Can a Physiologic Insight "Resuscitate" Research in Cardiopulmonary Resuscitation?

Anyone who has ever performed successful cardiopulmonary resuscitation (CPR) knows instantly that they have done something truly incredible. The significance of the act and the elemental feeling of usefulness is the same for all members of the extended healthcare team or the public who have just saved a life. CPR is also unique in that its core principles are well understood by many, or at least that’s what we think. It consists of clearing a patient’s airway, rhythmically pumping the thorax to circulate blood around the body, providing some ventilation to replace spontaneous breathing, and if a shockable arrhythmia is present, performing defibrillation.

Initial research into CPR yielded major gains. Campaigns to convert bystanders into competent responders have been success stories of major proportions, leading to marked improvements in rates of survival with good neurological outcome (1–3). However, recent clinical research into CPR has not improved its effectiveness. Large randomized controlled trials have tested the optimal types and doses of inotropic agents, bicarbonate, or anti-arrhythmic agents without identifying beneficial treatments (4–6). Perhaps the most striking (and highly powered) recent randomized controlled trial was the CCC (Continuous Chest Compression) study, which roughly translates into: ventilation is not sufficiently important to warrant interruption of CPR (7).

A novel insight in this issue of the Journal by Grieco and colleagues (pp. 728–737) has the potential to move beyond this impasse (8). The investigators began with careful observations of end-tidal CO2 (ETCO2) tracings during CPR in victims of cardiac arrest. Because exhaled CO2 reflects delivery of venous blood to the lungs, ETCO2 tracings are recommended to monitor the effectiveness of CPR. But during CPR, the ETCO2 tracing is characterized by oscillations; this is assumed to represent variable ventilation and exhalation of CO2. However, in many patients, these CPR-related oscillations were inconsistent or absent, presumably reflecting obstruction to airflow that prevented passage of the CPR-related CO2 oscillations to the ETCO2 monitor at the end of the endotracheal tube.

Because large airways were patent in each patient (endotracheal tubes were present in all), the investigators hypothesized that the airflow obstruction occurred in small airways. To quantify the degree of airway closure, they developed an Airway Opening Index, representing the changing ETCO2 values during CPR compared with the maximal ETCO2. Thus, a higher index reflects greater transmission of CO2 and greater airway patency.

Two sets of experiments confirmed that small airway closure could explain this phenomenon. Using a bench lung model, they reproduced the ETCO2 waveforms observed in the real patients by simulating airway closure, and demonstrated that incremental levels of positive end-expiratory pressure (PEEP) increased transmission of the oscillations, as well as elevating inspiratory flow and minute ventilation. In addition, the highest ETCO2 value during CPR represented the closest estimate of the actual alveolar CO2. Each of these findings was recapitulated in human cadavers (Thiel model), where the alveoli were loaded with CO2 before CPR; although PEEP improved transmission of oscillations, elevating the airway index, it did not compromise the effect of chest compressions on intrathoracic pressure.

If upheld, these findings could have immense application to the conduct of CPR in several direct ways. First, although perfusion with poorly oxygenated blood is preferable to no perfusion, without at least some ventilation, the perfusing blood will ultimately contain very little oxygen. Given that oxygenating tissues is the primary function of circulating blood, how could it be that ventilation during CPR seems not to be important? The idea that small airways close during CPR and prevent alveolar delivery of the applied breath could explain this paradox, because if the airways could be opened by titrated PEEP, the importance of ventilation during CPR could be properly evaluated. Second, it is possible that many trials reporting failure of carefully considered therapies (e.g., inotropic agents, antiarrhythmic medications, and even advanced airway techniques [9]) may represent false-negative results because ventilation was inadequate in so many patients. This is because if return of circulation does not occur rapidly, then ongoing CPR with inadequate ventilation would render any cointervention ineffective. Third, measuring the Airway Opening Index (or a version of it) should be easy, given that ETCO2 monitors are now widely available, including in ambulances. Thus, small airway closure during CPR could be detected and eliminated with small levels of PEEP.

Although this study, and another promising to increase perfusion during CPR using inhaled nitric oxide (10), are grounds for optimism, there are important reasons to be cautious. Several steps are required to corroborate this theory of small airway closure. The results need to be reproduced by others, and the nature and locus of the airway closure need to be better understood, perhaps using novel imaging. Although the cadaver studies were reassuring about the lack of adverse effects of PEEP, the optimal level of PEEP, balancing airway patency against perfusion, needs to be understood in individual patients. Finally, clinical trialists in the field will need all of their accumulated insight and experience to guard against a false-negative (or, far less likely, a false-positive) controlled trial.

In conclusion, CPR is a very special gift. Research in the field made early gains, but more recently progress has been slow. The physiologic insight by Grieco and colleagues has the potential to make a major positive difference to the field: If treating small airway closure works, survival after CPR might increase, and more important, so too might the quality of life among survivors.