Reply to Rezoagli et al.: CO2 Oscillation during Cardiopulmonary Resuscitation: The Role of Respiratory System Compliance.

From the Authors:

We read with great interest the letter by Rezoagli and colleagues, and we are grateful to the authors for opening a constructive discussion on the intrathoracic airway closure phenomenon that we recently reported in a relevant proportion of patients receiving cardiopulmonary resuscitation (CPR) (1). Rezoagli and coworkers suggest that the nonoscillating, steady capnogram reported in patients and cadavers may be in part explained by the drop in respiratory system compliance caused by cardiac arrest and CPR-induced lung volume loss. As a consequence, they hypothesized that the low airway opening index detected in some patients (nonoscillating or poorly oscillating capnogram) reflects the absence of fresh gas flow entering the system, and may be caused by the scarce volume displacement produced by chest compressions at a constant pressure in case of low compliance, rather than airway closure. Indeed, we previously reported that chest compressions induce huge lung volume loss (2), which can result in a reduction in respiratory system compliance (3, 4).

We think that several points support the major role played by airway closure/patency to explain our findings. First, the volume generated by chest compressions was measured as the volume entering the system during chest decompression. A careful examination of the tracings that we reported in patients (5), lung model, and cadavers (1, 2) shows that the flow is limited (“capped”) or absent during decompression. A limitation of flow (6) secondary to airway closure during CPR can occur during decompression because of a dynamic drop in the transmural pressure of the airways. In case of airway patency, the driving force of the flow during chest decompression is the respiratory system recoil force, which lowers alveolar pressure, generating a pressure gradient throughout the airways. In case of airway closure, the alveolar pressure is no longer transmitted to the airway opening, thereby hampering any inspiratory flow. Interestingly, in case of low compliance, this recoil pressure could potentially be higher.

Second, Rezoagli and coworkers suggest that, “For a given pressure applied to the sternum during a chest compression, the gas volume displacement is directly proportional to the compliance of the respiratory system.” This could have been true in the patients undergoing manual CPR. Although guidelines suggest to maintain chest compression depth within a range (7), rescuers may perform CPR with a constant pressure, yielding variable volume displacement according to different compliance. In the experiments conducted in human cadavers and on a bench model, we used a mechanical device (LUCAS 2; Jolife AB/Physio-Control), which is a piston that provides constant chest compression depth, independent from the pressure needed to achieve it. Thus, all the cadavers likely received the same ventral-to-dorsal displacement, which resulted in different changes in intrathoracic/alveolar pressure according to respiratory system and lung compliance; accordingly, we provide a figure showing that the change in the intrathoracic pressure resulting from chest compressions (i.e., a dependent variable if the LUCAS is used) was variable in the three cadavers we studied (Figure 1, left). There was no correlation between the changes in airway opening index and the intrathoracic pressure change resulting from chest compression (Figure 1, right). Moreover, in an exploratory analysis [Supplementary Table 2 in the manuscript (1)], we tested whether airway opening index was affected by whether chest compressions were performed manually or with a mechanical device (constant ventral-to-dorsal displacement, variable applied pressure); there was no difference in our cohort.

Figure 1.

Three cadavers were studied during chest compressions and pressure-regulated ventilation at a rate of 10/min at three different positive end-expiratory pressure (PEEP) levels (0, 5, and 10 cm H2O) in a sequential order for 5 minutes: each cadaver was studied twice while administering 5% and 10% CO2, yielding 18 individual observations (2). (Left) Individual values of the intrathoracic pressure change produced by chest compressions in the three studied cadavers at three PEEP levels. Medians and interquartile ranges are displayed. (Right) Lack of relationship between the changes in the intrathoracic pressure (estimated by esophageal pressure) resulting from chest compressions and airway opening index. Because the ventral-to-dorsal displacement of the thorax caused by the mechanical device was constant, the intrathoracic pressure change resulting from chest compressions reflects system elastance (and compliance).

Third, the ventilatory mode used in our cohort of patients and in experimental settings was pressure-regulated and maintained some positive pressure at end-expiration. This could result in low Vt insufflation as compliance diminishes (8), as suggested by the authors. Nevertheless, because of the continuous chest compression strategy, two or three compressions/decompressions occur during the 1-second high-pressure time of cardiopulmonary ventilation mode. The huge positive pressure generated by chest compression (reducing transpulmonary pressure) is expected to markedly reduce insufflated volume, almost irrespective of the compliance. Conversely, inspired Vt mostly depends on the volume driven by chest decompressions occurring at high airway pressure, which again depends on the recoil pressure expected to be limited by airway closure.

Finally, the fact that we used a mode maintaining positive pressure at end expiration reduced the incidence of airway closure. Without any expiratory positive pressure, we would have seen many more patients having no transmission of chest compression on airway pressure, reinforcing the independence from compliance.

On the whole, the hypothesis that low compliance may in part explain low-oscillating, steady capnograms appears unlikely because of the preeminent role of chest decompression in this context, whose efficiency in driving ventilation appears to critically depend on patency/closure of the airways.