Introduction:

the PaO2/FiO2 ratio adjusted to barometric pressure is used to assess oxygenation in respiratory failure; however, it does not take PaCO2 into account, unlike the PaO2/FAO2 ratio (PaO2 divided by the alveolar fraction of oxygen).

Objective:

to evaluate the agreement between PaO2/FiO2 and PaO2/FAO2.

Material and methods:

observational, ambispective, cross-sectional and analytical study in patients older than 18 years, under invasive mechanical ventilation with respiratory failure and with different PaCO2 values.

Results:

agreement was analyzed in 64 gasometric measurements taken at admission and when the highest PaCO2 was documented, of 32 patients. The Bland and Altman analysis showed a mean of the differences (bias) of 13 and a 95% limit of agreement, between 54 and -24. Agreement is better when PaO2/FiO2 is greater than 150. Twenty two percent of the measurements did not agree in the assignment to groups with mild, moderate, or severe hypoxemia.

Conclusions:

there is moderate agreement between PaO2/FAO2 and PaO2/FiO2; however, in patients with severe hypoxemia, agreement is better. In patients with PaO2/FiO2 ≥ 150, hypercapnia partly explains the low agreement. PaO2/FAO2 is possibly a better way to assess the degree of oxygenation disturbance.