Reappraisal of Mouth-to-Mouth Ventilation During Bystander-Initiated CPR
To the Editor:
The “reappraisal” of the literature on mouth-to-mouth ventilation during bystander-initiated CPR, by a working group of the Basic Life Support and Pediatric Life Support subcommittees of the American Heart Association (AHA),1 is misleading and incomplete. There is no convincing evidence that the low incidence of initiation of CPR out of hospital by lay bystanders is the result of fear of becoming infected by mouth-to-mouth ventilation. Such fear should not be promoted. If such fear exists, however, it should be mitigated by explaining that initiating CPR is safe and by carrying a pocket-size barrier for ventilation of strangers. The errors in this article concerning behavioral, educational, epidemiological, and logistics issues will be summarized in a separate letter by Braslow and Brennan.
Although the article says “… it is not intended to change any current AHA recommendations,” its publication has created confusion and the erroneous impression for laypersons and the media that in sudden coma, bystanders will save lives by merely pushing on the sternum (step C, circulation support). In cardiac arrest, oxygenated blood must be circulated to restore heartbeat and to keep the brain viable, requiring “head tilt plus blowing plus pumping.” The article suggests that mouth-to-mouth ventilation can be omitted in various forms of sudden loss of consciousness.1 Laypersons cannot differentiate between various forms of sudden coma and between the absence versus presence of a weak pulse. Coma always results in upper airway obstruction if the neck is flexed (references 26 to 31 in the article by Becker et al),2 3 4 5 6 as experienced by anesthesiologists every day. There are 20 million general anesthesias given in the United States each year. The data in Figure 1 are misleading1 because Gordon’s measurements of 1950 (reference 24 in the article by Becker et al) were made via tracheal tubes.
The omission of lifesaving step A (upper airway control by backward tilt of the head; reference 26 in the article by Becker et al)2 3 4 5 6 and step B (breathing support by mouth-to-mouth; references 27 and 30 in the article by Becker et al) will harm not only victims of sudden cardiac death but also those of trauma (particularly head trauma with impact apnea), intoxication, asphyxiation (common in children), and other causes of sudden coma with or without pulselessness. Indeed, step A was not even suggested by the authors1 during sternal compressions alone; at least the shoulders should be elevated so the head assumes spontaneously a backward-tilted position that might support a patent hypopharyngeal air passage (references 26 to 31 in the article by Becker et al).3 If there is still a faint pulse, steps A and B alone can often reverse the dying process. Even in cases of out-of-hospital sudden ventricular fibrillation, reoxygenation should precede countershocks, which will not result in heartbeat after >2 to 3 minutes of untreated ventricular fibrillation (reference 33 in the article by Becker et al).
This article’s “historical rationale” includes errors and omissions.1 For comatose humans with pulse and a natural air passage, the failure of chest-pressure methods to ventilate was documented in 1957.3 For humans both with or without pulse, the need to combine steps A, B, and C was documented in 19602 : in 30 anesthetized and curarized adult patients and in 12 patients with cardiac arrest, ventilation produced by sternal compressions (step C) alone was measured, with or without backward tilt of the head, and with or without tracheal tube. In all 30 supine, horizontal, curarized patients, the unsupported head assumed a semiflexed position that resulted in zero tidal volumes by sternal compressions. With backward tilt of the head by elevation of the shoulders (a measure ignored in the article by Becker et al), tidal volumes were zero in 16 and only 25 to 200 mL in 14 patients.2 In all 12 patients with cardiac arrest, in spite of tracheal tube in place, forceful sternal compressions alone moved essentially no ventilation.2 The authors listed this article (reference 31) but might not have studied these, the only data published thus far on ventilation produced by sternal compressions in humans,2 because that article predates the earlier citations in MEDLINE. If chest pressure alone ventilates via an open airway, it would be with unpredictably inadequate tidal volumes below resting lung volume, which causes lung collapse, particularly in terminal patients with congested lungs. Hence, not elastic recoil of the chest, but positive-pressure inflations are needed to ventilate and oxygenate.
In discussing ventilation requirements, we agree with the article by Becker et al that low minute ventilation is sufficient to normalize arterial Po2 and Pco2 during the low blood flow produced by external CPR in patients with sudden cardiac arrest. However, reoxygenation in asphyxiation (reference 29 of the article by Becker et al) and ventilation after restoration of spontaneous circulation, both with high blood flow, require hyperventilation, which is possible even with exhaled air (references 27, 29, and 30 of the article by Becker et al). Also, the authors’ concern about gastric insufflation by mouth-to-mouth ventilation1 is exaggerated. Regurgitation can also occur with sternal compressions alone, and gastric insufflation during mouth-to-mouth ventilation is self-limiting (reference 29 of the article by Becker et al). Furthermore, the presently recommended ventilation:compression ratios of 2:15 or 1:5, which is a compromise, were based on a study in dogs,7 which was not quoted.
Animal data of ventilation by chest compressions alone, the main argument presented for the “reappraisal,” have no clinical relevance. Rats, pigs, and dogs have straight airways that do not obstruct, even in the absence of a tracheal tube. Humans have kinked airways that always obstruct in the absence of backward tilt of the head in coma (references 26 to 31 in the article by Becker et al).2 3 4 5 6 Also, the thoraces of animals are more compliant and have greater elastic recoil. Furthermore, the authors’ admission that after 6 minutes, ventilation is needed1 is a concern, because response times for paramedics are usually longer than that.
In sudden ventricular fibrillation cardiac arrest with healthy lungs, arterial oxygen levels indeed remain near normal during a long period of no blood flow without CPR.8 This occurrence does not obviate the need for mouth-to-mouth ventilation, because sternal compressions alone in the presence of complete airway obstruction recirculate venous blood and cause arterial deoxygenation within <60 seconds.8
Clinical studies by the excellent CPCR registry of Belgium (references 38 and 39 in the article by Becker et al) have been overinterpreted. The fact that 10% of prehospital cardiac arrest patients survived among those who were found on arrival of the ambulance with bystanders doing only chest compressions does not prove that mouth-to-mouth ventilation is not needed; it may have been used earlier or was not necessary because of short arrest times with continued breathing (gasping). Moreover, the efficacy of individual resuscitation measures must be determined by physiological measurements in humans, rather than clinical, epidemiological, statistically significant outcome correlations, which do not prove cause-effect relationships on mechanisms.
What and how to teach the public must be extremely simple, with only 1 sequence of steps, ie, A-B-C. We agree with the authors’ first recommendation1 that CPR guidelines should not be changed at this time. Their second recommendation, to restudy details of mouth-to-mouth ventilation, is laudable but has low priority. Higher priority should be given to promoting the motivation and skill acquisition of the public with use of media and self-training systems; to early automatic external defibrillation by first responders; and to clinical documentation of cerebral resuscitation with mild hypothermia and blood pressure support. We strongly disagree with their third recommendation1 that “clinical trials of chest compressions without mouth-to-mouth ventilation are ethical,” for the obvious reason that randomly withholding ventilation and thereby condoning the moving of deoxygenated blood would be irresponsible.
- Copyright © 1998 by American Heart Association
Becker LB, Berg RA, Pepe PE, Idris AH, Aufderheide TP, Barnes TA, Stratton SJ, Chandra NC. A reappraisal of mouth-to-mouth ventilation during bystander-initiated cardiopulmonary resuscitation: a statement for healthcare professionals from the Ventilation Working Group of the Basic Life Support and Pediatric Life Support Subcommittees, American Heart Association. Circulation. 1997;96:2102–2112.
Safar P, Brown TC, Holtey WH, Wilder R. Ventilation and circulation with closed chest cardiac massage in man. JAMA. 1961;176:574–576.
Safar P. The failure of manual artificial respiration. J Appl Physiol. 1959;4:84–88.
Elam JO, Greene DG, Schneider MA, Ruben HM, Gordon AS, Hustead RF, Benson DW, Clements JA, Ruben A. Head-tilt method of oral resuscitation. JAMA. 1960;172:812–815.
Harris CS, Kirimli B, Safar P. Ventilation-cardiac compression rates and ratios in cardiopulmonary resuscitation. Anesthesiology. 1967; 28:806–813.
Lesser R, Bircher N, Safar P, Stezoski W. Sternal compression before ventilation in cardiopulmonary resuscitation. Prehospital and Disaster Medicine. 1985;1(suppl I):239–241.
We appreciate the thoughtful comments from Dr Safar and his colleagues regarding the role of mouth-to-mouth ventilation in adult CPR. The world will remain forever in the debt of James Elam, MD, who rediscovered the value of mouth-to-mouth ventilation in the 1950s, and Dr Safar, who meticulously documented the superiority of expired air ventilation to maintain blood oxygen levels. The combination of the Safar-Elam techniques with the chest compression rediscovered by Kouwenhoven, Jude, and Knickerbocker ushered in the modern era of CPR in the 1960s and has saved literally thousands of lives.
By the 1990s, however, a number of questions had arisen about CPR: Why is CPR so difficult for laypeople to learn and remember? Why is CPR started on such a small percentage of witnessed cardiac arrests? Is poorly performed CPR perhaps capable of doing harm? Are there different causes of cardiac arrest that merit modifications of the traditional “pump and blow” of basic CPR?
Our Ventilation Working Group started its work with open minds. For this report, we decided to address the particular CPR feature of mouth-to-mouth ventilation. This focus was because of indirect evidence suggesting that the disagreeable features of mouth-to-mouth ventilation, combined with a growing fear of disease transmission, were making people reluctant to start CPR. Adding mouth-to-mouth ventilation to the task of chest compression may create a complex psychomotor skill that is simply too difficult for lay rescuers to remember at the time of an event as dramatic as sudden cardiac arrest.
We decided to review the entire body of research regarding chest compression and ventilation in CPR. We noted a scarcity of human research on this topic and think there are areas that merit further investigation. The purpose of the article was to stimulate interest and promote CPR research. We reached the following conclusions:
1. Two-person CPR with chest compression and mouth-to-mouth ventilation, performed by experienced professionals, is indisputably the most effective method to maintain brain and heart viability and to prolong ventricular fibrillation during cardiac arrest.
2. Single-person CPR with chest compression and mouth-to-mouth ventilation, performed by inexperienced laypeople, is probably less effective but should be taught without reservation in all AHA CPR courses.
3. We completely reject the hypothesis that mouth-to-mouth ventilation should be eliminated from CPR training. This conclusion is a distortion of our work and is without valid scientific support.
4. Single-person CPR with chest compression but without mouth-to-mouth ventilation is suboptimal and is not recommended by the AHA. However, the optimal single rescuer technique for lay rescuers is not known because it must both “optimize” the physiology of promoting oxygenated blood flow as well as promote maximal community-wide implementation (which may be affected by ease of training, skill retention, willingness to act, and ability). Animal studies suggest that mouth-to-mouth plus chest compression improves oxygen saturation. But as the single rescuer switches from compression to ventilation, fewer compressions are performed, which may adversely impact circulation of blood and myocardial perfusion. The obtainment of additional knowledge on these competing processes seems important, and these processes have not yet been well studied.
5. Human data on lay single-rescuer CPR without mouth-to-mouth ventilation are scant. The data from the Belgian CPR registry and reports from the Netherlands fail to find any disadvantage for patients who received compression but not mouth-to-mouth ventilation. Moreover, our BLS instructors need practical guidance when confronted with a CPR learner they are certain will not perform mouth-to-mouth ventilation in an emergency. In such situations, the CPR learner must be encouraged to “at least do something.” There is good evidence that chest compression alone is far better than no rescue attempts at all.
6. We found a theme in the research regarding adults who collapse with sudden fibrillation and well-oxygenated blood. These individuals compose a significant proportion of adult out-of-hospital arrests and may maintain acceptable blood oxygen levels for a few minutes after collapse with chest compression alone. We could hypothesize that mouth-to-mouth ventilation could be delayed for these patients for a few minutes. This window of time, however, is limited, and ventilation will eventually be required. These data suggest an avenue for future research. We encourage creative research on this hypothesis; it is ethical if properly designed.
7. Further research is needed to improve community CPR rates. We cannot expect lay rescuers to make judgments about the cause of collapse nor to switch from 1 rescue sequence to another based on the passage of time. For successful community CPR, the techniques must be easy to learn, remember, and perform during a crisis. We have called for more research because we agree that our current data do not provide complete answers to some important questions. We believe that the public is best served by conducting research to answer unresolved issues.
8. Finally, readers should not mistake “confusion” with legitimate controversy. That letters with divergent viewpoints are printed reflects the fact that scientific controversy still exists. In our deliberations, we adhere to the principles of evidence-based decision making. CPR has been advocated for nearly 40 years, yet most victims of cardiac arrest never receive CPR efforts, and survival rates are poor. These controversies have existed for too long without resolution. Let us resolve to perform the research to turn scientific controversy into consensus among scientists. This duty falls to all of us as guardians of public health and safety.
The Safar Resuscitation Research Center has been a beacon of quality research on the topic of CPR. We appreciate their comments and their interest in this topic. We are sure they join with us in encouraging more research evaluation in this rich and dynamic area of resuscitation.