Influence of F_IO2 on P_aCO2 During Noninvasive Ventilation in Patients with COPD: What Will Be Constant Over Time?

To the Editor:

I read with interest the original article entitled “Influence of F_IO2 on P_aCO2 During Noninvasive Ventilation in Patients with COPD,” in which 17 subjects with stable COPD were evaluated after stabilization of acute respiratory failure.¹

Of the 17 subjects, 8 had pneumonia, and 9 had COPD exacerbations. All subjects were chronic retainers of P_aCO2 at > 45 mm Hg and were treated by noninvasive ventilation with F_IO2 < 0.50 and, after 40 min, F_IO2 = 1. The objective of the study was to monitor the changes in P_aCO2 and secondarily to observe changes in breathing frequency, tidal volume, minute volume, and Glasgow coma scale. The authors found no significant alteration in either P_aCO2 or the secondary outcomes assessed.¹

My opinions regarding these findings and their applicability to clinical practice are as follows.

1. It has long been known² that hyperoxia may increase pulmonary dead space, which reverses hypoxemic pulmonary vasoconstriction in the existing regional lung. Consequent to this, hypoxemic pulmonary perfusion would increase in poorly ventilated areas of the lung because of increase in dead space and P_aCO2. This effect would be more pronounced in patients with COPD and P_aO2 < 60 mm Hg. The authors reported basal P_aO2 values of 101.4 ± 21.7 mm Hg, with an average S_pO2 of 94.3 ± 2.2%.

2. The response time of exposure to high F_IO2 for 40 min might not be the same as that in longer time periods, and therefore, the authors' findings may not be relevant to clinical practice.

   For the patient with respiratory failure and hypoxemia, exposure to F_IO2 might be for longer than 40 min; instead, it could be for longer periods of 6–8 h. During such prolonged periods, a patient with moderate-to-severe COPD would require an increase in the minute volume to maintain a constant P_aCO2 with high exposure to F_IO2. This might be beyond a patient's ventilatory capacity. Even with noninvasive ventilation, the patient may require adjustments in inspiratory pressure levels and noninvasive ventilation strategies that allow a preset tidal volume to be maintained.³

3. For these 17 subjects during the phase of stabilization, the authors provided no spirometric data, no plethysmographic data, and no other data to assess the degree of COPD severity.⁴ In a patient with COPD, depending on the degree of lung hyperinflation and hypoxemia, the increase in F_IO2 could reverse the hypoxemic pulmonary vasoconstriction and alter ventilation/perfusion, causing an increase in dead space.

   In contrast, the response to changes in neurological status (encephalopathy) measured by the Glasgow coma scale during so short a period as 40 min might not be the same as that in long time periods at high F_IO2 because of the diffusing capacity for carbon monoxide in cerebrospinal fluid.^5,6

The authors did not provide data on dead space measurement or previous pulmonary hypertension values, and hypercapnia could affect clinical measurements.

In their final analysis, the authors evaluated the changes in P_aCO2, breathing frequency, minute volume, and tidal volume in a group of stable subjects undergoing noninvasive ventilation with an F_IO2 of 1 for a short time period. Their findings might not be comparable to findings obtained in subjects with different degrees of COPD exacerbation and exposed for prolonged periods to F_IO2, for whom oxygen therapy has some impact on survival.

I suggest that the foregoing observations be taken into account when considering the use of high F_IO2 to improve oxygenation in hypercapnic patients.

• Copyright © 2014 by Daedalus Enterprises