Abstract & Commentary

Supplemental Oxygen Administration in the Morbidly Obese: When Less is Better

By Andrew M. Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle, is Associate Editor for Critical Care Alert.

Dr. Luks reports no financial relationship to this field of study.

Synopsis: This single-center, double-blind, randomized, controlled crossover trial demonstrated that administration of 100% oxygen to stable patients with obesity hypoventilation syndrome leads to decreased minute ventilation, increasing dead-space to tidal volume ratio, and worsening hypercapnia.

Source: Wijesinghe M, et al. The effect of supplemental oxygen on hypercapnia in subjects with obesity-associated hypoventilation: A randomized, cross-over clinical study. Chest 2011;139:1018-1024.

Current guidelines for the use of emergency oxygen recommend its judicious use in patients with obesity-associated hypoventilation (OAH) in order to avoid worsening hypercarbia,1 but these guidelines have largely been based on anecdotal clinical experience rather than systematically collected evidence. Wijesinghe and colleagues sought to fill this gap in the literature by conducting a randomized, placebo-controlled trial of supplemental oxygen administration in this patient population to determine whether administration of 100% oxygen leads to worsening hypercarbia.

They included stable obese patients (BMI > 30 kg/m2) with evidence of daytime hypercarbia, defined as the presence of transcutaneous partial pressure of carbon dioxide (PtCO2) ≥ 45 mmHg, and excluded those patients who also had chronic obstructive pulmonary disease (COPD) based on pulmonary function testing results, another disorder associated with chronic hypercarbia, or who were already being treated with bilevel positive airway pressure ventilation. Each subject was studied on two occasions over a 7-day period. On one occasion, they were administered supplemental oxygen with an F1O2 of 1.0 for 20 minutes while on the other occasion they were breathing air for the same duration. The order of the treatments was determined randomly and subjects breathed each gas mixture through a full-face continuous positive airway pressure (CPAP) mask without positive airway pressure held in place with elastic straps around the head. The authors used a surrogate measure for arterial PCO2, PtCO2, which was measured using a combined SpO2/PtCO2 monitor. Minute ventilation was measured using a flow sensor attached to the CPAP mask's expiratory port. Mixed expired CO2 was measured using volumetric capnography and, along with the PtCO2, was used to calculate the dead-space to tidal volume ratio (Vd/Vt) using the Bohr-Enghoff equation.

A total of 23 patients with BMI ranging from 37.4 to 83.6 kg/m2 (median 50.7 kg/m2) were included in the study analysis. The median PtCO2 in this patient population was 47 mmHg (range, 45-67 mmHg) while the median SpO2 was 97% (range, 80-99%). Testing was terminated in three subjects while breathing an F1O2 of 1.0 because the PtCO2 rose ≥ 10 mmHg. Mean PtCO2 increased from 48.7 to 52.7 mmHg while breathing oxygen compared to a decrease from 48.6 to 47.7 while breathing air. PtCO2 increased by > 4 mmHg in 10 of 23 patients (44%) while breathing an F1O2 of 1.0 and none of the patients while breathing air. The rise in PtCO2 weakly correlated with the decrease in baseline saturation below normal. Minute ventilation fell from 10.1 to 8.0 L/min while breathing an F1O2 of 1.0 compared to a decrease from 9.6 to 9.0 L/min while breathing air. Mean Vd/Vt increased from 0.57 to 0.65 while breathing supplemental oxygen and was unchanged while breathing air. The increase in Vd/Vt while breathing oxygen was associated with a significant mean reduction in alveolar volume but no change in dead space volume. Each study was stopped after 20 minutes and the investigators did not determine whether PtCO2 would continue to rise or plateau over a longer time period of oxygen administration.


It is a virtual certainty that if patients come into the emergency room with hypoxemia, they will be placed on supplemental oxygen regardless of their clinical stability and, more often than not, the goal will be to raise their oxygen saturation percentage to the upper-90s range. For many patients this is a safe intervention but in the population of patients with baseline hypercarbia, concern has persisted for a long time based on clinical experience that doing so can worsen hypercarbia and, as a result, mental status. The study by Wijesinghe and colleagues provides nice evidence to support this long-standing suspicion in patients with obesity-associated hypoventilation and provides solid support for claims made in recently published guidelines on this subject.1

The study may be limited to some extent by the fact that they used transcutaneous assessments of the PCO2 rather than arterial blood gases, but the available data suggest that this method has reasonable agreement with arterial values in patients with COPD2 and obesity,3 and there is no strong reason to question the validity of their results as it pertains to the question being addressed in this study. The study should not, however, be viewed as justification for using PtCO2 to monitor unstable inpatients or emergency room patients. It was conducted in stable outpatients and, as with many of the new less-invasive hemodynamic monitors available these days, the accuracy and precision of this monitoring system may fall off considerably in less stable patients, particularly those with hemodynamic instability.

The phenomenon observed in this study has long been attributed to the fact that supplemental oxygen administration decreases the hypoxic drive to breathe these patients depend on in the face of baseline hypercarbia but, as the authors demonstrate, other factors likely play an important role. Supplemental oxygen raises the alveolar PO2 throughout the lungs, including those areas receiving poor ventilation, and, as a result, leads to release of hypoxic pulmonary vasoconstriction necessary to maintain adequate ventilation-perfusion ratios. Increased blood flow to poorly ventilated areas will impair CO2 elimination and worsen hypercarbia. This will decrease the mixed expired CO2 and, as a result, account for the observed change in Vd/Vt and alveolar volume seen in this study. The increasing hypercarbia is also due in part to the Haldane effect, whereby the higher PaO2 and subsequent increase in oxygen binding to hemoglobin leads to decreased carbon dioxide transport in the form of carbamino groups bound to the ends of the hemoglobin chains.

What is particularly interesting about the results of the study above is that this physiology and the resultant changes in CO2 occur very quickly following supplemental oxygen administration. Unfortunately, the authors did not continue oxygen administration beyond 20 minutes. It would have been useful to see the trend over a longer time, particularly given that many patients are left on supplemental oxygen in the emergency room while unobserved for longer than 20 minutes in busy, understaffed emergency rooms.

While the patients included in the study were known to be hypercarbic at baseline, it is important to note that for many morbidly obese patients who arrive in the emergency room or an outpatient clinic with hypoxemia, baseline ventilatory status may not be known at that time. This study tells us that the response in such cases should not be to err on the side of raising their saturations to nearly 98% but rather should focus on getting them up to values between only 88% and 94%. The overall difference in oxygen delivery to the tissues when oxygen saturation approaches 98% is only marginally better than it is with saturations in the 88%-94% range. More importantly, the benefits of such a marginal rise in oxygen delivery are more than offset by the rise in CO2 and potential alterations in the patient's mental status and the associated interventions, such as intubation for airway protection, that may follow.


  1. O'Driscoll BR, et al. BTS guidelines for emergency oxygen use in adult patients. Thorax 2008;63(suppl 6):vi1-vi68.
  2. Cox M, et al. Non-invasive monitoring of CO2 levels in patients using NIV for AECOPD. Thorax 2006;61:363-364.
  3. Maniscalco M, et al. Evaluation of a transcutaneous carbon dioxide monitor in severe obesity. Intensive Care Med 2008; 34:1340-1344.