Continuous monitoring of intrinsic PEEP based on expired CO2 kinetics: an experimental validation study

Abstract

Background: Quantification of intrinsic PEEP (PEEPi) has important implications for patients subjected to invasive mechanical ventilation. A new non-invasive breath-by-breath method (etCO₂D) for determination of PEEPi is evaluated.

Methods: In 12 mechanically ventilated pigs, dynamic hyperinflation was induced by interposing a resistance in the endotracheal tube. Airway pressure, flow, and exhaled CO₂ were measured at the airway opening. Combining different I:E ratios, respiratory rates, and tidal volumes, 52 different levels of PEEPi (range 1.8-11.7 cmH₂O; mean 8.45 ± 0.32 cmH₂O) were studied. The etCO₂D is based on the detection of the end-tidal dilution of the capnogram. This is measured at the airway opening by means of a CO₂ sensor in which a 2-mm leak is added to the sensing chamber. This allows to detect a capnogram dilution with fresh air when the pressure coming from the ventilator exceeds the PEEPi. This method was compared with the occlusion method.

Results: The etCO₂D method detected PEEPi step changes of 0.2 cmH₂O. Reference and etCO₂D PEEPi presented a good correlation (R² 0.80, P < 0.0001) and good agreement, bias - 0.26, and limits of agreement ± 1.96 SD (2.23, - 2.74) (P < 0.0001).

Conclusions: The etCO₂D method is a promising accurate simple way of continuously measure and monitor PEEPi. Its clinical validity needs, however, to be confirmed in clinical studies and in conditions with heterogeneous lung diseases.

Keywords: CO2; Dynamic hyperinflation; Intrinsic PEEP; Mechanical ventilation; Volumetric capnography.

Fig. 1

Experimental setup: single-limb circuit configuration. The experimental setup used to validate the end-tidal dilution method (etCO₂D) to estimate auto-PEEP. See text for detailed explanations of the components. In this configuration, the exhaled air (black arrow) is always confronted with the continuous flow/pressure of fresh gas generated by the ventilator (white arrow). Whenever the end-expiratory pressure of the animal exceeds the pressure/flow coming from the ventilator, a complete normal capnogram is obtained from the CO₂ sensor. If, however, the expiratory pressure of the ventilator exceeds the end-expiratory pressure of the patient, the capnogram will be diluted. In this evaluation, different levels of intrinsic PEEP were created and detected by stepwise increasing the level of end-expiratory pressure from the baseline value of 2 cmH₂O in 1 and 0.2 cmH₂O steps until de first dilution of the end-tidal portion of the capnogram was diluted. This pressure corresponded to the level of PEEPi

Fig. 2

Representative volume-based capnograms during incremental PEEP with 1 cmH₂O (a–c) and 0.2 cmH₂O (d, e) steps to determine the etCO₂D PEEPi. The previously determined occlusion method yielded a PEEP value of 5.4 cmH₂O. a A full normal capnogram is represented. b When the expiratory pressure in the ventilator is increased from 5 to 6 cmH2O, the end-tidal dilution becomes visually evident. c When the end-expiratory pressure is increased further to 7 cmH₂O, the capnogram becomes diluted at an earlier point during expiration

Fig. 3

Regression analysis. Reference PEEPi (OM) and etCO₂D PEEPi method had a good linear correlation (R² 0.80, P < 0,001)

Fig. 4

Bland-Altman analysis between the two PEEPi methods, OM minus etCO2D. OM, occlusion method