Ventilation de l'arrêt cardiaque : Pour par Michel RAMAKERS

Faut-il encore ventiler l’arrêt cardiaque ?

Pour

M Ramakers Praticien Hospitalier Service de Réanimation Polyvalente CH Mémorial, Saint Lô

Le souffle c’est la vie……

Merci de votre attention

Réanimation Cardio Pulmonaire de base (RCPB) :

vers l’abandon de la ventilation ? Pourquoi ?

La ventilation : un progrès

RESUSCITATION Liss

FIGURE 1. Warm ashes, burning ex- crement, or hot water applied to the victim's abdomen were thought to be beneficial in restoring heat and life to the body. Figures 1 through 8 are re- produced with permission of the Mu- seum of Science and Industry, Chi- cago, Illinois.

FIGURE 2. Whipping the victim with stinging nettles was considered helpful in "waking" him from a "'deep sleep."

Other methods were developed in the 1700s in response to a growing number of deaths by drowning. Inver- sion (Figure 4), practiced in Egypt 3,500 years ago, was popular in Europe and the New World. The victim was hung by his feet, with chest pressure to aid expiration and pressure release to aid inspiration.

Another resuscitation method involved placing the victim across a trotting horse, tt was hoped that the rhythmic compression of the victim's chest as his body bounced would re- store breathing. In the waterfront vari- ation of this method, the victim was placed over a barrel (Figure 5). Rolling the barrel forward released pressure on the victim's chest, allowing inspiration to occur. Rolling it backward re- turned the weight of the body to the chest, causing compression and ex- halation. Sometimes a vict im was

placed inside a barrel which was rolled to aid ventilation.

Because of the increase in deaths by drowning during this time, societies were formed to organize efforts at resuscitation. England's Royal Humane Society, founded in 1774, was preceded by the Socie ty for Recovery of Drowned Persons, which began in Amsterdam in 1767. Dutch recommendations 8 included the following: 1) warming the victim, which often required transporting him from the scene of the drowning, but could be accomplished by lighting a fire near the victim, burying him in warm sand, putting him in a warm bath, or placing him in bed with one or two volunteers; 2) removing swallowed or aspirated water by positioning the victim's head lower than his feet and applying manual pressure to his abdomen; vomit ing was induced by tickling the victim's throat with a feather; 3) stimulating the victim, particularly the lungs, stomach, and in- testines, by such means as rectal and oral fumigation with tobacco smoke. Snuff was blown into the vict im's nose, or mixtures of salt, oil, and water were squirted into his mouth; 4) restoring breath was attempted with a bellows or a mouth-to-mouth method; and 5) bloodletting.

All these methods had been In use for some time, as illustrated by the report of the hanging, resuscitation, and recovery of Anne Green in 1650. 9

Over the years, however, all except warming were discarded, largely as a result of the research of Benjamin Brodie in England and Leroy d'Etiolles in France.

Brodie denounced fumigation in 1811 after demonstrat ing that four ounces of strong tobacco would kill a dog, and one ounce would kill a cat.1 Ten years later, in a lecture on asphyx- ia, he noted that two to three minutes after breathing ceases the heart stops beating, after which no method of artificial ventilation is of any value. He believed that patients who recovered did so whether or not artificial ventilation was given, and he thought that warming the victim was the most important factor in resuscitation.

In 1829, Leroy d'Etiolles 1 demonstrated that overdis tent ion of the lungs by a bellows could kill an animal easily, and this method was discontinued. Mouth-to-mouth resuscita- t ion was not popular; therefore, warming the victim became the main- stay of resuscitation efforts.

PULMONARY RESUSCITATION In the same paper in which he

noted the dangers of the bellows, d'Etiolles also suggested a manual method of ventilation of applying pressure to the chest and abdomen. 1 In 1831, Dalrymple reviewed the technique and stated that side-to-side compression of the vict im's chest, either by hand or by two operators

118/66 Annals of Emergency Medicine 15:1 January 1986

•  La base de la ressuscitation fût pendant des siècles de réchauffer le corps de la victime et de le « stimuler » physiquement par des méthodes plus ou moins barbares

•  En 1892 des auteurs Français recommandaient de tirer fortement et de façon rythmique sur la langue

•  Différentes techniques de ventilation artificielles sont décrites au début du XIXème siècle mais ne connaîssent pas un grand succès

RESUSCITATION Liss

FIGURE 1. Warm ashes, burning ex- crement, or hot water applied to the victim's abdomen were thought to be beneficial in restoring heat and life to the body. Figures 1 through 8 are re- produced with permission of the Mu- seum of Science and Industry, Chi- cago, Illinois.

FIGURE 2. Whipping the victim with stinging nettles was considered helpful in "waking" him from a "'deep sleep."

Other methods were developed in the 1700s in response to a growing number of deaths by drowning. Inver- sion (Figure 4), practiced in Egypt 3,500 years ago, was popular in Europe and the New World. The victim was hung by his feet, with chest pressure to aid expiration and pressure release to aid inspiration.

Another resuscitation method involved placing the victim across a trotting horse, tt was hoped that the rhythmic compression of the victim's chest as his body bounced would re- store breathing. In the waterfront vari- ation of this method, the victim was placed over a barrel (Figure 5). Rolling the barrel forward released pressure on the victim's chest, allowing inspiration to occur. Rolling it backward re- turned the weight of the body to the chest, causing compression and ex- halation. Sometimes a vict im was

placed inside a barrel which was rolled to aid ventilation.

Because of the increase in deaths by drowning during this time, societies were formed to organize efforts at resuscitation. England's Royal Humane Society, founded in 1774, was preceded by the Socie ty for Recovery of Drowned Persons, which began in Amsterdam in 1767. Dutch recommendations 8 included the following: 1) warming the victim, which often required transporting him from the scene of the drowning, but could be accomplished by lighting a fire near the victim, burying him in warm sand, putting him in a warm bath, or placing him in bed with one or two volunteers; 2) removing swallowed or aspirated water by positioning the victim's head lower than his feet and applying manual pressure to his abdomen; vomit ing was induced by tickling the victim's throat with a feather; 3) stimulating the victim, particularly the lungs, stomach, and in- testines, by such means as rectal and oral fumigation with tobacco smoke. Snuff was blown into the vict im's nose, or mixtures of salt, oil, and water were squirted into his mouth; 4) restoring breath was attempted with a bellows or a mouth-to-mouth method; and 5) bloodletting.

All these methods had been In use for some time, as illustrated by the report of the hanging, resuscitation, and recovery of Anne Green in 1650. 9

Over the years, however, all except warming were discarded, largely as a result of the research of Benjamin Brodie in England and Leroy d'Etiolles in France.

Brodie denounced fumigation in 1811 after demonstrat ing that four ounces of strong tobacco would kill a dog, and one ounce would kill a cat.1 Ten years later, in a lecture on asphyx- ia, he noted that two to three minutes after breathing ceases the heart stops beating, after which no method of artificial ventilation is of any value. He believed that patients who recovered did so whether or not artificial ventilation was given, and he thought that warming the victim was the most important factor in resuscitation.

In 1829, Leroy d'Etiolles 1 demonstrated that overdis tent ion of the lungs by a bellows could kill an animal easily, and this method was discontinued. Mouth-to-mouth resuscita- t ion was not popular; therefore, warming the victim became the main- stay of resuscitation efforts.

PULMONARY RESUSCITATION In the same paper in which he

noted the dangers of the bellows, d'Etiolles also suggested a manual method of ventilation of applying pressure to the chest and abdomen. 1 In 1831, Dalrymple reviewed the technique and stated that side-to-side compression of the vict im's chest, either by hand or by two operators

118/66 Annals of Emergency Medicine 15:1 January 1986

Mais tout ne fût pas une réussite……..

All these studies were flawed, however, because the subjects were intu- bated. A number of articles in 1958 cited the inadequacy of all m a n u a l t echniques in the absence of endotracheal i n t u b a t i o n because the airway was no t effect ively pa ten t in e i t h e r the p rone or s u p i n e position. 33"36 These studies demonstrated the superiority of mouth-to-mouth resuscitation. The technician used his hands to mainta in an open airway, and exhaled air was found to be safe and effective for ven t i l a t ing a person in vent i la tory arrest. 37 Keith had pre- saged this development in 1909, when he stated the following:

My mind is also open to the con- viction that the ancient method of mouth-to-mouth insufflation with expiratory compression of the chest may not prove more effective than either; at least, if it should happen fhat I may be found in an apparently drowned condition, I sincerely hope that my rescuer will apply this prompt method to me as

15:1 January 1986

my first aid. It is air that my lungs and blood then will stand urgently in need of, not pressure, for if the pulmonary circulation has ceased, such pressure is, upon the evidence at present at our disposal, more likely to weaken than to strength- en the heart. With the patient in the prone position, the operator will have great difficulty in know- ing whether or not air is entering and leaving the lungs freely; with direct inflation one knows the effect immediately by placing the hands on the epigastrium; the hand is also needed there to produce expiration. 1

Fifty years after Keith's comments, in- ves t iga to r s f i na l ly had r ecogn ized mouth- to -mouth resuscitat ion as the most effective means of artificial ventilation without an artificial airway.

CARDIOPULMONARY RESUSCITATION

Closed-chest massage was the next

Annals of Emergency Medicine

FIGURE 6, The Dalrymple method.

FIGURE 7. The Marshall Hall method.

FIGURE 8. The Schafer prone pressure method used pressure applied to the victim's back, which forced the abdomen against the diaphragm and caused expiration. Inspiration occurred when the pressure was released.

major advance in cardiopulmonary resuscitation. Its introduction in 196038 e l imina ted the need for open-chest massage, which rarely was successful o u t s i d e t he o p e r a t i n g room.39, 40 Closed-chest massage, which could be performed virtually anywhere, became very popular and was endorsed by the A m e r i c a n Hea r t Assoc i a t i on , the American National Red Cross, the In- dustrial Medical Association, and the Uni ted States Public Health Service. Chest compression to pump the vic- t im's blood replaced chest compression to ventilate the victim.

In 1966 the Ad Hoc Committee on Cardiopulmonary Resuscitation of the Na t iona l Academy of Sciences and National Research Council published its first s tatement, establishing CPR as we know it today. 41 Although this committee cited mouth-to-mouth re- susc i ta t ion as superior to all other m a n u a l t e c h n i q u e s , the Silvester, Nielsen, and back pressure/hip lift techniques still were recommended if the m o u t h - t o - m o u t h method could not be used. The committee also rec- o m m e n d e d o p e n - c h e s t massage if c losed-chest massage could no t be done.

69/121

Ø  1831 Darlympe propose de passer un large bandage derrière le patient puis de le croiser sur la poitrine

Ø  1856 Marshall hall : déplacement de la victime 16 fois par minute de l’estomac (expiration) sur le côté (inspiration)

Ø  1878 Benjamin Howard : compression postérieure initiale de la victime puis compression des dernières côtes en décubitus dorsal

All these studies were flawed, however, because the subjects were intu- bated. A number of articles in 1958 cited the inadequacy of all m a n u a l t echniques in the absence of endotracheal i n t u b a t i o n because the airway was no t effect ively pa ten t in e i t h e r the p ro n e or s u p i n e position. 33"36 These studies demonstrated the superiority of mouth-to-mouth resuscitation. The technician used his hands to mainta in an open airway, and exhaled air was found to be safe and effective for ven t i l a t ing a person in vent i la tory arrest. 37 Keith had pre- saged this development in 1909, when he stated the following:


15:1 January 1986











69/121

All these studies were flawed, however, because the subjects were intu- bated. A number of articles in 1958 cited the inadequacy of all m a n u a l t echniques in the absence of endotracheal i n t u b a t i o n because the airway was no t effect ively pa ten t in e i t h e r the p r o ne or s u p i n e position. 33"36 These studies demonstrated the superiority of mouth-to-mouth resuscitation. The technician used his hands to mainta in an open airway, and exhaled air was found to be safe and effective for ven t i l a t ing a person in vent i la tory arrest. 37 Keith had pre- saged this development in 1909, when he stated the following:


15:1 January 1986











69/121

Et enfin….

Ø « Il monta, et se coucha sur l'enfant; il mit sa bouche sur sa bouche, ses yeux sur ses yeux, ses mains sur ses mains, et il s'étendit sur lui. Et la chair de l'enfant se réchauffa. »

Ø 1732 : réanimation d’un mineur

Ø 1802 : 500 cas de réanimation de nouveaux nés

Ø Technique délaissée (voire méprisée) par les Médecins…

Ø 1950 travaux de J Elam

Ø 1958 Research Council of the National Academy of sciences : recommande le bouche-à-bouche comme la technique de choix

European Resuscitation Council Guidelines for Resuscitation 2010 Section 2. Adult basic life support and use of automated external

defibrillators

R.W. Koster et al. / Resuscitation 81 (2010) 1277–1292 1281

Fig. 2.9. Interlock the fingers of your hands. Keep your arms straight.

If your initial rescue breath does not make the chest rise asin normal breathing, then before your next attempt:

• look into the victim’s mouth and remove any obstruction;• recheck that there is adequate head tilt and chin lift;• do not attempt more than two breaths each time before

returning to chest compressions.If there is more than one rescuer present, another res-

cuer should take over delivering CPR every 2 min to preventfatigue. Ensure that interruption of chest compressions isminimal during the changeover of rescuers. For this purpose,and to count 30 compressions at the required rate, it maybe helpful for the rescuer performing chest compressionsto count out loud. Experienced rescuers could do combinedtwo-rescuer CPR and in that situation they should exchangeroles/places every 2 min.

6b. Chest-compression-only CPR may be used as follows:• If you are not trained, or are unwilling to give rescue breaths,

give chest compressions only.• If only chest compressions are given, these should be con-

tinuous, at a rate of at least 100 min−1 (but not exceeding120 min−1).

7. Do not interrupt resuscitation until:• professional help arrives and takes over; or• the victim starts to wake up: to move, open eyes and to

breathe normally; or• you become exhausted.

Opening the airway

The jaw thrust is not recommended for lay rescuers becauseit is difficult to learn and perform and may itself cause spinal

Fig. 2.10. Press down on the sternum at least 5 cm.

movement.49 Therefore, the lay rescuer should open the airwayusing a head-tilt-chin-lift manoeuvre for both injured and non-injured victims.

Recognition of cardiorespiratory arrest

Checking the carotid pulse (or any other pulse) is an inaccu-rate method of confirming the presence or absence of circulation,both for lay rescuers and for professionals.50–52 There is, however,no evidence that checking for movement, breathing or cough-ing (“signs of a circulation”) is diagnostically superior. Healthcare

Fig. 2.11. Blow steadily into his mouth whilst watching for his chest to rise.

Et si ?

Reluctance of Internists and Medical Nurses to perform mouth-to-mouth ventilation BE Brenner, Arch Int Med, 1993

No. contacted

Responding, No.(%)

%

Unkwnown Trauma Child Gay Elderly

Resident 82 81(99) 54 36 99 21 64

Staff physisian 510 352(69) 57 60 81 16 55

Registrered Nurse 112 96(86) 20 32 75 10 33

Ø  Résidents, Médecins, Infirmières

Ø  Questionnaire avec différents scénarios

Ø  Volonté de réaliser du bouche-à-bouche dans ces différentes situations

Ø  Risque de contracter une infection

Ø  Peur d’une procédure en justice

Copyright 1995 by the American Medical Associa9on. All Rights Reserved. Applicable FARS/DFARS Restric9ons Apply to Government Use. American Medical Associa9on, 515 N. State St, Chicago, IL 60610. ??diteur American Medical Associa9on.

2

Bystander Cardiopulmonary Resuscita4on: Concerns About Mouth-‐to-‐Mouth Contact. Locke, Catherine; Berg, Robert; Sanders, Arthur; Davis, Melinda; MA, MEd; Milander, Melinda; Kern, Karl; Ewy, Gordon Archives of Internal Medicine. 155(9):938-‐943, May 8, 1995.

Figure 1 . Percentage of respondents "definitely" or "probably" willing to perform cardiopulmonary resuscita9on (CPR) with strangers using different CPR techniques. CC+V indicates chest compressions plus mouth-‐to-‐mouth ven9la9on; CC, chest compressions alone

Copyright 1995 by the American Medical Associa9on. All Rights Reserved. Applicable FARS/DFARS Restric9ons Apply to Government Use. American Medical Associa9on, 515 N. State St, Chicago, IL 60610. ??diteur American Medical Associa9on.

3

Figure 2 Bystander Cardiopulmonary Resuscita4on: Concerns About Mouth-‐to-‐Mouth Contact. Locke, Catherine; Berg, Robert; Sanders, Arthur; Davis, Melinda; MA, MEd; Milander, Melinda; Kern, Karl; Ewy, Gordon Archives of Internal Medicine. 155(9):938-‐943, May 8, 1995.

Figure 2 . Percentage of respondents "definitely" or "probably" willing to perform cardiopulmonary resuscita9on (CPR) with friends or rela9ves using different CPR techniques. CC+V indicates chest compressions plus mouth-‐to-‐mouth ven9la9on; CC, chest compressions alone

Attitudes toward the performance of bystander cardiopulmonary resuscitation in Japan T Taniguchi, Resuscitation, 2007 Attitudes toward the performance of bystander cardiopulmonary resuscitation 85

Table 3 Percentage of respondents willing to perform chest compression plus mouth-to-mouth ventilation/chestcompression alone

Scenarios

Stranger Trauma Child Elderly RelativeHigh school students 14.8/52.6* 18.1/50.9* 36.8*/63.6* 25.0/57.2* 41.1*/68.2*

Our previous study 13/73 18/66 50/80 23/70 53/85

High school teachers 28.5/75.2 30.4/70.8 51.7*/84.9 36.7/74.0 64.5/87.8Our previous study 25/76 27/65 41/85 31/77 64/90

EMTs 27.5*/100 22.8*/99.3 86.6/100 44.3/100 92.6/100Our previous study 67/97 68/96 85/94 42/86 96/99

Medical nurses 22.6*/88.9 19.5*/81.2 61.0*/92.2 35.7*/86.3 79.6*/96.5Our previous study 34/87 36/81 85/94 42/86 88/96

Medical students 51.2*/96.6 41.9*/93.9 87.2/98.9 77.7/97.8 92.7/99.4Our previous study 61/96 63/90 91/99 71/95 95/98

EMTs, emergency medical technicians* P < 0.05 vs. our previous study.

CC plus MMV in all scenarios than in our previousstudy.

Reasons for not performing CC plus MMV(Figure 1)

Among high school students and teachers, only9—10% who declined to perform CC plus MMV statedthat they would not do so because of concernabout disease transmission, more than 70% claimedthat they declined because of the fear of poorknowledge and/or imperfect performance of CCplus MMV. More than 70% of EMTs and medical stu-dents who declined to perform CC plus MMV refusedbecause of this concern. Moreover, 50% nurses whodeclined based their refusal on the fear of diseasetransmission. Among medical nurses and medicalstudents, the degree of concern about diseasetransmission was significantly higher than that inour previous study. In other categories, the reasonsfor not performing CC plus MMV were not signifi-cantly different between the present and previousstudy.

Discussion

Only 15—50% of laypeople and health care providerswere likely to perform CC plus MMV, especially ona stranger and a trauma victim with blood on face,but 50—100% were likely to perform CC only, whichwas the same as the results of our previous studyperformed in 1998. The reasons for unwillingnessamong laypeople to perform CC plus MMV weretheir inadequate knowledge and/or doubt that theywould be capable of performing CC plus MMV, while

the main reason among health care providers wasthe fear of contracting disease.

The present study demonstrated that Japanesehigh school students were reluctant to perform CCplus MMV on a stranger or trauma victim with bloodon face, and they were more reluctant to performCC only in all scenarios than in our previous study.Moreover, the present study showed that the reasonfor unwillingness to perform CC plus MMV amonglaypeople was their poor knowledge and/or fear ofinadequate performance. The present study sug-gested that one of the reasons of these findings maybe a lower % of high school students trained in CPRcourses than that of our previous study. Neverthe-less, the number of previous CPR training coursesin other categories was higher than in our previousstudy. Our previous study indicated that repeatedand effective CPR training could increase their

Figure 1 Reasons for not performing CPR. Open barsindicate the fear of poor knowledge and/or imperfectperformance of CPR; solid bars, concern about dis-ease transmission; teachers, high school teachers; EMTs,emergency medical technicians; nurses, medical nurses.

Effectivement…

Lieu Période RCP - RCP + V + MCE MCE

RA Waalewijn Resuscitation 2001

Pays-Bas (Amsterdam) 1995/1997

429/922

46,5% 493/922

53,5% 437/493

88,6% 41/493

8,3%

T Iwami Circulation 2007

Japon (Osaka) 1998/2003

3550/4877

72,8% 1327/4877

27,2% 783/1327

59% 544/1327

41%

K Bohm Circulation 2007 Suède 1990/2005 / 11275

8209/11275

73% 1145/11275

10%

SOS Kanto The Lancet 2007

Japon (Kanto) 2002/2003

2917/4068

71,7% 1151/4068

28,3% 712/1151

61,9% 439/1151

38,1%

TM Olasveengen Acta Anasthesiol Scand 2008

Norvège (Oslo) 2003/2006

269/695

39% 426/695

61% 287/426

66% 145/426

34%

MEH Ong Resuscitation 2008 Singapour 2001/2004

1695/2136

79,4% 441/2136

20,6% 287/441

65,1% 154/241

34,9%

T Ogawa BMJ 2011 Japon 2005/2007

56851/101781

55,7% 44930/101781

44,3% 19328/40035

48,3% 20707/40035

51,7%

Mais si on ne fait rien ce n’est peut-être pas pire ?

Critère de jugement RCP + RCP - Test statistique

RA Waalewijn Resuscitation 2001 Sortie vivant 14% 6% P < 0,001


Bonne évolution neurologique à J30 3,5% 2,1% /

SOS Kanto The Lancet 2007 Bonne évolution neurologique à J30 5% 2% 2,4 (1,6 – 3,4)


Sortie vivant 11,8% 9% /

MEH Ong Resuscitation 2008

Sortie vivant ou survie à J30 2,7% 0,5% /

Critère de jugement RCP - MCE seul Test statistique

RA Waalewijn Resuscitation 2001 Sortie vivant 6% 15%


Bonne évolution neurologique à J30 2,1% 3,5% 1,70 (1,02 – 2,84)

SOS Kanto The Lancet 2007

Bonne évolution neurologique à J30 2% 6% 3,0 (1,9 – 4,7)

TM Olasveengen Acta Anasthesiol Scand 2008 Sortie vivant 9% 10% /

MEH Ong Resuscitation 2008 Sortie vivant ou vivant à J30 0,5% 2,6% 5,0 (1,5 - 16,4)

Critère de jugement RCP - MCE + V Test statistique

RA Waalewijn Resuscitation 2001 Sortie vivant 6% 14% /

T Iwami Circulation 2007 Bonne évolution neurologique à J30 2,1% 3,6% 1,74 ( 1,12 – 2,71)

SOS Kanto The Lancet 2007 Bonne évolution neurologique à J30 2% 4% /


Sortie vivant 9% 13% /

MEH Ong Resuscitation 2008 Sortie vivant ou vivant à J30 0,5% 2,8% 5,4 (2,1 – 14,0)

Différence d’efficacité des deux techniques ?

Critère de jugement MCE + V MCE seul Test statistique

RA Waalewijn Resuscitation 2001 Sortie vivant 14% 15% p = 0,713

T Iwami Circulation 2007 Bonne évolution neurologique à J30 3,5% 3,6% /

K Bohm Circulation 2007 Vivant à J30 7,2% 6,7% 1,10 (0,86 – 1,40)

SOS Kanto The Lancet 2007 Bonne évolution neurologique à J30 4% 6% 1,5 (0,9 – 2,5)

TM Olasveengen Acta Anasthesiol Scand 2008 Sortie vivant 13% 10% p = 0,647

MEH Ong Resuscitation 2008 Sortie vivant ou vivant à J30 2,8% 2,6% 0,9 (0, 3 – 3,1)

T Ogawa BMJ 2011

Bonne évolution neurologique à J30 5,6% 4,6% 1,17 (1,02 – 1,35)

Une mauvaise interprétation….

« The outcome after CPR with chest compression alone is similar to that after chest compression with mouth-to-

mouth ventilation…. »

The New England

Journal

of

Medicine

© Copyr ight, 2000, by the Massachusett s Medical Society

VOLUME 342

M

AY

25, 2000

NUMBER 21

1546

·

May 25, 2000

CARDIOPULMONARY RESUSCITATION BY CHEST COMPRESSION ALONE OR WITH MOUTH-TO-MOUTH VENTILATION

A

LFRED

H

ALLSTROM

, P

H

.D., L

EONARD

C

OBB

, M.D., E

LISE

J

OHNSON

, B.A.,

AND

M

ICHAEL

C

OPASS

, M.D.

A

BSTRACT

Background

Despite extensive training of citizensof Seattle in cardiopulmonary resuscitation (CPR),bystanders do not perform CPR in almost half of wit-nessed cardiac arrests. Instructions in chest compres-sion plus mouth-to-mouth ventilation given by dis-patchers over the telephone can require 2.4 minutes.In experimental studies, chest compression alone isassociated with survival rates similar to those withchest compression plus mouth-to-mouth ventilation.We conducted a randomized study to compare CPRby chest compression alone with CPR by chest com-pression plus mouth-to-mouth ventilation.

Methods

The setting of the trial was an urban, fire-department–based, emergency-medical-care systemwith central dispatching. In a randomized manner,telephone dispatchers gave bystanders at the sceneof apparent cardiac arrest instructions in either chestcompression alone or chest compression plus mouth-to-mouth ventilation. The primary end point was sur-vival to hospital discharge.

Results

Data were analyzed for 241 patients ran-domly assigned to receive chest compression aloneand 279 assigned to chest compression plus mouth-to-mouth ventilation. Complete instructions weredelivered in 62 percent of episodes for the group re-ceiving chest compression plus mouth-to-mouthventilation and 81 percent of episodes for the groupreceiving chest compression alone (P=0.005). In-structions for compression required 1.4 minutes lessto complete than instructions for compression plusmouth-to-mouth ventilation. Survival to hospital dis-charge was better among patients assigned to chestcompression alone than among those assigned tochest compression plus mouth-to-mouth ventilation(14.6 percent vs. 10.4 percent), but the differencewas not statistically significant (P=0.18).

Conclusions

The outcome after CPR with chestcompression alone is similar to that after chest com-pression with mouth-to-mouth ventilation, and chestcompression alone may be the preferred approachfor bystanders inexperienced in CPR. (N Engl J Med2000;342:1546-53.)

©2000, Massachusetts Medical Society.

From the Department of Biostatistics (A.H., E.J.) and the Departmentof Medicine (L.C., M.C.), University of Washington, and Medic I, Seattle.Address reprint requests to Dr. Hallstrom at 1107 NE 45th St., Suite 505,Seattle, WA 98105-4689.

LTHOUGH bystander-initiated cardiopul-monary resuscitation (CPR) has been asso-ciated with an increase of 50 percent ormore in survival after out-of-hospital cardi-

ac arrest, and despite extensive training of citizens inCPR techniques,

1,2

approximately half of the victims ofwitnessed out-of-hospital cardiac arrests in the Seat-tle–King County, Washington, area during the pastfew decades did not receive bystander-initiated CPR.To address this problem, investigators in King Coun-ty initiated a program in which dispatchers weretaught to instruct callers in how to initiate CPR.

3,4

Theinstructions included airway management, mouth-to-mouth ventilation, and chest compression. The in-vestigators reported that dispatcher-instructed CPRby bystanders was associated with a rate of survivalto hospital discharge that was similar to the historicalexperience with bystander-initiated CPR, that thetime required to provide the instructions averaged 2.4minutes, and that the most common reason for notcompleting the instructions was the arrival of emer-gency-medical-services personnel.

Since the average interval to a response in Seattlewas 3.1 minutes, as compared with 4.5 minutes inthe suburban communities where the King Countystudy was conducted, it was unclear whether imple-menting such a program in Seattle might simply be adrain on dispatch-center resources. In addition, stud-ies in animals, particularly those by Meursing et al.,

5

demonstrated that central arterial oxygenation re-mains relatively high for a substantial time after theonset of ventricular fibrillation.

In 1989 we therefore began a preliminary trial ofdispatcher-instructed bystander CPR that comparedthe value of instructions for chest compression onlywith that of standard instructions for chest compres-sion plus mouth-to-mouth ventilation. After approx-

A

The New England Journal of Medicine Downloaded from nejm.org at INSERM DISC DOC on November 7, 2014. For personal use only. No other uses without permission.

Copyright © 2000 Massachusetts Medical Society. All rights reserved.

1550

·

May 25, 2000

The New England Journal of Medicine

excluded from the primary analysis because the pa-tients apparently had drug overdoses, alcohol intox-ication, or carbon monoxide poisoning. Among thesepatients, 80.7 percent (88 of 109) of those receivingchest compression plus mouth-to-mouth ventilationand 75.7 percent (78 of 103) of those receiving chestcompression alone survived to hospital discharge (P=0.39). The survival status of 15 patients was unknown.When included episodes and episodes that were ex-cluded because of apparent drug overdose, alcoholintoxication, or carbon monoxide poisoning are com-bined, survival to hospital discharge was 30.2 per-cent (117 of 387) in the group receiving chest com-pression plus mouth-to-mouth ventilation and 32.9percent (113 of 343) in the group receiving chestcompression alone, a nonsignificant difference of 2.7percentage points (95 percent confidence interval,¡4.1 to 9.5 percentage points).

The dispatchers mistakenly diagnosed a total of260 cases not due to cardiac arrest as being due tocardiac arrest and randomly assigned them to treat-ment groups. Another 10 cases that were not due tocardiac arrest were not diagnosed as due to cardiacarrest, but were still assigned to treatment for un-known reasons. Among these 270 episodes, instruc-tions were started in only 14 and completed in 7, pri-marily because the process of getting the patient tothe floor resulted in additional information that al-lowed the dispatcher to realize the mistake. Data fromthe dispatch tapes, paramedics’ reports, and telephoneinterviews showed no serious adverse effects on ei-ther the patient or the caller in any of these 270 ep-isodes, and in particular in any of the 14 episodes inwhich CPR was started after instruction from thedispatcher.

Only 20 dispatcher-instructed bystanders (14 giveninstructions for chest compression plus mouth-to-mouth ventilation, 4 given instructions for chest com-

*Because of rounding, not all percentages total 100.†The coexisting conditions were cancer, cardiac disease, and diabetes.

T

ABLE

3.

C

HARACTERISTICS

OF

P

ATIENTS

AND

E

PISODES

.*

C

HARACTERISTIC

C

HEST

C

OMPRESSIONPLUS

M

OUTH

-

TO

-M

OUTH

V

ENTILATION

(N=279)

C

HEST

C

OMPRESSION

A

LONE

(N=241)

Mean age (yr) 68.5 67.9Male sex (%) 64.9 62.2Race or ethnic group (%)

WhiteBlackNative AmericanAsianHispanicOther

81.08.10.08.91.60.4

72.216.10.99.10.90.9

Season (%)Winter (Nov.–Feb.)Spring (March–June)Summer or fall (July–Oct.)

33.735.530.8

32.839.028.2

Location (%)HomeOther residencePublic, indoorsPublic, outdoors

89.24.32.53.9

87.15.45.02.5

Episode unwitnessed (%) 43.2 41.4Mean first-unit response interval (min) 4.0 4.1First cardiac rhythm or state (%)

Ventricular fibrillationAsystolePulseless electrical activityVentricular tachycardiaUnknown

41.640.916.10.41.1

44.437.817.00.80.0

No. of coexisting conditions (%)†0123

25.849.121.13.9

27.049.020.73.3

Acute symptoms before episode (%)None«1 hr before episode>1 hr before episode

76.33.9

19.7

78.46.6

14.9CPR instructions completed (%) 61.6 80.5Possible adverse effects on patient (%) 1.8 3.7

*CI denotes confidence interval.

T

ABLE

4.

P

RIMARY

AND

S

ECONDARY

O

UTCOMES

A

CCORDING

TO

T

REATMENT

G

ROUP

.

O

UTCOME

C

HEST

C

OMPRESSIONPLUS

M

OUTH

-

TO

-M

OUTH

V

ENTILATION

C

HEST

C

OMPRESSION

A

LONE

T

WO

-S

IDED

P V

ALUE

D

IFFERENCE

(95% CI)*

no./total no. (%) %

Discharged alive (primary outcome)

29/278 (10.4) 35/240 (14.6) 0.18 4.2 (¡1.5 to 9.8)

Admitted to the hospital 95/279 (34.1) 97/241 (40.2) 0.15 6.1 (¡2.1 to 15.0)



The New England

Journal

of

Medicine


VOLUME 342

M

AY

25, 2000

NUMBER 21

1546

·

May 25, 2000


A

LFRED

H

ALLSTROM

, P

H

.D., L

EONARD

C

OBB

, M.D., E

LISE

J

OHNSON

, B.A.,

AND

M

ICHAEL

C

OPASS

, M.D.

A

BSTRACT

Background


Methods


Results


Conclusions






1,2


3,4



5



A



1548

·

May 25, 2000

The New England Journal of Medicine

Figure 1.

Protocol for Standard Instructions for CPR by Chest Compression Combined with Mouth-to-MouthVentilation.The instructions for CPR by chest compression alone do not include the shaded sections. In this exampleit is assumed that the victim is male.

!

I can tell you how to help until the!medics arrive. Do you want to help?

Help is on!the way.

Yes

No

No

Is there someone else!there who can help?

Tell that person!exactly what I say.

Can you get the phone near him?

Listen carefully. I’ll tell you what to do.

Get him flat on his back on the floor.!Strip his chest. Kneel by his side.!Pinch the nose. With the other hand,!

lift the chin so the head bends back.!Completely cover his mouth with yours.!Force 2 deep breaths of air into the lungs.!

Just as if you were blowing up a big balloon.!Remember: !

Flat on his back. Strip the chest.!Pinch the nose. With the other hand,!

lift the chin so the head bends back.!Force 2 breaths.!Then come back to the phone!

Put the heel of your hand on the center of the chest!right between the nipples.!

Put your other hand on top of that hand.!Push down firmly only on the heels of your hands,!

1 or 2 in. (2.5 or 5 cm). Do it 15 times.!Just as if you were pumping the chest.!

Make sure the heel of your hand is on the center of!the chest right between the nipples.!

Pump 15 times. Then pinch the nose and lift the chin!so the head bends back. Two more breaths!and pump the chest 15 times. Keep doing it!!

Pump the chest 15 times. Then 2 breaths.!Keep pumping on the chest until help can take over!!I’ll be hanging up now. Help is on the way.

Is he awake or breathing normally?

Listen carefully! I’ll tell you what to do next.

YesStop

Yes YesNo

Yes

No



Ø  Les instructions complètes de RCP sont délivrées chez 62% des personnes dans le groupe avec ventilation et 81% dans le groupe MCE seul (p<o,oo5)

Ø  L’arrivée du service médical d’urgence est la première raison expliquant cela : 20,8% groupe ventilation, 7,9% dans le groupe MCE seul

Ø  Donc, une partie non négligeable (mais non connue) des patients du groupe ventilation n’ont pas (ou peu) bénéficier de MCE et/ou de ventilation

Ø  Le pronostic de diffère pas, que l’on fasse ou pas un MCE le plus rapidement possible

The New England

Journal

of

Medicine


VOLUME 342

M

AY

25, 2000

NUMBER 21

1546

·

May 25, 2000


A

LFRED

H

ALLSTROM

, P

H

.D., L

EONARD

C

OBB

, M.D., E

LISE

J

OHNSON

, B.A.,

AND

M

ICHAEL

C

OPASS

, M.D.

A

BSTRACT

Background


Methods


Results


Conclusions






1,2


3,4



5



A



Compression-only CPR or standard in out-of-hospital cardiac arrest L Svensson nejm, 2010

Compression-Only CPR or Standard CPR

n engl j med 363;5 nejm.org july 29, 2010 439

receiving standard CPR. There were no signifi-cant differences between the two groups with respect to the other secondary end points.

Subgroup AnalysesThe rates of the primary outcome of 30-day sur-vival and the secondary outcome of 1-day survival did not differ significantly among the subgroups studied (Fig. 2A and 2B). Specifically, the rate of the primary end point did not vary significantly with age (P = 0.50), interval between call and first EMS response (P = 0.95), or first cardiac rhythm (P = 0.99). Adjustment for the baseline character-istics did not change the results.

There was no significant difference in the rates of survival between the two groups after data from patients under 18 years of age were exclud-ed. Nor did the rates of survival differ signifi-cantly between the groups for patients who re-ceived treatment other than the treatment they had been randomly assigned to receive. Details of these subgroup comparisons, with respect to the pri-mary and secondary end points, and compari-sons of patients whose cardiac arrest was classi-fied as uncertain and those with “true” cardiac arrest are provided in the Supplementary Appen-dix, available at NEJM.org.

Loss to Follow-upInformation on follow-up was unavailable for 132 of 1952 patients (6.8%), the main reason being loss of the corresponding EMS field reports, oc-

curring primarily in a small number of EMS dis-tricts. We therefore performed a subgroup analy-sis excluding districts where more than 18% of patients were lost to follow-up. No difference from the main results was found.

Discussion

Our nationwide, randomized study of witnessed out-of-hospital cardiac arrest shows that giving instructions for compression-only CPR before the arrival of EMS personnel does not significantly improve the outcome of patients as compared with standard CPR. Neither the 1-day nor 30-day rates of survival differed significantly between the two groups. Furthermore, there was no signifi-cant difference in the rates of survival among various subgroups. The findings were similar ir-respective of whether the data were analyzed ac-cording to the assigned treatment (the primary analysis) or the treatment received. Our results are in agreement with those from previously pub-lished retrospective registry studies.4,5,8

Previous studies in animals have shown no dif-ferences in survival or neurologic outcomes with standard CPR and compression-only CPR.3,9 One investigation even showed adverse outcomes re-lated to the interruption of chest compression in order to perform mouth-to-mouth ventilation.10 Complete occlusion of the airways does not re-duce the chances of survival if reasonable circu-lation is provided by chest compression.11

Table 3. Survival Outcomes in the Study Population, According to Treatment Group.*

OutcomeCompression-

Only CPRStandard

CPRTwo-Sided

P ValueDifference (95% CI)

no. of patients/total no. (%) percentage points

Primary analysis

30-Day survival 54/620 (8.7) 46/656 (7.0) 0.26 1.7 (−1.2 to 4.6)

1-Day survival 147/613 (24.0) 136/652 (20.9) 0.18 3.1 (−1.5 to 7.7)

Survival to discharge from hospital 54/282 (19.1) 44/297 (14.8) 0.16 4.3 (−1.8 to 10.5)

Per-protocol analysis

30-Day survival 39/461 (8.5) 43/575 (7.5) 0.56 1.0 (−2.3 to 4.3)

1-Day survival 115/457 (25.2) 123/571 (21.5) 0.17 3.6 (−1.6 to 8.8)

Survival to discharge from hospital 39/220 (17.7) 42/261 (16.1) 0.63 1.6 (−5.1 to 8.4)

* Data from 1276 patients were included in the primary analysis, and data from 1036 were included in the per-protocol analysis. Data for survival to discharge were missing for many patients who died before day 30. CI denotes confidence interval, and CPR cardiopulmonary resuscitation.



5 x 100 = 500 10 x 100 = 1000 Si on est seul…..

Quality of chest compressions during continuous CPR; comparison between chest compression-only CPR and conventional CPR C Nishiyama Resuscitation, 2010

1154 C. Nishiyama et al. / Resuscitation 81 (2010) 1152–1155

Fig. 2. CPR quality index, the proportion of chest compressions with appropri-ate depth among the total chest compressions during 20-s CPR period, for chestcompression-only CPR and conventional CPR. Error bar indicates standard deviation.

tional CPR group over time. The decay of CPR was greater duringthe chest compression-only CPR, and the intergroup difference inthe CPR quality index increased, reaching statistical significance at61–80 s period (p = 0.003).

3.3. Interruption of chest compressions during 2-min CPR

Time to the first resuscitation (either chest compressionor ventilation) was 32.0 ± 8.0 s and 35.0 ± 9.0 s in the chestcompression-only CPR group and in the conventional CPRgroup, respectively (p = 0.005). No-flow time was 32.0 ± 9.0 s and81.0 ± 10.0 s in the chest compression-only CPR group and the con-ventional CPR group, respectively (p < 0.001).

4. Discussion

To our knowledge, this is the first RCT indicating a difference inthe time-dependent deterioration of chest compressions betweenchest compression-only CPR and conventional CPR with compres-sion/ventilation ratio of 30:2. It demonstrated that the quality ofchest compressions declined more rapidly in chest compression-only CPR than in conventional CPR and the difference reachedstatistical significance 1 min after start of CPR. Previous studiesevaluating the influence of rescuer fatigue on the quality of chestcompressions had some weak points including the non-controlledstudy design, biased participants such as medical students,12 andno CPR training before skills evaluation.13 In the latter study, itis not clear whether their fatigue or lack of CPR knowledge skillshindered their CPR performance, because their CPR were evalu-ated without prior CPR training. In addition, they compared chestcompression-only CPR with the conventional CPR with compres-sion/ventilation ratio of 15:2.12 Hence, our study provided validand updated evidence on the time-dependent CPR deterioration.

Although the number of total chest compressions did not changeover time, the percentage of correct compressions with sufficientdepth was gradually reduced with ongoing resuscitation as pre-viously reported.12,15,16 Most plausible explanation of this qualitydeterioration is the cumulative fatigue resulting from continuouschest compressions.1,10,11,15 One clinical study evaluating actualin-hospital resuscitation also found time-dependent deteriorationof chest compression by rescuer’s fatigue.17 This inference wasalready confirmed by some studies using physiological markerssuch as heart rate, oxygen saturation by pulse oximetry, blood

lactate concentration, and neuromuscular function.18–21 In thesereports, physical fitness may be beneficial to CPR providers toensure the adequacy of chest compressions during CPR.10,16,18–21 Inthe conventional CPR, longer hands-off time caused by ventilationsmight serve as a rest and result in a recovery from fatigue.

The 2005 American Heart Association guidelines for CPR andemergency cardiovascular care recommend that rescuers changetheir chest compressions every 2 min during conventional CPR,1

but it does not refer to a change during chest compression-only CPR due to insufficient evidence. Rescuers generally do notbecome aware when their fatigue-induced compression impair-ment begins.11 Unless the rescuer complains of fatigue, he/sheshould be replaced with others in order to maintain the qualityof chest compressions. According to the findings of this study, werecommend that rescuers should change their roles in CPR every1 min for chest compression-only CPR.

Even if chest compression-only CPR has the drawback of easyfatigability, much larger numbers of chest compressions during 2-min CPR than the conventional CPR8,9 would amply compensatefor this weakness. It is simpler, easier to learn and perform8,9 andis expected to increase the number of lay people who could per-form bystander-initiated CPR.9,22 Bearing these features in mind,we should disseminate chest compression-only CPR in order toincrease bystander CPR and improve survival after OHCA.

This study has some limitations. First, we did not measure theintermediate factors that lowered the quality of CPR and could notthoroughly infer the biological mechanism of chest compressiondecay. Second, the resuscitation skills were evaluated by a case-based scenario test and actual resuscitation performances wereunknown. Third, although rescuers’ characteristics such as gen-der, height, and weight also would influence the quality of chestcompressions,10 we did not evaluate these factors in this study. Weare planning to assess the associations between rescuers’ charac-teristics and CPR quality in the next study. Finally, we cannot ruleout the possibility that the difference in duration of CPR trainingmight affect the outcomes.”

5. Conclusions

Quality of chest compressions rapidly declined in the chestcompression-only CPR compared with the conventional CPR. Werecommend that rescuers should change their roles in CPR every1 min to maintain the quality of chest compressions during chestcompression-only CPR.

Conflict of interest statement

There are no conflicts of interest to declare.

Role of funding source

This study was supported by a Grant-in-Aid for Health andLabour Sciences Research Grants (H16-Shinkin-02) from theJapanese Ministry of Health, Labour and Welfare.

Acknowledgements

We gratefully acknowledge Masaaki Matsumoto, KatsuharuHirai, Shohei Nakai, Keiji Akatsuka, Tatsuo Azuma, Katsuo Ogura,Eiji Ohtani, Nobuyuki Iwai, Masato Ando, Kazushi Nakajima,Yasuyuki Shinkai, Katsuya Ito, Seiji Kasatani, and Rei Suzuki forinstruction in the CPR training program. We also thank all membersof the Japanese Population-based Utstein-style Study with Basicand Advanced Life Support Education (J-PULSE), and the faculty ofKyoto University School of Public Health for helpful comments on

Ø  Etude randomisée, volontaires âgés de plus de 18 ans

Ø  2 groupes de 104 et 105 personnes, instructions délivrées sur 2h (CPR-only) et 3h (conventional CPR)

Ø  Evaluation de la qualité du MCE (mannequin) pendant 2 min sur des périodes de 20 sec

Ø  Diminution de la qualité du MCE dans le temps, plus marquée dans le groupe « CPR-only »

Ø  On note néanmoins un nombre moins important de compressions dans le groupe « conventional CPR »

Le MCE assure une ventilation satisfaisante ?

Does compression-only cardiopulmonary resuscitation generate adequate passive ventilation during cardiac arrest ? CD Deakin Resuscitation, 2007

56 C.D. Deakin et al.

Table 1 Demographics and time intervals from collapse to ambulance arrival and arrival in hospital (VF: ventricularfibrillation; PEA: pulseless electrical activity; LMA: laryngeal mask airway)

Age(years)

Weight(estimated)(kg)

Initialrhythm

Intubatedpre-hospital?

Bystander CPRadministered

Time from collapse toambulance arrival (min)

Time from collapse tohospital arrival (min)

55 75 VF LMA Yes 7 4745 60 Asystole Yes No 17 4382 80 Asystole Yes No 4 4071 55 PEA Yes Yes 12 4882 80 Asystole Yes No 11 4064 85 VF Yes Yes 6 5464 60 Asystole Yes Yes 7 3774 70 PEA No No 10 3466 75 Asystole No Yes 0 471 80 VF Yes Yes 5 2871 120 PEA LMA Yes 7 4876 60 Asystole Yes Yes 6 3259 65 PEA Yes Yes 9 4347 90 PEA No No 9 2964 80 Asystole No Yes 8 4459 65 VF Yes Yes 5 4680 90 PEA LMA No 10 39

Figure 1 Typical respiratory variables recorded during resuscitation, showing cyclical manual ventilation (A) andinterspersed passive ventilation (B) from chest compressions delivered by the LUCAS thumper. This example shows tidalvolumes of approximately 700 ml from manual ventilation delivered using a self-inflating bag, followed by volumes ofapproximately 60 ml from passive ventilation. Corresponding end-tidal CO2 measurements are also shown.

Ø  Patients ayant un ACR en dehors de l’hôpital, pris en charge aux urgences, intubés et ventilés

Ø  MCE par le systéme LUCAS en annulant la décompression active, rythme de 100/min

Ø  Mesure des volumes expirés et de l’EtCO2

Ø  Analyse de la ventilation par MCE seul durant la pose de la voie veineuse centrale (environ une minute)

Ø  17 patients inclus

Ø  Vt médian : 41,5 ml(33,0 - 62,1ml)

Ø  Volume minute CO2 médian : 19,5ml (15,9 – 33,8 ml; normales 150 – 180 ml)

Adequate passive ventilation during cardiac arrest 57

Figure 2 Tidal volume (Vt):deadspace (Vd) ratio duringpassive ventilation generated by LUCAS thumper (n = 16).The boundaries of the box indicate the 25th and 75thpercentile, and the line within the box marks the median.Whiskers above and below the box indicate the 90th and10th percentiles, respectively. Outlying points are shownas full circles.

CO2 minute volume

The median minute volume CO2 was 19.5 ml (range15.9—33.8 ml; normal range 150—180 ml).

End-tidal CO2

The median end-tidal CO2 recorded during passiveventilation was 0.93 kPa (range 0.0—4.6 kPa).

Deadspace

Median anatomical deadspace (Vd) was 162.7 ml(range 119.7—206.6 ml). In all study patients, pas-sive tidal volume (Vt) was less than the measuredanatomical deadspace. Median tidal volume as apercentage of deadspace was 27% (range 0—75%).Results are shown graphically in Figure 2.

Discussion

Passive ventilation occurring as a result ofcompression-only CPR in humans appears to beineffective in generating tidal volumes adequatefor gas exchange. In all patients, passive tidalvolumes were significantly less than the patient’sestimated deadspace. Additionally, CO2 minute vol-ume, an estimate of actual gas exchange, wasapproximately 10% of the normal range. We areunaware of any other study that has measured

end-tidal CO2 during measurement of passive venti-lation. EtCO2 is a measure of alveolar gas exchange,its production requiring both alveolar ventila-tion and pulmonary capillary blood flow. Duringcompression-only CPR, we found sustained levelsof EtCO2 in most patients, suggesting that alveo-lar gas exchange was occurring, despite measuredpassive tidal volumes being consistently less thanthe estimated anatomical deadspace. This sug-gests that gas transport mechanisms similar tothose described in high frequency ventilation maybe occurring due to the relatively high respira-tory frequency. These mechanisms include directbulk flow, longitudinal dispersion, Pendeluft venti-lation, asymmetric velocity profiles, and moleculardiffusion.25 Mechanical agitation from externalchest compression may also affect mixing of res-piratory gases. Although this study suggests thathigh frequency mechanisms may result in some gasexchange during compression-only CPR, the CO2minute volumes documented in this study suggestthat it is inadequate to sustain alveolar oxygendelivery at a level able to maintain adequate longterm tissue oxygen delivery.

Although the primary aim of ventilation is todeliver oxygen, the well established methodologyused for this study is based on measurements ofexhaled CO2 which in addition to enabling mea-surement of tidal volume, allow indirect calculationof deadspace. The use of this CO2-based techniquetherefore provides additional data that gives a bet-ter understanding of respiratory physiology thancould be obtained by other techniques.

The study was undertaken using a mechanicaldevice to deliver optimal external chest compres-sion in order to avoid bias from poor quality manualchest compression. Airway patentcy was ensured bytracheal intubation. In the pre-hospital setting ofbystander CPR, poorer quality chest compression26

and partial airway obstruction27 are common andboth likely to result in even smaller passive tidalvolumes than those recorded in this study.

Following the introduction of external chestcompression for cardiac arrest, Safar studied 30anaesthetised healthy patients subject to phar-macological neuromuscular blockade.28 In patientswith an open airway, ‘‘. . .rhythmic firm pressureover the lower half of the sternum at a rateof 1 compression per second. . .’’ generated anaverage tidal volume of 156 ml with a range of0—390 ml. The tidal volumes were larger than theestimated dead-space volume in 57% but smallerin 43% patients. The same paper documented 12further patients in cardiac arrest in whom ‘‘. . .acuffed tracheal tube was inserted and closed chestmassage performed with maximal force.’’ No ven-

Interruption trop longue du MCE avec la ventilation ?

Single rescuer cardiopulmonary resuscitation : Can anyone perform to the guidelines 2000 recommandations ? TA Higdon Resuscitaion, 2006

Ø  Etude sur mannequin

Ø  24 pompiers ayant reçu une formation à la RCPB dans les 2 dernières années

Ø  Nouvelle technique de RCP : compressions thoraciques continues

Ø  Pratique des 2 types de RCP avec enregistrement des paramètres par le mannequin et par un enregistrement vidéo

Ø  Première technique testée randomisée

36 T.A. Higdon et al.

Table 1 Data acquisition

Resusci Anne compression data Resusci Anne ventilation data Video dataAverage depth of chest

compressionAverage ventilation volume Did they assess victim responsiveness?

Average number of compressionsper minute

Average number ofventilations per minute

Did they call for help/911?

Average compression rate Minute volume Time to first compressionTotal number of delivered

compressionsTotal number of deliveredventilations

Verify unsuccessful ventilations attemptsnot registered by manikin

Note that ‘‘compressions per minute’’ refers to the number of compressions delivered in a 1-min period and ‘‘compression rate’’refers to the rate of compression during an uninterrupted stretch of compressions.

ing capabilities of these manikins were enabled toinclude quantitative and qualitative data on bothchest compressions and ventilations. The videorecordings were used to evaluate variables notrecorded by the manikins (Table 1).

This study was approved by the University ofArizona Institutional Review Board (IRB). Signedinformed consent was obtained from each partic-ipant.

Statistical analysis of the data was performedusing commercially available Statview 5.0 statis-tical software (SAS Institute, Cary, NC). Paired t-tests and/or Chi square were used to compare per-formance on STD-CPR versus UCC-CPR. Data arereported as mean ± S.E.M.

Results

All 24 paramedics completed both the STD-CPR andthe UCC-CPR branches of the study successfully. Ofthese 24 paramedics, three (13%) were female. Themean age was 36 ± 1 years and ranged from 26 to48 years. Thirteen (54%) were certified CPR instruc-tors and all 24 (100%) had taken at least one CPRcertification class within 2 years. Twenty of the 24(83%) had performed CPR in an emergency situationat least once.

During the performance of standard single res-cuer CPR, the average pause for rescue breathingwas 10 ± 1 s with a range of 7—19 s. The mean num-ber of ventilations delivered per minute during STD-CPR was 6 ± 0.4 breaths/min and the mean minute

ventilation was 4.7 ± 0.5 L/min. Table 2 showsthe comparison between UCC-CPR and STD-CPR.The compression rate was not different betweenthe two types of CPR (99 ± 5 versus 92 ± 4). Dueto the lengthy interruptions of chest compres-sions to deliver the two breaths during STD-CPR,paramedics delivered significantly fewer compres-sions while performing STD-CPR than when per-forming UCC-CPR (44 ± 2 versus 88 ± 5; p < 0.0001).

Similar to previous studies2,6 we found thatthe initial delay prior to providing chest compres-sions was much longer with STD-CPR than UCC-CPR(27 ± 1.2 versus 9 ± 0.8 s, p < 0.0001). This was duein part to the time spent assessing and positioningthe manikin and providing two initial rescue breathsprior to starting chest compressions according tothe AHA BLS protocol for the single rescuer.

When performing UCC-CPR, 15/22 paramedics(68%) delivered the minimal goal of at least 80 chestcompressions per minute whereas none achieved 80compressions per minute when performing STD-CPR(!2 = 22.76, p < 0.001)

A questionnaire completed anonymously byall participating paramedics following testingdescribed the attitudes of this population regardingCPR and mouth-to-mouth ventilation. When askedif they would, if off duty, perform standard CPRon a stranger who collapsed in a public place, 2of 24 (8%) answered that they would definitely doso. However, given the same scenario, 22/24 (92%)responded that they would definitely be willing toperform uninterrupted chest compressions. All butone respondent indicated that they thought the

Table 2 Comparison of CPR techniques

STD-CPR (15:2) CC-CPR p-valueTime to 1st compression 27 ± 1.2 s 9 ± 0.8 s <0.0001Pause in compressions for rescue breaths 10 ± 1 s NA NACompression rate 99 ± 5 92 ± 4 0.23Compressions delivered/min 44 ± 2 88 ± 5 <0.0001

Duration of ventilations during cardiopulmonary resuscitation by lay rescuers and first reponders : relationship between delivering chest compressions and outcomes SG Beesems Circulation, 2013

Ø  Prospective, observationnelle

Ø  Patient ayant un ACR extrahospitalier, bénéficiant d’une RCP par un témoin un pompier ou un policier, et équipé d’un défibrillateur automatique (DA)

Ø  Analyse de l’enregistrement du DA afin de déterminer le nombre de compressions thoraciques et la durée des périodes de ventilation

Ø  Sont exclus les tracés non analysables, une période analysable < 2min, les RCP avec uniquement un MCE, LES RCP avec un rapport compressions ventilation de 15/2

Ø  199 inclusions

Ø  Durée interruption pour 2 insufflations : 7 sec (6 - 9)

1586 Circulation April 16, 2013

responders) equipped with an AED (LIFEPAK 500 or LIFEPAK 1000, Physio Control Inc, Redmond, WA). In the Netherlands, firefighters and policemen are dispatched as part of the organized response to cardiac arrest but are considered lay rescuers because their training includes only the standard ERC basic life support/AED courses for lay rescuers.

Besides the dispatched first responder, bystanders are also able to perform basic life support and to defibrillate with an onsite AED be-fore the arrival of an ambulance. In the past, such lay rescuers had generally received a standard ERC basic life support and AED course. All rescuers were trained to perform CPR according to the 2005 guidelines, which included a compression/ventilation ratio of 30:2.

Data SourceThe Amsterdam Resuscitation Study (ARREST) is an ongoing, pro-spective registry of all out-of-hospital cardiac arrests in the Dutch prov-ince of North Holland. All data are collected according to the Utstein recommendations. The Medical Ethics Review Board of the Academic Medical Center in Amsterdam approved the study and gave a waiver for the requirement of (written) informed consent. Details of the design of the data collection in the ARREST study are described elsewhere.9

Study Design and Data CollectionThe investigation was a prospective study of all persons who suffered out-of-hospital cardiac arrest, an AED was attached, and received CPR by trained lay rescuers in the period of September 2010 until March 2011 in the Dutch province North Holland.

Medical students collected all AED ECG recordings shortly after a cardiac arrest. These data were stored and analyzed with dedicated software specific for each type of AED.

For the purpose of this study, we included only AEDs for which the impedance recording (Physio Control LP500, LP1000, or LPCR+) or the displacement transducer (Zoll AED Plus, ZOLL Inc., Chelmsford, MA) allowed accurate determination of chest compressions.

Recordings were eligible for analysis if the AED had recorded at least the first complete compression-ventilation cycle from the notification of “start CPR” to “stop CPR” by the voice prompt of the AED before the AED was disconnected by emergency medical service personnel. We excluded ECGs with a compression/ventilation ratio other than 30:2 and ECGs that were not analyzable because of technical deficiencies.

We differentiated the dispatched first responders from the other lay rescuers by the AED used. LP1000 and LP500 were used solely by dispatched first responders; LPCR+ and the Zoll AED Plus were used solely by nondispatched onsite rescuers.

Data AnalysisAll recordings were annotated for initiation and termination of a compression period. For our analysis, we selected the first and, when available, the last complete cycle of CPR of an AED recording to

ensure that the possible effect of rescuer fatigue on ventilation dura-tion and compressions was not overlooked in our analysis. The be-ginning of a cycle was defined as the moment the AED instructs the rescuer to start the resuscitation effort (marked “CPR prompt” in the recording). If the first identifiable compression was given before the notification of CPR prompt, that first compression was marked as the start of the compression period. The 2-minute interval ended with the last compression given after the AED instructed the rescuer to stop CPR (“stop CPR” in the recording; Figure 1).

The AED connect period consisted of 1 full cycle in 85 cases, 2 full cycles in 66 cases, and ≥3 cycles in 48 cases. Not all cycles in those with >1 cycle were analyzable; therefore, in 12 of those cases, only 1 cycle was included in the analysis.

A ventilation period was defined as a regular interruption in chest compressions during the CPR cycle for >3 seconds when impedance changes suggested ventilations and/or when the AED had given ven-tilation prompts. An interruption of chest compressions of <3 seconds was not considered a true attempt to ventilate.

The chest compression fraction was the proportion of the total re-suscitation time without spontaneous circulation during which chest compressions were administered, averaged over the cycles analyzed in our study. We analyzed the duration of each ventilation and chest compression period, as well as the number of chest compressions and ventilations delivered during each 2-minute CPR cycle. We calculated the average duration of the ventilation period by adding the duration of all ventilation cycles in the first cycle and (when available) the last cycle and divided by the number of ventilation periods.

Follow-upSurvival to discharge was verified by contacting the hospital to which the patient had been transported. We retrieved data on neurological outcome at discharge from the hospital charts and estimated the ce-rebral performance category: 1=good cerebral performance. 2=mod-erate cerebral disability, 3=severe cerebral disability, 4=coma or vegetative state, and 5=death.

Statistical AnalysesStatistical analyses were performed with standard software (SPSS ver-sion 18.0 for Mac, SPSS Inc, Chicago, IL). Time intervals and other median values were expressed as medians (25th–75th percentiles). Baseline comparisons were analyzed by calculating the χ2 statistic or 1-way ANOVA. The paired t test was used to determine statistical sig-nificance between the number of compressions between periods 1 and 2. The number of ventilations delivered by dispatched first responders and onsite rescuers was analyzed with the Mann-Whitney U test.

We examined the association between ventilation pause and sur-vival. We measured the distribution of relevant baseline factors possibly associated with survival. These factors were age, sex, wit-nessed collapse, time interval from emergency call to attachment of

Figure 1. Schematic time frame of 1 cycle of an electronic recording from an automatic external defibrillator (AED) showing the ECG (black line) and the impedance channel (green line) that reflects chest compressions. The 2 slower and shallower deflections during the ventilation pauses reflect the impedance change caused by 2 insufflations. The AED voice prompt “start CPR” (cardiopulmonary resuscitation) was marked as period 1 start (P1s). The first identifiable compression after the moment the compressions were started was marked C1, even if it occurred before P1s. Likewise, the beginning of a period of ventilation was marked V1. We finished a period with the last compression after the AED prompted “stop CPR” (P1e), and we marked the last compression of that cycle as V3, even if this occurred after the voice prompt to stop CPR.

at INSERM - DISC on November 9, 2014http://circ.ahajournals.org/Downloaded from



including time from emergency call to attachment of the AED, VF as initial rhythm, and type of lay rescuer, were unevenly distributed between the ventilation groups. After adjustment for baseline factors, ventilation pause duration was not associated with significantly better or worse survival (Table 4). If analyzed for VF cases only, the results of the multivariable analysis were similar (data not shown). Post hoc secondary analyses using alternative grouping of ventilation duration (3–5, 6–9, 10) or only ventilation group variables of 3 to 12 and ≥13 seconds (data not shown) did not demonstrate quantitatively different results.

DiscussionThe results of our study show that lay rescuers require a median ventilation time of 7 seconds (25th–75th percentile, 6–9 sec-onds) to complete 2 ventilation attempts. Only 21% are able to fully meet the ventilation guidelines of 2010. However, 97% of all rescuers provided chest compressions above the recom-mended minimum of 60 chest compressions in 1 minute, 88% of all administered >70 chest compressions in 1 minute, and 63% administered >80 chest compressions in 1 minute.

Interruptions in chest compressions for rhythm analysis, rescue shocks, ventilations, or human error cause a rapid decrease in the aortic relaxation (diastolic) pressure and thereby cause a decrease in coronary and cerebral perfusion

pressures.10,11 Earlier investigations demonstrated an associa-tion between the proportion of resuscitation time that chest compressions are administered and survival to hospital dis-charge after out-of-hospital cardiac arrest.10 Therefore, the AHA and ERC 2010 guidelines emphasized the importance of minimizing the time without chest compressions.1,2

The long interruption time for ventilations is cited as a jus-tification for compression-only CPR to allow a sufficient num-ber of chest compressions.6,7 We found that the time to provide the 2 ventilation breaths was even shorter in real life than it was in the studies with manikins4,5 and that the great majority of all rescuers provided chest compressions above the recommended minimum of 60 chest compressions in 1 minute. Raising the chest compression rate to at least 100 per minute in the 2010 guidelines could make the time for a ventilation pause even less critical. However, the intention was to not exceed a chest compression rate of >120 per minute. Nevertheless, in 23% of the cases, the chest compression rate was >120 per minute with potentially a small adverse influence on survival.12

Predictors of SurvivalIn our study, better survival was observed with longer ventila-tion times and the lowest chest compression fractions. This find-ing is not in agreement with the findings of other studies with

Table 2. Ratio of Compressions and Ventilations Delivered

Ventilation Duration, s

3–5 6–7 8–9 10–12 ≥13 P Value*

Cases, n (%) 42 (21) 58 (29) 50 (25) 28 (14) 21 (11)

Chest compression rate/min, median† 107 (101–121) 105 (102–118) 113 (103–126) 111 (101–118) 106 (96–116) 0.18

Chest compression rate >100/min, % 81 80 88 82 72 0.73

Chest compression rate >120/min, % 26 19 34 14 14 0.39

Compressions/ventilations delivered, n/min‡ 95/3 84/3 84/3 84/3 70/2

≥60 chest compressions delivered/min, % 98 98 100 97 86 0.042

≥70 chest compressions delivered/min, % 95 93 96 89 43 <0.001

≥80 chest compressions delivered/min, % 93 66 72 54 19 <0.001

Chest compression fraction, median, %† 74 (68–79) 66 (61–70) 62 (57–66) 63 (54–74) 57 (49–63) <0.001

Survival, % (n/N) 12 (5/42) 22 (13/58) 26 (13/50) 29 (8/28) 43 (9/21) 0.007

*P value for trend.†Chest compression fraction is presented as median (25th–75th percentile).‡Numbers indicate the amount of compressions and single ventilations delivered in each minute.

Table 3. Distribution of the Baseline Factors in the Ventilation Groups

Ventilation Duration, s

3–5 6–7 8–9 10–12 >13 P Value*

Cases, n (%) 42 (21) 58 (29) 50 (25) 28 (14) 21 (11)

Patient age, mean±SD, y 66±19 66±15 65±14 65±15 65±15 0.99

Patient sex, male, n (%) 26 (62) 43 (74) 31 (62) 22 (79) 16 (76) 0.27

Witnessed collapse, n (%) 30 (71) 45 (78) 33 (66) 20 (71) 18 (86) 0.63

Dispatched first responder/onsite rescuers, n 38/4 42/16 37/13 16/12 11/10 <0.001

VF as initial rhythm, n (%) 15 (36) 26 (45) 32 (64) 15 (54) 12 (58) 0.030

Time from emergency call to AED attachment, min† 7:03 (6:16–9:01) 6:48 (5:16–9:29) 6:46 (4:23–9:37) 5:21 (1:44–8:38) 4:30 (1:03-7:18) 0.029

AED indicates automatic external defibrillator; and VF, ventricular fibrillation.*P value for trend.†Time interval is presented in median in minutes (25th to 75th percentile).



Beesems et al Interruptions of Chest Compressions 1589

long interruptions of chest compressions, mainly caused by pauses associated with defibrillation shocks.13,14 This paradox can be attributed to the fact that other baseline factors that are more important for predicting survival were unevenly distrib-uted between the groups of ventilation duration. After adjust-ment for the baseline factors, the ventilation pause duration was not associated with significantly better or worse survival (Table 4). It is also possible that the suggested detrimental effect of perishock pauses does not apply to pauses for ventilation.

It must also be noted that in the study that found a signifi-cant relationship between lower chest compression fractions and survival, the relation was seen mainly in the patients with chest compression fractions <60% and as low as 20%.10 In our study, the lowest chest compression fractions were never below 60%; the great majority of rescuers even could deliver >60 chest compressions per minute.

Are All Lay Rescuers the Same?None of the lay rescuers in our study were part of emergency medical services, but police and firefighters may have had a more strict retraining schedule than other lay rescuers and may have been involved in >1 case of a true cardiac arrest. This matches our earlier observations that AEDs from dispatched first responders were used on average twice a year, whereas onsite AEDs are used on average once in 30 years.9 Although we indeed observed that onsite rescuers were more frequently represented in the longest ventilation groups, this finding did not reach sta-tistical significance, and the difference in ventilation pauses was neither clinically or statistically significant, nor was the number of delivered chest compressions per minute. There is therefore no suggestion that onsite lay rescuers with no or minimal past experience did worse than the dispatched first responders.

What Is the Scientific Basis for the Current Recommendations in the 2010 Guidelines?No direct scientific evidence supports or refutes the recommended maximal insufflation time of 5 seconds. There is evidence15,16 that supports the recommendation that the insufflation volume should not exceed 600 mL and that short insufflation times increase the risk of gastric dilatation.17,18 The recommended 5-second interruption time for 2 ventilations is the mathematical consequence of the intention to deliver at least 60 chest compressions per minute at a rate of 100 per minute, given a compression/ventilation ratio of 30:2. Our study shows that

a chest compression fraction of >60%, compatible with good survival, is achieved in all ventilation groups except the longest.10 The importance of the minimal number of compressions delivered per minute is emphasized in a recent study in which the group of patients who received 75 to 100 compressions per minute had significantly more return of spontaneous circulation than those with fewer compressions delivered.12 Patients in all ventilation groups except the group with pauses ≥13 seconds received ≥70 compressions in 1 minute in the great majority of cases. Therefore, it appears that interruptions for ventilations are less critical for delivering sufficient compressions per minute than suggested by the current guidelines.

Longer insufflation periods proved compatible with excel-lent survival rates. There is no justification to recommend these very strict and short insufflation times, and it is possible to allow the recommended maximum time for ventilations to be at least 10 seconds without detriment to survival.

LimitationsIn most recordings, the impedance signal allowed identifica-tion of ventilation breaths, but in some cases, we could observe only the absence of compressions. In that situation, we are not certain that the pause in compressions, which we have defined as a ventilation period, was actually used to administer effec-tive ventilations.

Information about the quality of the delivered chest com-pressions was absent in the recordings except in the Zoll AED recordings. This is an important factor because our results show that even with a relatively long interruption for ventila-tion, the number of administered chest compressions per min-ute remains acceptable.

Likewise, no information was available about the quality of the ventilations provided. The study of Odegaard et al19 demonstrated that there is an association between duration and quality of ven-tilation breaths. Only half of the ventilation attempts by lay res-cuers on a manikin were successful. This could mean that rapid insufflations by rescuers who achieved very short interruptions of chest compressions do not have a positive effect on survival.

In general, lay rescuers follow the voice prompts of the AED. However, sometimes a more experienced lay rescuer would start chest compressions just before the voice prompt, and sometimes chest compressions were started just after the voice prompt. In both cases, we marked the first chest com-pression as the start of the cycle.

Table 4. Survival Analyses

VariableOR (95% CI), Univariable

Analysis P ValueOR (95% CI), Multivariable

Analysis P Value

Ventilation duration of 3–5 s Reference Reference

Ventilation duration of 6–7 s 2.14 (0.70–6.55) 0.183 1.62 (0.43–6.10) 0.48



Ventilation duration ≥13 s 5.55 (1.55–19.8) 0.008 2.38 (0.46–12.1) 0.30

Time from emergency call to AED attachment 0.82 (0.74–0.90) <0.001 0.81 (0.71–0.92) <0.001

Dispatched first responder/onsite rescuers 0.29 (0.14–0.58) <0.001 0.67 (0.27–1.64) 0.38

VF as initial rhythm 26.2 (7.77–88.22) <0.001 32.6 (8.86–120.1) <0.001

CI indicates confidence interval; OR, odds ratio for survival; and VF, ventricular fibrillation.


RCP spécialisée : ventile-t-on bien ?

Do we hyperventilate cardiac arrest patients ? F John Resuscitation, 2007 Do we hyperventilate cardiac arrest patients? 85

Figure 3 A typical recording (30 s) of ventilatory vari-ables during resuscitation, demonstrating persistentlyhigh airway pressures and high respiratory rate.

in two patients. In 11/12 (91.7%) patients, the air-way pressure remained positive for more than 90%of the time. This contrasts with pre-hospital stud-ies that document a positive airway pressure for50%1 and 47.3%2 of the time, despite higher ventila-tion rates. The difference between the two studiesmay be related to the use of the LUCAS thumper todeliver chest compressions which is more efficientthan manual compression12,13 and may contributeto an overall increase in mean intrathoracic pres-sure. The median airway pressure in this studywas 13.9 cmH2O which therefore requires a centralvenous pressure in excess of this value to enablevenous return to the heart. Although central venouspressures are generally higher during cardiac arrestbecause of venous pooling, airway pressures ofthis magnitude are likely to impair venous return.Increased airway pressures also reduce coronaryand cerebral perfusion pressures and animal studieshave shown that this may have further detrimentalhaemodynamic consequences.1,5,7

Animal studies have demonstrated the adverseeffects of hyperventilation on survival1—7 and it islikely that the similar human findings in this studywould also have an adverse impact on outcome.Guidelines on respiratory rates are well known, butit would appear that in practice they are not beingobserved, probably because of the performance ofCPR by overzealous but well-trained medical staff,as suggested in other studies.2 This highlights achallenge in translating theory and manikin-basedresuscitation training into practice. The use of audi-ble tone guidance may well assist in the delivery ofappropriate ventilation rates as has been demon-strated in a laboratory study.9

All reported studies of ventilation during car-diac arrest have demonstrated varying degrees ofhyperventilation. It is likely that hyperventilationis a widespread problem, both during hospital andpre-hospital resuscitation. The need to ventilatepatients at an appropriate rate should be empha-sised during all resuscitation training.

Conflict of interest

No author has any conflict of interest with the con-tents of this study.

References

1. Aufderheide TP, Sigurdsson G, Pirrallo RG, et al. Hyper-ventilation-induced hypotension during cardiopulmonaryresuscitation. Circulation 2004;109:1960—5.

2. Aufderheide TP, Lurie KG. Death by hyperventilation: a com-mon and life-threatening problem during cardiopulmonaryresuscitation. Crit Care Med 2004;32:S345—51.

3. Cheifetz IM, Craig DM, Quick G, et al. Increasing tidalvolumes and pulmonary overdistention adversely affect pul-monary vascular mechanics and cardiac output in a pediatricswine model. Crit Care Med 1998;26:710—6.

4. Karlsson T, Stjernstrom EL, Stjernstrom H, et al. Central andregional blood flow during hyperventilation. An experimen-tal study in the pig. Acta Anaesthesiol Scand 1994;38:180—6.

5. Pepe PE, Raedler C, Lurie KG, Wigginton JG. Emergency ven-tilatory management in hemorrhagic states: elemental ordetrimental? J Trauma 2003;54:1048—55.

6. Theres H, Binkau J, Laule M, et al. Phase-related changesin right ventricular cardiac output under volume-controlledmechanical ventilation with positive end-expiratory pres-sure. Crit Care Med 1999;27:953—8.

7. Yannopoulos D, Tang W, Roussos C, et al. Reducing ventilationfrequency during cardiopulmonary resuscitation in a porcinemodel of cardiac arrest. Respir Care 2005;50:628—35.

8. Rosengarten PL, Tuxen DV, Dziukas L, et al. Circulatoryarrest induced by intermittent positive pressure ventila-tion in a patient with severe asthma. Anaesth Intens Care1991;19:118—21.

9. Milander MM, Hiscok PS, Sanders AB, et al. Chest compressionand ventilation rates during cardiopulmonary resuscitation:the effects of audible tone guidance. Acad Emerg Med1995;2:708—13.

10. Nolan JP, Deakin CD, Soar J, et al. European ResuscitationCouncil guidelines for resuscitation 2005. Section 4. Adultadvanced life support. Resuscitation 2005;67:S39—86.

11. Anon. Part 1: Introduction to the International Guidelines2000 for CPR and ECC. A consensus on science. EuropeanResuscitation Council. Resuscitation 2000;46:3—15.

12. Steen S, Liao Q, Pierre L, et al. Evaluation of LUCAS,a new device for automatic mechanical compression andactive decompression resuscitation. Resuscitation 2002;55:285—99.

13. Rubertsson S, Karlsten R. Increased cortical cerebral bloodflow with LUCAS; a new device for mechanical chestcompressions compared to standard external compressionsduring experimental cardiopulmonary resuscitation. Resus-citation 2005;65:357—63.

Ø  Patients ayant un ACR extrahospitalier et admis aux urgences, intubés et ventilés

Ø  MCE par le Système LUCAS

Ø  Ventilation manuelle par ballon par un médecin sénior

Ø  Mesure de la FR, Vt, Paw, Peep, Pmean, Pi, EtCO2

Ø  La FR maximum et médiane sont mesurée sur la période de ventilation continue la plus longue

Ø  12 patients inclus

Ø  Hyperventilation en rapport avec une augmentation de la fréquence ventilatoire

84 J.F. O’Neill, C.D. Deakin

Table 1 Summary of ventilatory variables resulting from manual ventilation during cardiopulmonary resuscitationof 12 patients

Median Min MaxPatient weight (kg) 80.0 60 120Time from initial arrest (min) 43.0 29 56Minute volume (l/min) 13.0 4.6 21.3Respiratory rate—–median (min−1) 21.0 7 37Respiratory rate—–max (min−1) 25.5 9 41Tidal volume (ml) 618.5 374 923Peak end-expiratory pressure (cmH2O) 1.3 0 6.9Mean airway pressure (cmH2O) 13.9 5.1 37.4Peak inspiratory pressure (cmH2O) 60.6 46 106.1Compliance-dynamic (ml/cmH2O) 20.4 5 68.2% Time airway pressure >0 cmH2O (%) 95.3 87.9 100

Figure 1 Box and whisker plot showing distribution ofmean airway pressure during manual ventilation in 12patients during cardiac arrest. The boundaries of the boxindicate the 25th and 75th percentile, and the line withinthe box marks the median. Whiskers above and below thebox indicate the 90th and 10th percentiles, respectively.Outlying points are shown as full circles.

out-of-hospital. No patient survived. Evidence ofaspiration was present in three patients.

Median tidal volume was in excess of 10 ml/kg in3/12 patients.

Figure 1 shows the distribution of mean airwaypressure. Figure 2 shows distribution of respiratoryrate. Figure 3 shows a typical recording (30 s) ofventilatory variables during resuscitation, demon-strating persistently high airway pressures and highrespiratory rate.

Discussion

The results demonstrate that in this sample ofhospital patients undergoing manual ventilation

during CPR, hyperventilation occurred frequently.Hyperventilation was caused by excess respiratoryrates rather than excessive tidal volumes. Therespiratory rate was at least double that recom-mended in 9/12 (75%) patients whilst the tidalvolume was no higher than the recommended10 ml/kg10 in 9/12 (75%) patients. The respiratoryrates are similar to findings previously reportedin hospital1,2,9 and pre-hospital studies.1,2 This isthe first study we are aware of to report humanin vivo tidal volumes during cardiopulmonaryresuscitation.

The airway pressures recorded were high, witha maximum peak airway pressure over 100 cmH2O

Figure 2 Box and whisker plot showing respiratory rateduring manual ventilation in 12 patients during cardiacarrest. The boundaries of the box indicate the 25th and75th percentile, and the line within the box marks themedian. Whiskers above and below the box indicate the90th and 10th percentiles, respectively. Outlying pointsare shown as full circles. The bold horizontal line showsthe recommended respiratory rate.10

Hyperventilation-induced hypotension during cardiopulmonary resuscitation TP Aufderheide Circulation, 2004

results, animal studies were performed to determine thepotential hemodynamic and survival rate consequences ofexcessive ventilation rates.

MethodsClinical Observational StudyThis study was performed with an exception from informed consentrequirements for emergency research (21 §CFR Part 50.24) aftercommunity consultation and public notification. It was part of butunrelated to another study for which the Food and Drug Adminis-tration had approved an investigational device exemption. TheHuman Research Review Committee at the Medical College ofWisconsin approved the study.The clinical observational study was performed in the City of

Milwaukee, where basic life support and advanced life support EMSpersonnel respond in a tiered manner. Care is provided according toAHA guidelines. For the study, an additional research team includinga physician and paramedic were dispatched to the scene of eachpatient. Entry criteria for the study were (1) adult patients (presumedor known to be !21 years) believed to be in cardiac arrest ofpresumed cardiac cause and (2) patients who were successfullyintubated with an endotracheal tube who were undergoing CPR at thetime of scene arrival of the research team. A portable pressuremonitor (Propaq, Welch Allyn Protocol, Inc) was used for electronicmeasurement of airway pressures, a surrogate for intrathoracicpressures. After arrival at the scene and after patient intubation, theresearch team connected the noninvasive intrathoracic pressuresensor between the endotracheal tube and the bag-valve resuscitator.Ventilations were then continuously recorded until resuscitationattempts were discontinued or the patient was resuscitated. There area variety of factors that may affect ventilation rate throughout theresuscitation efforts, including the practice of hyperventilating im-mediately before and after intubation. For this reason, we sought todetermine the maximum ventilation rate, defined as the highestventilation rate recorded during CPR over a 16-second periodoccurring at least 2 minutes after intubation. The ventilation fre-quency, duration, and percentage of time in which a positive pressurewas recorded in the lungs were then calculated with a digital caliper.The first 7 consecutive cases constitute group 1. After recognizing

that rescuers were consistently hyperventilating patients in cardiacarrest, investigators immediately retrained all EMS personnel toprovide ventilations at a rate of 12 breaths per minute during CPRafter establishment of a secured airway. The duration of eachventilation was not addressed during retraining. The subsequent 6consecutive cases (after retraining) constitute group 2. Data werealso analyzed by combining groups 1 and 2 (group 3). Differencesbetween the means of groups 1 and 2 were statistically analyzed byANOVA. A probability value of !0.05 was considered statisticallysignificant. All data are expressed as mean"SEM.

Results: Clinical Observational StudyThe average age of the 13 consecutive patients (6 women, 7 men)was 63"5.8 years (range, 34 to 96); 3 patients had an initial rhythmof ventricular fibrillation (VF), 5 had pulseless electrical activity, and5 had asystole. Overall, the maximum ventilation rate was observedan average of 18.8"11.9 minutes after intubation (range, 2 to 39minutes). No patient survived. The average maximum ventilationrate for group 1 patients was 37"4 breaths per minute (range, 19 to49), ventilation duration was 0.85"0.07 seconds/breath, and thepercentage of time in which a positive pressure was recorded in theairway was 50"4% (Table 1). After retraining, 3 of 6 group 2patients had ventilation rates !26 breaths per minute. The ventila-tion rate for these 6 patients was slower than in group 1 patients, at22"3 breaths per minute (range, 15 to 31). However, ventilationduration was significantly longer than in group 1 patients (1.18"0.06versus 0.85"0.07 seconds/breath, respectively, P!0.05). As a result,the percentage of time in which a positive pressure was recorded inthe airway was similar in group 2 and group 1 patients (44.5"8.2%versus 50"4%, respectively) (P#NS). Combining groups 1 and 2

(group 3), the ventilation rate for all 13 patients was 30 breaths perminute (twice the AHA-recommended rate).Individual recordings provide insight into the rate and duration of

ventilations provided by professional rescuers. Figure 1A representsdelivery of CPR relatively close to AHA guidelines. Only one suchcase was observed. Figure 1, B, C, and D illustrate representativeexamples of hyperventilation observed in the majority of casesbefore retraining. After retraining, slower ventilation rates were seenin group 2 patients, but ventilation duration was more prolonged(Figure 1E). As a result, the percentage of time in which a positivepressure was recorded in the airway was not significantly differentbetween groups 1 and 2.

Animal StudiesThe porcine hemodynamic and survival studies were approved by theCommittee of Animal Experimentation at the University of Minne-sota. The animals received care in compliance with the 1996 Guidefor the Care and Use of Laboratory Animals by the NationalResearch Council. The animal preparation and surgical techniqueshave been previously described in detail.3 Briefly, each animalreceived 10 mL (100 mg/mL) of intramuscular ketamine HCl forinitial sedation, followed by intravenous propofol (2.3-mg/kg bolusand then a constant intravenous infusion of 165 "g/kg per minute).During the preparatory phase, animals were ventilated with room airby a positive-pressure ventilator (Harvard Apparatus Co). The rateand tidal volume were adjusted to maintain an arterial carbon dioxide(PaCO2) at 40 mm Hg and oxygen saturation $90%, based onanalysis of arterial blood gases (IL Synthesis, InstrumentationLaboratory).Central aortic and right atrial pressures were recorded continu-

ously using a micromanometer-tipped catheter (Mikro-Tip Trans-ducer, Millar Instruments). All animals were treated with heparin(100 U/kg IV) as a single bolus once catheters were in place.Intrathoracic pressures were measured continuously with a micro-manometer-tipped catheter positioned within the trachea, 2 cm belowthe tip of the endotracheal tube at the level of the carina. End-tidalcarbon dioxide (ETCO2) was recorded continuously (CO2SMO Plus,Novametrix Medical Systems).

Resuscitation ProtocolsVentricular fibrillation was induced by using a 5F bipolar pacingcatheter (St Jude Medical Corp) placed into the right ventricle, withalternating current at 7 V and 60 Hz. As soon as VF was induced, thepositive-pressure ventilator was disconnected from the animal. After6 minutes of untreated VF, closed-chest standard CPR was per-formed continuously with a pneumatically-driven automatic pistondevice (CPR Controller, AMBU International).3 The compressionrate was 100 per minute with a 50% duty cycle, and the compressiondepth was 25% of the anterior-posterior diameter of the chest wall.After each compression, the chest wall was allowed to recoilcompletely and without any impedance from the compressiondevice. Pressure-controlled, synchronous ventilations were per-formed with a semiautomatic ventilator (Demand Valve ModelL063–05R, Life Support Products Inc) at a constant flow rate of 160L/min. Ventilation was initiated during the decompression phase ofCPR, and each breath was delivered over a 1-second period of time.

TABLE 1. Clinical Observational Study: Maximum VentilationRate, Duration, and Percentage of Time in Which a PositivePressure Was Recorded in the Lungs (Mean!SEM)

GroupVentilation Rate

(Breaths per Minute)Ventilation Duration

(Seconds per Breath)% PositivePressure

Group 1 37"4* 0.85"0.07† 50"4%

Group 2 22"3* 1.18"0.06† 44.5"8.2%

Group 3 30"3.2 1.0"0.7 47.3"4.3%

*P!0.05; †P!0.05; group 1, first 7 consecutive cases; group 2, subsequent6 consecutive cases (after retraining); group 3, groups 1 and 2 combined.

Aufderheide et al Hyperventilation-Induced Hypotension During CPR 1961


Ø  Etude clinique observationnelle

Ø  Patients ayant fait un ACR extrahospitalier pris en charge par l’équipe médicale d’urgence (EMU)

Ø  Mesure de la FR et de la durée moyenne d’un cycle respiratoire chez des patients intubés ventilés

Ø  Première phase: 7 ACR consécutifs (groupe 1)

Ø  Deuxième phase : formation de tout le personnel de l’EMU sur la fréquence respiratoire de 12/min

Ø  Troisième phase : 6 ACR consécutifs (groupe 2)

During the first 2 minutes of CPR, a compression-to-ventilation ratioof 5:1 was used on all animals.

Hemodynamic Protocol (Protocol I)After the initial 2 minutes of CPR, each animal received 3 differentventilation rates (12, 20, and 30 breaths per minute) in a computer-generated random order, with each phase lasting for 2 minutes. These3 different ventilation rate interventions were delivered in anasynchronous manner, either every 5 seconds (12 per minute), every3 seconds (20 per minute), or every other second (30 per minute),with each breath delivered over a period of 1 second.During CPR, aortic, right atrial, and intrathoracic pressures were

continuously recorded. ETCO2 and O2 saturation were also measuredcontinuously and recorded every minute. Arterial blood gases werecollected before induction of VF and at the end of each ventilationrate phase (after minute 8, 10, 12, and 14 of cardiac arrest).

Survival Protocol (Protocol II)Ventilation during the first 2 minutes of CPR was deliveredsynchronously with a 5:1 compression-to-ventilation ratio. After theinitial 2 minutes of CPR, each animal was randomized to receive 4minutes of CPR with 1 of 3 different ventilation modes: (1) 12breaths per minute with 100% O2; (2) 30 breaths per minute with100% O2; or (3) 30 breaths per minute with 5% CO2 and 95% O2.Five percent CO2 was added to inspiratory gases in the third group toevaluate the effect of hyperventilation on survival in the absence ofhypocarbia. During these interventions, ventilations were deliveredin an asynchronous manner every 5 seconds (12/min) or every othersecond (30/min), with each ventilation delivered over a period of 1second.During CPR, aortic, right atrial, and intrathoracic pressures as well

as ETCO2 and O2 saturation were continuously recorded. Arterialblood gas samples were assessed before induction of VF and at theend of each ventilation phase.At the end of each protocol, the animals were shocked with a

biphasic defibrillator (M Series, Zoll Medical Corp) using 150 J, upto 3 times, as needed.5 If resuscitation was successful, animals wereventilated with a ventilator and supplemental oxygen. Return ofspontaneous circulation (ROSC) was defined as a palpable pulseover 5 minutes. Survival was defined as a stable blood-perfusingrhythm generating a measurable blood pressure over the first hour ofobservation after resuscitation. No other therapeutic interventionswere performed after ROSC.At the end of each study protocol, the animals were euthanized

with an intravenous bolus of 60 mg propofol and then 10 mLpotassium chloride.All values are expressed as mean!SEM. Coronary perfusion

pressure was calculated as the difference between aortic diastolic andright atrial diastolic pressures. For each animal, 10 measurementswere performed for both aortic diastolic and right atrial diastolicpressures, and the average difference was used as the representativevalue for each animal. Mean intrathoracic pressure was measured asthe time-averaged value from continuous measurements acquiredover a 10-second period. Comparison between groups was done byANOVA and paired t test. Survival was calculated with !2 andFisher’s exact tests. A probability value of "0.05 was consideredstatistically significant.

ResultsAnimal Hemodynamic StudiesIncreased ventilation rate was associated with significantlyhigher mean intrathoracic pressures (P"0.0001) and signifi-cantly lower coronary perfusion pressures (P#0.03) andsignificantly higher arterial pH, but no change in PaO2 (Table2). There was also an increase in right atrial diastolic pressurewith increased ventilation rate (Figure 2). This was onlysignificantly lower in the 12–breaths/min versus 30–breaths/min groups (3.5!1.1 versus 7.3!1.0 mm Hg, P#0.02). The

Figure 1. A, This 16-second intrathoracic pressure recordingdepicts CPR performed relatively close to AHA guidelines.Large-amplitude waves represent ventilations (11 breaths perminute). Small-amplitude waves represent chest compressions(90 compressions per minute). B, This 64-second intrathoracicpressure recording (from group 1) demonstrates a ventilationrate of 47 breaths per minute. C, This 16-second intrathoracicpressure recording (from group 1) represents a ventilation rateof 38 breaths per minute. D, This 16-second intrathoracic pres-sure recording (from group 1) represents a ventilation rate of 34breaths per minute. E, After retraining, this 16-second recordingfrom a group 2 patient demonstrates a slower ventilation rate(11 breaths per minute) but increased ventilation duration (over4 seconds/breath), leaving little time (20%) during CPR fordevelopment of low or negative intrathoracic pressure.



Hyperventilation-induced hypotension during cardiopulmonary resuscitation TP Aufderheide Circulation, 2004

ROSC rate was 3 of 9 pigs; 2 of 3 pigs that survived received12 ventilations per minute as the terminal ventilation ratesequence.

Animal Survival StudiesThe survival rate in pigs ventilated at 12 breaths per minute(100% O2) was 6 of 7 (86%), compared with a survival rateof 1 of 7 (17%) at a rate of 30 breaths per minute (100% O2),and 1/7 (17%) at a ventilation rate of 30 breaths per minute(5% CO2/95% O2) (P!0.006) (Figure 3). Mean intrathoracicpressures were significantly higher with the higher ventilationrates (P"0.0001), and coronary perfusion pressures werelower (Table 3). Changes in arterial blood gases and ETCO2

with hyperventilation are shown in Table 4. Pigs ventilated at30 breaths per minute (100% O2) had lower levels of PaCO2(Table 4). Supplemental CO2 resulted in correction of hypo-capnia (Figure 3 and Table 4).

DiscussionThese results demonstrate that ventilation rates during theprehospital application of CPR in a city with well-trainedEMS personnel were observed to be far in excess of thoserecommended by the AHA. To our knowledge, this repre-sents the first time that ventilation frequency, duration, andpercent positive airway pressure have been objectively andelectronically recorded during CPR performed by profes-sional rescuers at the scene of out-of-hospital cardiac arrests.Both rapid-rate, short-duration ventilations and slow-rate,long-duration ventilations contributed to a high percentage oftime that pressure in the chest was increased. As confirmedby the porcine hemodynamic and survival studies, excessiveventilation rates during CPR resulted in increased positiveintrathoracic pressures, decreased coronary perfusion, anddecreased survival rates.During the decompression phase of standard CPR, a small

vacuum is created within the chest relative to the rest of thebody every time the chest wall recoils back to its restingposition.10 This draws venous blood back into the rightheart.10 Accentuating this small vacuum with use of an

TABLE 2. Animal Protocol I: Changes in Hemodynamics andArterial Blood Gases With Three Different Ventilation RatesDelivered in Random Order (Mean!SEM)

Ventilation Rate, Breaths per Minute

12 20 30 P

Hemodynamics

SAP, mm Hg 68.8#4.7 62.7#4.2 60.1#3.6 0.33

CPP, mm Hg 23.4#1.0 19.5#1.8 16.9#1.8 0.03

MIP, mm Hg per minute 7.1#0.7 11.6#0.7 17.5#1.0 "0.0001

Arterial blood gases

pH 7.34#0.02 7.45#0.03 7.52#0.03 0.0006

PaCO2, mm Hg 22.7#2.7 15.6#2.2 11.6#1.5 0.005

PaO2, mm Hg 340.9#40.7 403.3#47.0 403.7#48.0 0.59

SAP, Systolic aortic pressure; CPP, coronary perfusion pressure; MIP, meanintrathoracic pressure.

Statistical analysis was done by ANOVA. A value of P"0.05 was consideredstatistically significant.

Figure 2. Hemodynamic Study (n!9). Changes in meanintrathoracic pressure (MIP), coronary perfusion pressure (CPP),and right atrial diastolic pressure (RA diastolic) with differentventilation rates during resuscitation in a porcine model of car-diac arrest. Probability value of "0.05 was considered statisti-cally significant, based on ANOVA analysis of the 3 groups.

Figure 3. Survival Study (n!7 pigs per group). Changes inmean intrathoracic pressure (MIP), arterial CO2 (PaCO2), coronaryperfusion pressure (CPP), and survival rate, with hyperventilationand correction of hypocapnia ($CO2). Probability value of "0.05was considered statistically significant, based on ANOVA analy-sis of the 3 groups.

Aufderheide et al Hyperventilation-Induced Hypotension During CPR 1963


Ø  Etude animale

Ø  Cochon intubés, ventilés, FV induite, début de la RCP 6 min après début FV

Ø  MCE 100/min, mécanique

Ø  Ventilation par une valve à la demande, durée du cycle 1 sec

Ø  CPR 2 min avec rapport compression/ventilation de 5/1

Ø  Puis 3 groupes de 7 cochons : Ø  FR 12/min, FIO2 100% (groupe1)

Ø  FR 30/min, FIO2 100% (groupe 2)

Ø  FR 30/min FIO2 95%, FICO2 5% (groupe 3)

Ø  CPR pendant 4 min

Ø  Choc électrique (3 max)

Ø  Mesure de la pression aortique, pression de l’OD, pression intra-thoracique

Ø  Survie à une heure

Peut-on optimiser la ventilation ?

Ø  Animale, prospective

Ø  RCP standard / RCP standard + valve d’impédance inspiratoire (VII)

Ø  15 cochons, FV, RCP standard avec ou sans VII sur 4 périodes de 7 min

Ø  MCE 80/min, FR 16 /min, Vt 450 ml, Ambu

Ø  Mesure du débit sanguin ventriculaire (DSV), débit sanguin cérébral (DSC), pression de perfusion coronarienne (PPC)

Optimizing Standard CardiopulmonaryResuscitation With an InspiratoryImpedance Threshold Valve*Keith G. Lurie, MD; Katherine A. Mulligan, BA; Scott McKnite, BS;Barry Detloff, BA; Paul Lindstrom, BS; and Karl H. Lindner, MD

Objectives: This study was designed to assess whether intermittent impedance of inspiratory gasexchange improves the efficiency of standard cardiopulmonary resuscitation (CPR).Background: Standard CPR relies on the natural elastic recoil of the chest to transiently decreaseintrathoracic pressures and thereby promote venous blood return to the heart. To furtherenhance the negative intrathoracic pressures during the "relaxation" phase of CPR, we tested thehypothesis that intermittent impedance to inspiratory gases during standard CPR increasescoronary perfusion pressures and vital organ perfusion.Methods: CPR was performed with a pneumatically driven automated device in a porcine modelof ventricular fibrillation. Eight pigs were randomized to initially receive standard CPR alone,while seven pigs initially received standard CPR plus intermittent impedance to inspiratory gasexchange with a threshold valve set to .40 cm H20. The compressiomventilation ratio was 5:1and the compression rate was 80/min. At 7-min intervals the impedance threshold valve (ITV) waseither added or removed from the ventilation circuit such that during the 28 min of CPR, eachanimal received two 7-min periods of CPR with the ITV and two 7-min periods without the valve.Results: Vital organ blood flow was significantly higher during CPR performed with the ITV thanduring CPR performed without the valve. Total left ventricular blood flow (mean± SEM)(mL/min/g) was 0.32±0.04 vs 0.23±0.03 without the ITV (p<0.05). Cerebral blood flow (mL/min/g) was 20% higher with the ITV (+ITV, 0.23±0.02; -ITV, 0.19±0.02; p<0.05). Each time theITV was removed, there was a statistically significant decrease in the vital organ blood flow andcoronary perfusion pressure.Conclusions: Intermittent impedance to inspiratory flow of respiratory gases during standardCPR significantly improves CPR efficiency during ventricular fibrillation. These studies under¬score the importance of lowering intrathoracic pressures during the relaxation phase of CPR.

(CHEST 1998; 113:1084-90)

Key words: active compression-decompression CPR; cardiac arrest; cardiopulmonary resuscitation; heart; impedancethreshold valve; ventricular fibrillation

Abbreviations: ACD=active compression-decompression; CPP=coronary perfusion pressure; CPR=cardiopulmonaryresuscitation; ITV=impedance threshold valve; NS=not significant

HP he potential value of increasing negative in--¦¦ trathoracic pressure during the decompressionphase of cardiopulmonary resuscitation (CPR) with a

new technique termed active compression-decom¬pression (ACD) CPR has been described recently.1-5ACD CPR enhances the bellows-like action of the

*From the Cardiac Arrhythmia Center (Dr. Lurie, Ms. Mulligan,and Messrs. McKnite, Detloff, and Lindstrom), CardiovascularDivision, University of Minnesota, Minneapolis, and the Depart¬ment of Anesthesiology and Critical Care Medicine (Dr. Lind¬ner), University of Ulm, Uhn, Germany.Supported by the American Heart Association.Manuscript received April 16, 1997; revision accepted September10, 1997.Reprint requests: Keith Lurie, MD, UMHC, Box 508, 420 Dela¬ware Street SE, Minneapolis, MN 55409

chest. Use of this method is associated with im¬proved hemodynamic status in animal models andhumans when compared with conventional manualCPR.15 More recently, we demonstrated that theefficacy of ACD CPR could be further improved byinsertion of an inspiratory impedance threshold valve(ITV) into the respiratory circuit.6 In a porcinemodel of ventricular fibrillation, we observed thatuse of the ITV during ACD CPR, which physiolog¬ically mimics the clinical Mueller maneuver, lowersintrathoracic pressure during the decompressionphase, thereby enhancing vital organ blood flow andlowering defibrillation energy requirements whencompared with ACD CPR alone.6 In the presentstudy, we hypothesized that the use of an inspiratoiy

1084 Laboratory and Animal Investigations

Downloaded From: http://journal.publications.chestnet.org/ by a Service Gestion du Siege User on 11/02/2014

Chest, 1998

impedance valve during standard closed-chest CPR,which also relies on the bellows-like action of thethorax, would enhance the negative intrathoracicpressure generated by the natural recoil of the chestduring the decompression phase. As such, the pur¬pose of the present investigation was to test thishypothesis by insertion of an inspiratory ITV into therespiratory circuit. We describe the measurementsol central aortic pressures, coronary perfusion pres¬sures, vital organ blood flow with radiolabeled mi-crospheres, and arterial blood gases in a well-established porcine model of ventricular fibrillation.

Materials and Methods

Experiments described in this report were approved by theCommittee on Animal Experimentation at the University ofMinnesota. Healthy domestic female farm pigs (28 to 33 kg) werefasted overnight and anesthetized with pentobarbital, as previ¬ously described.16 The anesthetic and surgical approach duringthe preparatory phase have also been described recently.16 Leftventricular and ascending aortic arch BPs as well as right atrialpressures were monitored using high-fidelity micromanometercatheters (Millar Instruments; Houston). A single aorta-left ven¬

tricular micromanometer catheter with a lumen for injectingradiolabeled microspheres was used. A 5F bipolar pacing cathe¬ter (Daig Inc; Minnetonka, Minn) was used to induce ventricularfibrillation.6 Core temperatures were maintained between 36.5and 38.5° C. Fifteen minutes prior to induction of ventricularfibrillation, 5,000 U of sodium heparin was administered IV.

Closed-chest CPR was performed with a 6.5-cm circularcompression pad positioned over the sternum. The automateddevice and the measurement of hemodynamic parameters havebeen described previously.16 Compression and decompressionexcursion was measured continuously by the voltage output of a

linear variable differential transformer.6 Compression-decom¬pression forces were similarly monitored continuously using a

piezo electric force transducer.6 These data, which included allhemodynamic measurements, assessment of the distal trachealpressure from a fluid-filled pressure transducer connected to thedistal end of the endotracheal tube, and measurements ofcompression/decompression chest excursion and compression/decompression forces, were digitized on-line (SUPERSCOPE IIv.295; GW Instruments; Somerville, Pa) and analyzed electroni¬cally using a computerized recording system (Power Macintosh7100/66 computer; Apple Computer; Cupertino, Calif).6The protocol was designed to compare standard CPR alone

with standard CPR plus an inspiratory ITV. Each pig served as itsown control. The experimental protocol is seen in the schematicin Figure 1.Once catheters were placed into the left ventricle, right atrium,

and aorta, the right atrial diastolic pressures were maintained at2.5 to 5 mm Hg with IV normal saline solution. Within 10 min ofinducing ventricular fibrillation, the first radiolabeled micro-

sphere was injected into the left ventricle to measure baselineblood flows. Once the reference arterial blood was collected,ventricular fibrillation was induced with a single 3- to 5-sapplication of alternating current directly to the right ventricle viathe pacing catheter. The point at which ventricular fibrillationwas induced was termed "time 0." At this point, the pigs were

assigned randomly to initially receive either CPR alone or CPRplus the ITV. The endotracheal tube (ET tube high-low JET,Mallinckrodt Inc; St. Louis) was immediately disconnected fromthe mechanical ventilator and the tube cuffpressure was assessedto ensure that it was adequate to seal the trachea. After 3 min ofventricular fibrillation, during which time no chest compressionsor ventilation was performed, CPR was initiated with the auto¬mated device. The compression rate was 80/min with a 50% dutycycle, a depth of 25% of the anterior-posterior diameter of thechest wall, and a velocity of 7.5 inches/s. Active upward move¬

ment of the compression pad was performed with the same

velocity as during the compression phase such that it did notimpede the natural recoil of the chest. Compression and decom¬pression excursion was continuously monitored and adjusted witha control module to ensure that compression depth was adequateand that during the decompression phase the pad did not impedechest wall relaxation.6

Ventilatory support and automated standard CPR were per¬formed simultaneously during CPR. Ventilation was provided bymanual bag ventilation (Ambu bag; Glostrup, Denmark) andoxygen (8 L/min) as previously described.3 Ventilatory supportwas continued throughout all experiments using manual bagventilation with 10 L oxygen supplementation. Hand-held venti¬lation was utilized after preliminary experiments showed that itwas easier to interpose manual ventilatory efforts at the end ofthe decompression phase with the bag ventilation than withmechanical ventilation. Moreover, the mechanical ventilatorsavailable to us (including the Harvard animal ventilator [HarvardApparatus; Dover, Mass] and the Siemens ventilator [Siemens;Munich Germany]), which we have previously used,136 had a

significant amount of resistance to inspiration. That inspiratoryresistance prevented us from testing our overall hypothesis.During CPR, respirations were delivered continuously at a rate of16/min (one breath every five chest compressions) at a constanttidal volume of approximately 450 mL. As previously described,ventilations were delivered during the decompression phase ofCPR.16The ITV in this study consisted of two 20 cm H20 threshold

valves (Ambu Anesthesia PEEP Valve 20, No. 194011000; Ambu,Inc; Glostrup, Denmark) connected in series between the endo¬tracheal tube and the Ambu bag such that during the decom¬pression phase, but in the absence of manual ventilation, thevalves opened only with greater than .40 cm H20 of inspiratorypressure. In this fashion, more than .40 cm H20 of intrathoracicpressure was required for inspiration of respiratory gases duringfour of every five compression cycles during performance of CPRwith the ITV. With standard CPR and without active bagventilation, use of these threshold valves in series resulted ineffectively no inspiratory movement of respiratory gases duringthe decompression phase of CPR. As shown in the protocol timeline, at 7-min intervals, the ITV was either added or removed

Time: 0 10 12 17 19 24 26

VF Start bead 1 ±ITVCPR(±ITV)

bead2 dTV bead 3 ±ITV bead 4

Figure 1. Experimental protocol.

CHEST/113/4/APRIL, 1998 1085


surgical interventions to avoid all unnecessary suffering. This studywas performed according to Utstein-style guidelines7 on 40 femalefarm pigs weighing 28 to 33 kg. Each animal received 7 mL (100mg/mL) of ketamine HCl (Ketaset, Fort Dodge Animal Health) IMfor initial sedation. Intravenous access with an 18-gauge angiocath-eter (Jelco Ethicon, Inc) was rapidly obtained through a lateral earvein. Propofol anesthesia (PropoFlo, Abbott Laboratories) (2.3mg/kg) was initially delivered as an intravenous bolus. While theanimals were spontaneously breathing but heavily sedated, they wereintubated with a 7.0F endotracheal tube (Medline Industries Inc).The animals were then given an additional 30 mg of propofol andwere maintained on a propofol infusion of 160 !g · kg!1 ·min!1 untiljust after induction of ventricular fibrillation.The animals were positioned in the supine position. Femoral artery

cannulation was performed under aseptic conditions, and arterialblood pressures were monitored and recorded as previously de-scribed.2,6 Continuous ECG monitoring was recorded with a lead IIECG. Data were digitized, recorded, and analyzed as previouslydescribed.2,3,6 Intrathoracic pressures were measured with a micro-manometer-tipped catheter positioned 2 cm below the tip of theendotracheal tube. End tidal CO2 (ETCO2) (CO2 SMO Plus Respi-ratory Profile Monitor, Novametrix Medical Systems), arterial pres-sures, and intrathoracic pressures were recorded continuously duringboth the preparatory phase and the experimental protocol. Animalsreceived 400 mL of normal saline before the induction of ventricularfibrillation. Temperature was recorded with a rectal thermometer andmaintained between 37.5°C and 39.5°C with either a fan and acooling blanket or a Bair Hugger, Temperature Management System(Augustine Medical, model 505), as needed.Animals were positioned on a rigid cradle for standard CPR with

an automated CPR device. A circular compression pad with adiameter of 6.4 cm was attached to a pneumatically drivencompression-piston device (CPR Controller, Ambu International)and positioned over the lower third of the sternum.During the preparatory phase, the animals were ventilated with

room air by use of a positive-pressure ventilator (Harvard ApparatusCo, model 607) at an average rate of 16 breaths per minute and atidal volume of 20 mL/kg. The rate was adjusted on the basis ofanalysis of arterial blood gases every 30 minutes (IL Synthesis,model 20, Instrumentation Laboratory).

ProtocolVentricular fibrillation was induced in the anesthetized animals witha 3-second, 60-Hz, 140- to 160-V AC shock applied across thethorax with 2 half-circle stainless steel surgical needles as electrodes.CPR was performed continuously at a rate of 80 compressions perminute, with a compression-decompression duty cycle of 50%.Compressions were performed to a depth of 25% of the anteropos-terior diameter of the thorax with a circular compression pad. Duringthe decompression phase, the compression pad was elevated at a rateof 7.5 in/s to allow for the natural recoil of the anterior chest wall.After ventricular fibrillation was induced, the intravenous saline andpropofol infusions were immediately discontinued, and the animalwas disconnected from the mechanical ventilator. After 6 minutes ofuntreated cardiac arrest, standard closed-chest CPR was deliveredcontinuously with an automated pneumatic piston device as de-scribed above. Thirty seconds before initiation of CPR, either a shamor an active ITV was attached to the endotracheal tube. The ITVsused in this study have been described previously in detail.5,6Assignment of each valve was made according to a computer-generated randomization list. Researchers were blinded to the kind ofvalve used until after the pigs were euthanized 24 hours afterresuscitation.During CPR, ventilation was delivered during the decompression

phase. Animals were ventilated during CPR with 100% O2 with ademand valve resuscitator (Life Support Products, Inc, model L063-05RM) through either a sham (nonfunctional) or active functionalITV (ResQ-Valve, CPRx LLC) at a compression-to-ventilation ratioof 5:1. It was not possible, when looking at the blue impedancevalves, to determine whether or not there was a silicone diaphragmwithin the valve. In addition, the silicone diaphragm venting ports

were occluded during the manufacturing of the sham valves, suchthat they functioned as a hollow conduit for respiratory gas ex-change. As such, half of the ITVs were made as sham valves and theother half were active. Figure 1 depicts the function of the ITVduring the chest compression and decompression phases of CPR.2,5,6After 12 minutes of ventricular fibrillation and a total of 6 minutes

of CPR, the impedance valve was removed, and each animal wasimmediately defibrillated with up to 3 sequential 200-J transthoracicmonophasic shocks (Lifepak 6, Physio-Control). Animals that weresuccessfully resuscitated were reconnected to the automatic ventila-tor. Animals that remained in cardiac arrest received a single dose ofintravenous epinephrine (0.045 mg/kg) and an additional 90 secondsof CPR with the previously assigned sham or active ITV. Eachanimal that remained in cardiac arrest then received up to 3additional sequential 200-J transthoracic shocks. All animals withsuccessful restoration of spontaneous circulation were treated withintravenous fluids. Dopamine, at a concentration of 1.6 mg/mL, wasadministered to maintain the systolic blood pressure at "70 mm Hg,as needed. Mechanical ventilation with supplemental oxygen (10L/min) was continued throughout the immediate postresuscitationperiod. The endotracheal tube was removed once the animal was ableto breathe independently, as judged by measurement of peak inspira-tory flow rates and maintenance of adequate ventilation as themechanical ventilation rate was progressively reduced. Each animalwas transferred to a heated holding area until it woke up and was ableto move around and drink water independently. The survivors wereheld in an observation area for 24 hours before undergoing furtherassessment. They then underwent euthanasia and autopsy 24 hoursafter resuscitation.Survival rates, complication rates, and neurological status were

evaluated 24 hours after resuscitation. Neurological function wasevaluated 24 hours after resuscitation by 2 investigators (S.M., T.Z.)who remained blinded to the device that was used. Evidence ofpulmonary congestion, as judged by blood gas analysis, was assessed1 to 3 hours after resuscitation. Pulmonary congestion was alsoassessed by observing for pink frothy exudate in the endotrachealtube during and after CPR and at autopsy. Neurological function wasassessed quantitatively, as described by Bircher, Safar, Vaagenes,and colleagues.8,9 The Swine Neurologic Deficit Score was used toevaluate level of consciousness, respiratory pattern, cranial nervefunction, motor and sensory function, and behavior evaluation,including ability to drink, chew, stand, and walk.8 The CerebralPerformance Score was also used.9 It is a neurological assessmentbased on a 5-point evaluation of the level of consciousness.

Statistical AnalysisHemodynamic and perfusion parameters were analyzed by ANOVA(a value of P#0.05 was considered statistically significant). The "2

test was used for survival rate analysis. A priori, the sample size was

Figure 1. Schematic of respiratory gas flow through ITV.

Lurie et al Impedance Valve Improves Outcome After VF in Pigs 125

by guest on November 3, 2014http://circ.ahajournals.org/Downloaded from

Ø  DSV moyen plus élevé dans le groupe avec VII : 0,32±0,11 ml/min/g vs 0,23±0,05 ml/min/g ; p < 0,05

Ø  DSC moyen plus élevé dans le groupe avec VII : 0,23±0,02 vs 0,19±0,02 ; p< 0,05

Ø  PPC moyenne est plus élevée dans le groupe avec VVI : 14,8±1,3 mm Hg vs 12,5±1,5 mm Hg ; p = 0,07

Ø  Augmentation de 20% de la PPC mais de 40% du DSV

Ø  Effet de l’augmentation du retour veineux mais également d’autre mécanismes permetant une majoration de la perfusion myocardique



(CHEST 1998; 113:1084-90)









Chest, 1998



(CHEST 1998; 113:1084-90)









2a.5o

0.6

0.54

1 0.4H

is"5 ° 3

> o.H

0.010 20

Minutes After VF

2 b.

o

&

0.4

0.3H

1!

0.010 20

Minutes After VF.i30

2c.3<A(A0)

.2 G»

itCOco

20 H

15

10

10 20Minutes After VF

30

Figure 2. Top (a): myocardial blood flow (mean±SEM) assessed 2, 9, 16, and 23 min after initiationof either standard CPR alone (open circles) or standard CPR plus the ITV (closed circles). After 7 minof CPR, the ITV was either added to or removed from the ventilatory circuit. Asterisk indicates p<0.05.Center (b): brain blood flow (mean±SEM) assessed 2, 9, 16, and 23 min after initiation of eitherstandard CPR alone (open circles) or standard CPR plus the ITV (closed circles). After 7 min of CPR,the ITV was either added to or removed from the ventilatory circuit. Asterisk indicates p<0.05 whenanalyzing the change in blood flow at different time points; double daggers indicate p<0.05 for theblood flow at a given time point. Bottom (c): the mean (±SEM) CPP was calculated by the differencebetween the diastolic aortic and right atrial pressures 2, 9, 16, and 23 min after initiation of eitherstandard CPR alone (open circles) or standard CPR plus the ITV (closed circles). After 7 min of CPR,the ITV was either added to or removed from the ventilatory circuit. Asterisk indicates p<0.05.

CHEST/113/4/APRIL, 1998 1087


Chest, 1998

Ø  Animale, prospective

Ø  40 cochons, 20 dans chaque groupe

Ø  Sédatés, ventilés

Ø  FV pendant 6 min puis RCP avec une valve factice ou une VII

Ø  Après 6 min de RCP, les valves sont retirées et l’animal est choqué ± Adrénaline en fonction d’une RACS

Ø  Critère de jugement : évolution neurologique à H 24

Use of an Inspiratory Impedance Valve ImprovesNeurologically Intact Survival in a Porcine Model of

Ventricular FibrillationKeith G. Lurie, MD; Todd Zielinski, MS; Scott McKnite, BS;

Tom Aufderheide, MD; Wolfgang Voelckel, MD

Background—This study evaluated the potential for an inspiratory impedance threshold valve (ITV) to improve 24-hoursurvival and neurological function in a pig model of cardiac arrest.

Methods and Results—Using a randomized, prospective, and blinded design, we compared the effects of a sham versusactive ITV on 24-hour survival and neurological function. After 6 minutes of ventricular fibrillation (VF), followed by6 minutes of cardiopulmonary resuscitation (CPR) with either a sham or an active valve, anesthetized pigs received 3sequential 200-J shocks. If VF persisted, they received epinephrine (0.045 mg/kg), 90 seconds of CPR, and 3 more 200-Jshocks. A total of 11 of 20 pigs (55%) in the sham versus 17 of 20 (85%) in the active valve group survived for 24 hours(P!0.05). Neurological scores were significantly higher with the active valve; the cerebral performance score(1"normal, 5"brain death) was 2.2#0.2 with the sham ITV versus 1.4#0.2 with the active valve (P!0.05). A totalof 1 of 11 in the sham versus 12 of 17 in the active valve group had completely normal neurological function (P!0.05).Peak end-tidal CO2 (PETCO2) values were significantly higher with the active valve (20.4#1.0) than the sham (16.8#1.5)(P!0.05). PETCO2 $18 mm Hg correlated with increased survival (P!0.05).

Conclusions—Use of a functional ITV during standard CPR significantly improved 24-hour survival rates and neurologicalrecovery. PETCO2 and systolic blood pressure were also significantly higher in the active valve group. These data supportfurther evaluation of ITV during standard CPR. (Circulation. 2002;105:124-129.)

Key Words: cardiac arrest ! fibrillation ! cardiopulmonary resuscitation ! valves ! survival ! arrhythmia ! brain

Survival rates remain poor for most patients who sufferfrom a cardiac arrest. Studies on the mechanism of blood

flow during cardiopulmonary resuscitation (CPR) have re-cently focused on the importance of the decompression phaseof CPR.1–4 During the decompression phase of standard CPR,a small vacuum is created within the chest relative to the restof the body every time the chest wall recoils back to itsresting position.5 This draws venous blood back into the rightheart. In addition, during the decompression phase of stan-dard CPR, air is drawn into the lungs. We previouslydescribed the use of an impedance threshold valve (ITV) toprevent the inflow of respiratory gases during the active chestwall recoil phase, or decompression phase, of standardCPR.4,5 The ITV is a small (35-mL) disposable plastic valvethat is attached to the endotracheal tube or a face mask. Itworks by allowing the rescuer to freely ventilate the patientbut impeding inspiratory airflow during the decompressionphase of CPR when the patient is not being actively venti-

lated. This creates a small vacuum within the chest to furtherenhance venous return.We recently demonstrated in a porcine model that use of

the ITV resulted in a nearly 2-fold increase in blood flow tothe brain and the heart after 6 minutes of ventricular fibril-lation and 6 minutes of standard CPR.6 Although use of theITV during standard CPR has been reported previously in 2studies involving $30 animals, to date there have been nodefinitive data in support of a survival benefit from the use ofthe ITV with standard CPR.4,6 Thus, the purpose of thisinvestigation was to test the hypothesis that the ITV wouldimprove neurological function and 24-hour survival in anestablished animal model of cardiac arrest during perfor-mance of standard CPR.

MethodsPreparatory PhaseThe study was approved by the Committee of Animal Experimen-tation at the University of Minnesota. Anesthesia was used in all

Received August 7, 2001; revision received October 12, 2001; accepted October 15, 2001.From the Cardiac Arrhythmia Center, University of Minnesota, Minneapolis (K.G.L., T.Z., S.M., W.V.); the Department of Emergency Medicine,

Medical College of Wisconsin, Milwaukee (T.A.); and the Department of Anesthesiology, Leopold-Franzens-University, Innsbruck, Austria (W.V.).Dr Keith G. Lurie is the principal investigator on an NIH SBIR grant awarded to CPRx LLC. He is a coinventor of the inspiratory impedance threshold

valve and founded CPRx LLC to develop this device.Correspondence to Keith G. Lurie, MD, Department of Medicine, University of Minnesota, MMC 508, AHC, 420 Delaware St SE, Minneapolis, MN

55455. E-mail [email protected]© 2002 American Heart Association, Inc.Circulation is available at http://www.circulationaha.org

124 by guest on November 3, 2014http://circ.ahajournals.org/Downloaded from

Circulation, 2002

calculated on the basis of expected differences in 24-hour survivalbetween groups. All data are expressed as mean!SEM.

ResultsSurvival OutcomesThe main study end points were 24-hour survival andneurological function. Twenty animals received standardCPR plus a sham valve, and 20 received standard CPR plusan active valve. Twenty-four-hour survival was 55% in thesham valve group and 85% in the active valve group(P"0.05).Neurological function was significantly improved in the

24-hour survivors that received treatment with the activevalve. The cerebral performance score, based on a 5-pointscoring system (1#normal, 5#brain death),9 was 2.2!0.2 foranimals treated with the sham valve versus 1.4!0.2 for thosetreated with the active valve (P"0.002). With the SwineNeurologic Deficit Score,8 there were similar statisticallysignificant improvements in the active valve group (Table 1,Figure 2). The neurological deficits in survivors treated with

the sham valve were striking. The animals often appeareddocile and disoriented, frequently walking into the wall of thecage without apparently realizing that it was there. Only 1 of11 survivors had a completely normal neurological score. Bycontrast, 12 of 17 survivors treated with the active valve hadnormal neurological function (P"0.05). Improved neurolog-ical function was observed in many of the different categoriesin animals treated with the active valve (Tables 1 and 2).The initial return of spontaneous circulation (ROSC) rate

was 80% in the sham valve group and 95% in the active valvegroup. For those animals that had ROSC after the first 3defibrillatory shocks, there was no significant difference inthe amount of energy needed between groups. Withoutepinephrine treatment (ie, with only Basic Life Support[BLS] techniques during the resuscitation), 35% of the pigs inthe sham valve group versus 60% in the active valve groupcould be resuscitated initially (P#NS). Five pigs in the shamvalve group and 2 in the active valve group died after beingsuccessfully resuscitated, but before 24 hours.We also analyzed 24-hour survival results in the subset of

animals that were resuscitated with DC shock alone, withoutepinephrine. With BLS followed by defibrillation, 24-hoursurvival in the sham controls was 30% versus 60% in theactive valve group (P#0.057). The neurological score in thissubgroup also showed a significant improvement in the activevalve group (Table 2). Similarly, when advanced life support(ALS) measures were needed and used, use of the ITV alsoresulted in a statistically significant improvement in theneurological status of the animals 24 hours after resuscitation.Epinephrine therapy had no observed beneficial impact on

neurological function in the sham valve group but didimprove the likelihood of successful resuscitation in both thesham and active valve groups. The cerebral performancescore was 2.0!0.4 in the sham valve group with BLS aloneversus 2.3!0.3 in the sham valve group that received ALS(epinephrine and additional shocks). The total neurologicaldeficit scores were similar as well, with and without epineph-rine treatment, in animals treated with the sham valve (Table2). By contrast, pigs requiring epinephrine in the active valve

TABLE 1. Twenty-Four Hour Survival and NeurologicalAssessment Score

Sham Valve (n#20) Active Valve (n#20)

24-hour survival, n (%) 11 (55)* 17 (85)*

Neurological assessment

Consciousness 25.0!6.2 10.6!4.4*

Respiratory pattern 10.8!8.5* 0.0!0.0*

Painful stimulus 13.3!4.1 4.7!2.1

Muscle tone 16.7!5.6 5.9!2.7

Standing 5.0!2.6 1.2!1.2

Walking 13.3!3.3 5.3!2.1*

Restraint 30.8!5.3* 12.9!4.8*

Total deficit score 16.4!3.3† 5.8!1.8†

*P"0.05.†P"0.02.

Figure 2. Pittsburgh neurological deficit score for all animalsreceiving standard CPR with either sham (n#11 of 20 survivors24 hours after resuscitation) or active (n#17 of 20 survivors 24hours after resuscitation) valve and subgroup of animals thatwere resuscitated without (w/o) epinephrine. All values aremean!SEM. *P"0.03.

TABLE 2. Twenty-Four Hour Survival and NeurologicalAssessment Score by Level of Care (BLS or ALS) Received

Sham Valve (n#20) Active Valve (n#20)

BLS ALS BLS ALS

24-hour survival, n (%) 6 (30) 5 (25) 12 (60) 5 (25)

Neurological assessment

Consciousness 25.0!9.2 25.0!9.2* 13.9!5.5 0.0!0.0*

Respiratory pattern 5.0!5.0 16.7!16.7* 0.0!0.0* 0.0!0.0

Painful stimulus 15.8!7.8 10.8!3.3* 6.6!2.6 0.0!0.0*

Muscle tone 18.3!8.9 12.5!6.0 6.0!2.8 6.3!6.3

Standing 6.7!4.2 3.3!3.3* 1.5!1.5 0.0!0.0*

Walking 13.3!5.6 13.3!4.2 6.2!2.7 2.5!2.5

Restraint 30.0!8.2 32.0!7.5* 15.4!6.0 5.0!5.0*

Total deficit score 16.3!3.4* 16.3!3.6† 7.0!2.2* 2.0!1.0†

*P"0.05.†P"0.002.

126 Circulation January 1/8, 2002



















Circulation, 2002

group that lived for 24 hours all had a normal cerebralperformance score.A total of 5 animals in the sham group and 4 in the active

valve group required dopamine supportive therapy afterresuscitation, secondary to hypotension. The duration ofsupport varied between 1 minute and 30 minutes and wassimilar between groups.

Hemodynamic OutcomesSystolic arterial pressures (Figure 3A) as well as peak ETCO2

(PETCO2) (Figure 4) levels were significantly higher in thegroup treated with the active valve. The systolic and diastolicpressures (Figure 3B) rose more rapidly and remained higherin the active valve group than in controls. With a !2 test, therewas a significantly greater chance for 24-hour survival when

the diastolic blood pressure was !21 mm Hg (80%) com-pared with animals with a diastolic blood pressure of"21 mm Hg (40%) (P"0.05).PETCO2 levels were significantly higher among survivors

than among the animals that died. There was a significantlygreater chance for 24-hour survival in animals with a maxi-mum PETCO2 level of#19 mm Hg (79%) than in animals witha peak PETCO2 value of "18 mm Hg (45%) (P"0.05).Changes in intrathoracic pressure were also monitored

continuously during the performance of CPR. As shown inFigure 5, intrathoracic pressures were consistently lower inthe active valve group. There was a significantly greaterchance for 24-hour survival when intrathoracic pressure was#1.5 mm Hg (80%) than in animals with thoracic pressure of"1.5 mm Hg (40%) (P"0.05).

Figure 3. Standard CPR was performed with either a sham oractive ITV. Systolic pressures (A) were recorded continuouslyduring 6-minute study period during CPR. Systolic pressures,plotted on y axis from 38 to 74 mm Hg, were significantly higherduring time points marked with asterisk. Diastolic arterial pres-sures (B) were also recorded continuously during study periodand from 16 to 34 mm Hg. Values are mean#SEM for timeperiod indicated on each graph. *P"0.05.

Figure 4. End-tidal CO2 values were measured over 6-minutestudy period. All values plotted from 10 to 24 mm Hg areexpressed as mean#SEM. Standard CPR was performed witheither a sham or active ITV. VF indicates ventricular fibrillation.*P"0.05.

Figure 5. Intrathoracic pressures were recorded continuouslyduring 6-minute study period. Values represent mean#SEM fortime period indicated on each graph. Standard CPR was per-formed with either a sham or active ITV. *P"0.05.

Lurie et al Impedance Valve Improves Outcome After VF in Pigs 127



















Circulation, 2002

Ø  Pas d’effet indésirables en particulier pas d’œdème pulmonaire clinique ni anatomopathologique

Ø Meilleur pronostic neurologique

Ø Meilleur EtCO2

Ø A évaluer avec d’autres études


















Le souffle c’est la vie……

Merci de votre attention

Health & Medicine

Ventilation de l'arrêt cardiaque : Pour par Michel RAMAKERS