Validation of Airway Occlusion Pressure as a Method of Assessing Breathing Effort During Noninvasive Ventilation

Background: The airway-occlusion pressure is used to estimate the muscle pressure (Pmus) and the occlusion pressure at 100 ms (P0.1) to assess respiratory drive in patients on mechanical ventilation. However, the validity of these maneuvers during noninvasive ventilation (NIV) has not been evaluated. This study was designed to validate the airway-occlusion pressure and the P0.1 described for mechanical ventilation during NIV in a bench model. Methods: This was a bench observational prospective study carried out during January and February 2024 in the ICU laboratory of the Hospital Británico of Buenos Aires. Results: In the non-leakage NIV scenarios with oronasal and total face mask, the NIV-airway-occlusion pressure increased with greater Pmus (P < .001). For a programmed Pmus of 5 cm H₂O, values around 4.5 cm H₂O were recorded for both oronasal and total face masks. At 10 cm H₂O, the values were ∼8 cm H₂O, and at 15 cm H₂O, they were ∼11 cm H₂O. With leaks, this difference worsened as leakage increased and the effort decreased. In the Bland-Altman analysis between mechanical ventilation-airway-occlusion pressure and NIV-airway-occlusion pressure without leakage for oronasal and total face masks, we found a good agreement for the 3 levels of Pmus with both types of masks. With regard to the values of NIV-airway-occlusion pressure with the helmet, Bland-Altman analysis showed a high bias and random error. Multivariate analysis found that NIV-airway-occlusion pressure depends on the type of interface, increased with Pmus, and decreased as leakage increased. The agreement of NIV-P0.1 was not good across all noninvasive measurements. Conclusions: This study constitutes a relevant contribution in the validation of indices to assess Pmus during NIV. In a laboratory setting, the measurement of airway-occlusion pressure in NIV may be used to assess effort estimation in the absence of leakage; however, it will likely be underestimated. P0.1 proved to be an unreliable method. These findings suggest the feasibility of assessing muscle effort during NIV.