Boletín Médico de Mixcoac: In CPR, Less May Be Better

Fifty years have passed since Kouwenhoven, Jude, and Knickerbocker1 proposed external chest compression to provide circulation of blood to the brain and heart after cardiac arrest. Shortly thereafter, mouth-to-mouth rescue breathing was adopted as an essential addition to this lifesaving procedure. Since that time, there has been very little fundamental change in the method or manner of cardiopulmonary resuscitation (CPR). Decades of observational studies have shown that survival is improved if CPR is performed by bystanders rather than being provided only when emergency medical services (EMS) staff arrives. The use of automated external defibrillators by bystanders and the use of in-hospital hypothermia in comatose patients have also been found to improve outcomes in patients with cardiac arrest.

Only relatively recently, however, have the fundamentals of the initial resuscitation been investigated. Focused, impressive laboratory research has resulted in a surge of interest in these fundamentals. Most of the studies have involved pigs with electrically induced ventricular fibrillation and have resulted in two distinct conclusions. First, the interruption of chest compression results in a lower coronary perfusion pressure and presumably less myocardial blood flow during CPR.2 Second, an increased frequency of positive-pressure ventilation reduces the survival rate. The detrimental effect of this high frequency of ventilation is thought to result from both the interruption of compression2and the obstruction of venous return to the central circulation because of high intrathoracic pressure during ventilation.3

These data, along with observational EMS field studies suggesting that outcomes are better with continuous chest compression and no rescue breathing, led the American Heart Association to advocate “hands-only” CPR4 for bystanders not trained or competent in CPR with rescue breathing.

Two articles in this issue of the Journal — one by Rea and colleagues5 and the other by Svensson and colleagues6 — describe studies that took advantage of emergency medical dispatchers' instructing bystanders to administer CPR. In the studies, patients with out-of-hospital cardiac arrest were randomly assigned to undergo one of two types of CPR performed by a bystander: either continuous chest compression without any attempts at ventilation, or chest compression with interruptions for rescue breathing by bystanders (the current standard).

The straightforward conclusion from the primary analyses of these studies is that continuous chest compression without active ventilation, which is simpler to teach and perform, results in a survival rate similar to that with chest compression with rescue breathing. Equally straightforward is the message that advocating continuous chest compression without ventilation by a bystander should increase the frequency of bystanders' effectively performing CPR and therefore increase the chances of survival after cardiac arrest. Performance of mouth-to-mouth rescue breathing is far more difficult than proper chest compression, and rescue breathing may be viewed with distaste and raise concerns about risks associated with mouth-to-mouth contact. One suggestion made by Rea and associates in their discussion deserves some attention: that mouth-to-mouth ventilation is performed so poorly by bystanders that this periodic interruption for “ventilation” succeeds solely in diminishing coronary flow.2 Nonetheless, CPR courses should teach rescue breathing, since it is important in cases of cardiac arrest due to obvious respiratory failure, which include most cardiac arrests in children and some in adults.

There was a trend toward better survival with continuous chest compression among patients whose arrests were due to cardiac causes5 and among patients whose initial cardiac rhythm was ventricular tachycardia or fibrillation6 rather than asystole or electromechanical dissociation. This trend is consistent with the benefit of continuous compression in animal models, in which arrest is caused by induced ventricular tachycardia or fibrillation.2 In dogs, after the sudden interruption of blood flow by means of ventricular fibrillation, the predicted steep decline in arterial oxygen saturation does not occur until many minutes after the start of resuscitation. The volume of oxygen in the lungs is relatively great when arrest occurs suddenly.7

On the other hand, it might be detrimental not to provide rescue breathing in patients with other causes of cardiac arrest. Oxygenation may be more compromised over longer periods of declining circulation, such as when there is hypotension resulting in electromechanical disassociation or a prolonged period of bradycardia before asystole.

An even more radical suggestion is that in patients with sudden cardiac arrest caused by ventricular tachycardia or fibrillation, it may be beneficial to restore circulation with blood that is moderately unsaturated with oxygen rather than with well-oxygenated blood. Although in animal models of ventricular fibrillation, a prolonged period (i.e., 8 minutes) without ventilation is detrimental,8 recent observational clinical studies support the idea that hyperoxia during recovery from cardiac arrest is detrimental to patients.9 Studies of isolated cardiac-tissue specimens10 have raised the possibility that initial reperfusion with hypoxemic blood may result in fewer injurious oxygen free radicals and less reperfusion injury.

These are intriguing observations and hypotheses, but more research is needed. To my knowledge, definitive studies in animals of hypoxic reperfusion after ventricular fibrillation have not yet been performed. Although clinical observational studies support the relative benefit of compression-only CPR performed by EMS personnel on arrival, there is a need for a randomized study in this setting as well. Particular attention needs to be paid to whether the benefits of compression-only CPR are apparent primarily in cardiac arrest with initial ventricular tachycardia or fibrillation rather than arrest with other rhythms. It will also be important to study the subgroup of patients who undergo bystander-performed compression-only CPR followed by EMS-performed compression-only CPR. There is certainly a need for rescue breathing after a prolonged period of CPR; in such cases, should the rescue breathing be provided while continuous chest compression is performed or should compression be interrupted to provide greater ventilation? A large study might also address whether rescue breathing is of critical importance in patients whose cardiac arrest is not associated with ventricular tachycardia or fibrillation. The answers to these questions may be paradigm-shifting and will almost certainly be lifesaving.

Disclosure forms provided by the author are available with the full text of this article at NEJM.org.