Early recognition of pulmonary dysfunction is critical in accurately and rapidly diagnosing respiratory illnesses such as acute hypoxic respiratory failure, ARDS, and community-acquired pneumonia because the prognosis is largely dependent on how soon the treatment is initiated after the diagnosis. Unfortunately, few easily accessible (noninvasive) techniques exist for measuring oxygenation needs. We congratulate Chalmers et al1 for the study in this issue of Respiratory Care in which they identified changes in 
as an early prognostic marker in patients with ARDS and community-acquired pneumonia. Specifically, the investigators examined a cohort of ∼3,000 subjects with ARDS or community-acquired pneumonia who were admitted to the ICU. The study identified 
to be easily accessible in electronic health records and, through careful monitoring, 
trajectory was a predictor of ventilator-free days.1
/
is the primary classification index and the accepted standard for determining the severity of respiratory illness. However, measurements are noncontinuous, require invasive regular arterial blood gas measurements, and are costly and time intensive for respiratory therapists and other clinicians. Alternatively, 
/
has been used to stratify ARDS by using the relationship of the oxygen-hemoglobin dissociation curve and 
.2 Although there are limitations with using 
to estimate arterial oxygen saturation (
) because some inaccuracies occur in hypoxic ranges,3 
is a strong surrogate for 
and has practicality for treating patients with ARDS over a conservative range (
of 55–80 mm Hg).4 The sigmoidal shape of the oxygen-hemoglobin dissociation curve (Fig. 1) represents the binding affinity between hemoglobin and oxygen, and highlights a direct curvilinear relationship of 
and 
from a 
of 25–80 mm Hg; however, after the genu (somewhere between 70–80 mm Hg), the relationship between 
and 
becomes less predictable, and increases in 
do not reliably improve oxygen saturation. By using this information, 
levels of 70–80 mm Hg are sufficient and do not justify an increase in 
. Chalmers et al1 concluded that 
can be used as an indicator of illness trajectory with similar accuracy to 
/
. A more novel finding was that decreased 
correlated with an increase in ventilator-free days in subjects with acute hypoxic respiratory failure who were in the ICU. A strength of their cohort was that 
was tightly maintained (ICU protocol driven) to achieve an oxygen saturation in the low 90%.
The oxyhemoglobin dissociation curve showing the relationship between the arterial oxygen saturation (
) and the partial pressure of arterial oxygen (
).
Interestingly, many ICUs err on the side of high normal oxygen levels. This practice base is hypoxia averse. Clinical indices set by the American-European Con-sensus Committee target a 
between 80 and 115 mm Hg to obtain “normoxemia.”5 Although it is crucial to keep the organs adequately oxygenated, little benefit in oxygenation occurs after an oxygen saturation of 93% (Fig. 1). More importantly, there is a risk of hyperoxygenation, which has been shown to increase barotrauma6 and is associated with decreased ventilator-free days and increased mortality.7 Furthermore, hyperoxia worsens outcomes after myocardial infarction and strokes. Therefore, for 
to be a useful metric, it has to be tailored to an 
/
of 90%–95%. This requires more frequent monitoring and adjustment of 
at the patient’s bedside. 
adjustments are the single most common intervention performed in the ICU. However, these interventions require expertise, equipment, and personnel to adjust and monitor 
levels. There is a clear need for an efficient, conservative, and proactive technology for recognizing and treating pulmonary injuries.
Closed loop control of 
in patients on mechanical ventilation grew out of an unmet need for providing oxygen in remote environments. For purposes of this editorial, closed loop control of PEEP will not be discussed. Specifically, closed loop control of 
was developed to provide the standard of care oxygenation in the absence of bedside personnel. In addition, remote or austere environments require the need to transport and carry oxygen. Thus, closed loop control of 
was also developed to conserve oxygen as a resource. A key study by Johannigman et al,8 demonstrated that closed loop control of 
improves efficacy (better target oxygen saturation leading to less incidence of hypoxia and hyperoxia) and efficiency (less oxygen usage – upto 50% reduction). Another unique aspect of closed loop control of 
is that the closed loop control algorithm possesses diagnostic utility. Specifically, the activity of the closed loop control of the 
algorithm identifies pulmonary (dys) function, a novel signature of the 
to closed loop control of 
ratio. In a preclinical ARDS model, the embedded closed loop control of the 
algorithm was incorporated into an alert decision support technology, called the smart oxygenation system, which provided earlier recognition of pulmonary dysfunction.9 In addition, by implementing a more-rapid initiation of life-saving interventions via the smart oxygenation system, the ARDS severity changed from severe to mild.10
In the context of the severe acute respiratory syndrome-coronavirus-2 pandemic, there has been an increased need for clinical expertise, ventilators, oxygen, and supplies associated with treating ARDS. Rapid, noninvasive diagnostic and therapeutic systems that efficiently and effectively provide goal-directed oxygen therapy are needed. Automated tasks or remote monitoring controls for oxygenation and ventilation would confer protection for respiratory therapists and other bedside clinicians who take care of patients who are highly contagious. Although there are commercially available ventilators that perform closed loop control of 
, currently none are approved by the FDA in the United States. The FDA has provided an emergency use authorization for closed loop control of 
for ventilators during this pandemic. Indices, such as changes in 
or 
/
that maintain oxygen saturation between 90% and 95% can be strong tools to assess lung function. Perhaps, when coupled to closed loop control of 
and/or smart oxygenation systems, these tools can lead to a quicker therapeutic response, thereby improving outcomes for acute lung injury.
Footnotes
- Correspondence: Catherine M Sampson PhD, Department of Anesthesiology, University of Texas Medical Branch, 301 University Blvd, Galveston, TX 77555-5302. E-mail: casampso{at}utmb.edu
See the Original Study on Page 1521
The authors have disclosed no conflicts of interest.
- Copyright © 2021 by Daedalus Enterprises
