interest in this question in the intensive care literature. In a 2016 systematic review of mortality in children who were critically ill, researchers evaluated 6 observational studies and could not draw strong conclusions about any association between hyperoxia and mortality.4  In a larger, more recent observational study involving >6000 patients, researchers concluded that severe hyperoxemia (Pao2 level > 300 mm Hg) was independently associated with mortality in a large quaternary PICU in the United States.5  The Oxy-PICU trial is an ongoing study of liberal (>94%) versus conservative (88%–92%) saturation targets for patients receiving respiratory support in PICU in Britain, with pilot feasibility data published to date and a full trial scheduled to begin in early 2020.6 

Hyperoxia has a multitude of proposed pathophysiological mechanisms, which may exacerbate respiratory disease in children. Both temporary and permanent morphologic changes to lung tissue secondary to hyperoxia can occur within hours of exposure. Free-radical species damage the capillary endothelium, creating a state of hyperpermeability, which can trigger pulmonary edema, and modulations to coagulation and fibrinolysis pathways lead to deposition of fibrin and platelet accumulation, altering the alveolar-capillary membrane.7  Distortion of cell signaling alters antioxidant mechanisms, hastening necrosis and apoptosis.7  Perhaps an even more relevant potential harm in pediatrics, in which viral lower respiratory tract infections commonly result in mucus plugging and alveolar collapse because of smaller airway size, is a phenomenon known as resorption atelectasis. Pressure gradients between alveoli and pulmonary capillary beds increase as higher concentrations of oxygen are inspired such that atelectasis distal to obstructed airways can hasten when breathing oxygen rather than air. Much of our understanding of this concept comes from the anesthesia literature, in which substantial effects on atelectasis may occur even with changes in fractional inspired oxygen from 21% to 30%.8  Children may be even more susceptible to these effects given baseline lower tidal volumes and/or smaller airway diameters.

Joseph Priestley initially named the element oxygen “dephlogisticated air.” That name, which sounds somewhat ridiculous to us today, referenced the phlogiston theory of combustion, eventually overturned by Lavoisier. It was Lavoisier who chose the word oxygene in French, which meant “acid generating,” because he believed oxygen was necessary for acid formation, a theory also overturned by later scientists. Thus, the history of the name for the element oxygen itself contains a history of scientific paradigm shifts.

Similar paradigm shifts routinely occur in medicine, and our thinking about supplemental oxygen may be amid such a shift. In the recent past, few would have questioned oxygen as a first-line therapy in a patient suspected of having a myocardial infarction, yet this practice has undergone reevaluation and reversal. We hope that the next frontier in pediatric research will investigate the possibility that supplemental oxygen, like other drugs, has the potential for harm when used in excess. Specifically, we suggest that a bell-shaped dose-response curve is likely to be present for other pediatric conditions, consistent with the recent adult literature.

The recent Nobel Prize awarded for work elucidating how the body responds to hypoxia provides an opportunity to marvel at the complex and tightly regulated physiology of oxygen delivery in the human body. It also provides an opportunity to note that such evolutionarily refined mechanisms likely do not exist for hyperoxia. Hyperoxia only became possible with the development of supplemental oxygen for clinical use in the modern era; thus, it stands to reason that there is still much we do not know about its physiology.

Although pulse oximetry is limited in its ability to assess the exact degree of hyperoxia in patients receiving supplemental oxygen, invasive measurement of arterial oxygenation is challenging in patients outside of the ICU. Furthermore, saturation (rather than Pao2) targets have been widely accepted as proxies for oxygenation in pediatric and adult trials. While awaiting definitive studies to determine the relationship between high oxygen saturation levels, tissue hyperoxia, and outcomes, strategies to avoid excessive doses of oxygen when saturations are high may be warranted. Practitioners should consider not only when to use oxygen to treat hypoxemia but also how to titrate the dose more judiciously.

Dr Ralston conceptualized the perspective and drafted the manuscript; Drs Lonhart and Schroeder drafted the manuscript; and all authors approved the final manuscript as submitted.

FUNDING: No external funding.

1
Chu
DK
,
Kim
LH-Y
,
Young
PJ
, et al
.
Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis
.
Lancet
.
2018
;
391
(
10131
):
1693
1705
2
Kayton
A
,
Timoney
P
,
Vargo
L
,
Perez
JA
.
A review of oxygen physiology and appropriate management of oxygen levels in premature neonates
.
Adv Neonatal Care
.
2018
;
18
(
2
):
98
104
3
Wyckoff
MH
,
Aziz
K
,
Escobedo
MB
, et al
.
Part 13: neonatal resuscitation. 2015 American Heart Association Guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care
.
Circulation
.
2015
;
132
(
18 Suppl 2
):
S543
-
60
4
Raman
S
,
Prince
NJ
,
Hoskote
A
,
Ray
S
,
Peters
MJ
.
Admission PaO2 and mortality in critically ill children: a cohort study and systematic review
.
Pediatr Crit Care Med
.
2016
;
17
(
10
):
e444
e450
5
Ramgopal
S
,
Dezfulian
C
,
Hickey
RW
, et al
.
Association of severe hyperoxemia events and mortality among patients admitted to a pediatric intensive care unit
.
JAMA Netw Open
.
2019
;
2
(
8
):
e199812
6
Peters
MJ
,
Jones
GAL
,
Wiley
D
, et al;
Oxy-PICU Investigators for the Paediatric Intensive Care Society Study Group (PICS-SG)
.
Conservative versus liberal oxygenation targets in critically ill children: the randomised multiple-centre pilot Oxy-PICU trial
.
Intensive Care Med
.
2018
;
44
(
8
):
1240
1248
7
Walsh
BK
,
Smallwood
CD
.
Pediatric oxygen therapy: a review and update
.
Respir Care
.
2017
;
62
(
6
):
645
661
8
Magnusson
L
,
Spahn
DR
.
New concepts of atelectasis during general anaesthesia
.
Br J Anaesth
.
2003
;
91
(
1
):
61
72

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.