ean Louis Gerard, MD, and André Khayat, MDepartments of Thoracic and Cardiovascular Surgery, Anesthesiology, Cardiology, and Emergency and Critical Care Medicine,
niversity Hospital, Caen, France
wbhp2t1
cpoardr
Background. The survival of patients after prolongedardiac arrest is still inadequate. Extracorporeal life sup-ort (ECLS) represents an alternative therapeutic methodor patients who do not respond to conventional cardio-ulmonary cerebral resuscitation. This technology issed to support the circulation of a patient with severeardiac failure.
Methods. Between June 1997 and January 2003, 40CLS procedures were performed in patients who pre-ented with refractory cardiac arrest. During externalardiac massage, the patient was connected to an extra-orporeal circuit by the insertion of an arterial andenous cannula through the femoral vessels. The extra-orporeal circuit included a centrifugal pump and anxygenator. Mean age was 42 � 15 years; the average timef external cardiac massage was 105 � 44 minutes.
Results. Once the circulation was restored, 22 patients
isefloosa
P
WetJcdrpcovmw
ardiovascular Surgery, University Hospital, 14033 Caen, France; e-mail:[email protected].
2005 by The Society of Thoracic Surgeonsublished by Elsevier Inc
ats.ctsnetjournDownloaded from
ere disconnected from the extracorporeal circulationecause of brain death or multiorgan failure; after 24ours, among the 18 survivors, 6 were weaned off theump, 9 were bridged to a ventricular assist device, andpatients were directly bridged to cardiac transplanta-
ion. Eight patients are alive and without any sequelae at8 month’s follow-up.Conclusions. In prolonged cardiac arrest with failing
onventional measures, rescue by extracorporeal supportrovides an ultimate therapeutic option with a goodutcome in survivors. Our results encourage the widerpplication of ECLS for refractory cardiocirculatory ar-est in selected patients. The high rate of neurologiceath needs further improvements in the early phase ofesuscitation maneuvers.
orldwide experience with cardiopulmonary resus-citation (CPR) has demonstrated that survival to
ischarge ranged from 8.2% to 22% in hospitalized pa-ients and was below 3% in out-of-hospital patients whoxperienced cardiac arrest [1–2]. Because of the lowurvival rate after prolonged CPR, methods that are moreggressive have been suggested to increase success.Extracorporeal life support (ECLS) refers to a technol-
gy that is used to support the circulation of a patientith severe cardiac failure. The physiologic objective is torovide temporary circulatory support to the vital organsnd to unload the failing heart as the injured myocar-ium attempts to recover. Indications for applying rescueCLS have not been clearly defined; however, guidelinesave been established to avoid futile efforts [3–6].Portable cardiopulmonary bypass (CPB) is a simple
nd effective system for controlling a patient’s circulatorynd respiratory functions on an emergency basis untilhysicians evaluate the cause of the patient’s hemody-amic shock and initiate further treatment [8, 9]. Because
his technology made consistent progress with miniatur-
ccepted for publication June 21, 2004.
ddress reprint requests to Dr Massetti, Department of Thoracic and
zed pumps and circuit biocompatibility, cardiothoracicurgeons and ECLS specialists have renewed their inter-st in the use of CPB as an emergency resuscitative toolor patients suffering from refractory cardiac arrest. Pub-ished series have demonstrated an overall survival ratef between 15% and 60% in this heterogeneous categoryf patients [4–11]. To evaluate the impact of ECLS on theurvival of patients who experienced a prolonged cardiacrrest, we retrospectively analyzed our experience.
atients and Methods
e reviewed the case histories of all patients treated withmergency CPB for prolonged cardiopulmonary arrest athe University Hospital of Caen between June 1997 andanuary 2003. Data collection was both retrospective andoncurrent. The tracking of each patient’s clinical courseuring the hospitalization was recorded. Vascular, neu-ologic, hemorrhagic, renal, and perfusion system com-lications were documented. We selected patients whoseonditions leading to CPR were thought to be of cardiacrigin. ECLS contraindications included previous irre-ersible brain damage, terminal malignancy, and age ofore than 75 years. Patients receiving ECLS support for
eaning from CPB, accidental hypothermia, or for respi-
179Ann Thorac Surg MASSETTI ET AL2005;79:178–84 EXTRACORPOREAL LIFE SUPPORT
CA
RD
IOV
AS
CU
LA
R
atory support were excluded. Briefly, patients werencluded into the ECLS therapy if they:
presented with refractory cardiac arrest requiringexternal cardiac massage;could not be returned to spontaneous circulationwithin 45 minutes;received ECLS in the hospital.
In ECLS-supported patients, we selected three groupsnd analyzed the population data, survival at 24 hours,nd weaning from ECLS followed by discharge from theospital. The group “survival at 24 hours” identified
hose patients who survived to cardiopulmonary resus-itation-advanced life support and showed no signs ofrreversible brain damage.
evice Descriptionhe hardware for emergent cardiopulmonary circulationonsisted of a Biomedicus portable bypass system (PBS)Medtronic, Inc, Minneapolis, MN) incorporating a cen-rifugal pump console and a water pump system. Areconnected tubing set was attached to a hollow-fiberembrane oxygenator with an integral heat exchanger
Maxima PRF, Medtronic, Inc), a constrained vortexump chamber, and a flow probe. The ECLS circuitonsisted of a closed Carmeda Bioactive Surface-coated
able 1. Clinical Pre-Extracorporeal Life Support Features oftudied Patients
AllPopulation
n � 40
Alive� 24 Hn � 18
Dischargedfrom Hospital
n � 8
ge (years) 42 � 15 41.4 � 17 34.6 � 15ex ratio (F/M) 17:23 8:10 4:4ut-of-Hospital CA 5 1 1nown cardiopathy 10 7 1waiting heart tx 4 3 0
ircuit of polyvinyl chloride tubing (Medtronic, Inc).annulas used were Biomedicus (17F to 25F), according
o the size of patients.
annulation Techniquence the decision to use ECLS support has been made,
he surgeon dissects the femoral vessels to the groin.eanwhile, the perfusionist assembles a circuit thateets the specific requirements of the patient’s size and
rimes it with Ringer’s lactate solution. Heparin 50 UI/kgs administered intravenously to the patient immediatelyefore cannulation of the vessels. Reinjection of a lowose of heparin allows the activated clotting time (ACT)
o be kept at between 150 and 180 seconds at full flow.Cannulation is peripheral (femorofemoral) using Biome-
nto the femoral vessels through a modified Seldingerechnique (surgical cut-down followed by vessel puncture).he distal tip of the arterial cannula is positioned in theommon iliac artery or distal abdominal aorta. The distal tipf the venous cannula is placed in the right atrium underchocardiography guidance and confirmed by chest radi-graphy. Limb ischemia used to be the major problem athe beginning of our experience; therefore, perfusion of theistal limb is accomplished with a small 8F catheter. Thisrterial shunt is instituted between the side port of therterial cannula and a point located some centimetersistally in the superficial femoral artery.Successful ECLS is defined as mean blood pressure of
t least 60 mm Hg and flows of at least 2.5 L/m2. Theiomedicus pump rpm is increased to approximately,500 rpm maximum or until desired flows are obtained.asopressor (norepinephrine) is infused to maintain aean systemic arterial pressure of more than 60 to 70 mmg. Ventricular filling and inotropic support maintain aulsatile flow through the native heart. The aim is toecompress the left heart and to minimize stasis and
herefore the risk of intracardiac clot.To accomplish mechanical decompression of the left
eart, we recently performed an atrial balloon septos-omy in one patient. A contralateral femoral vein ap-
able 2. Main Indications for Extracorporeal Life Support
180 MASSETTI ET AL Ann Thorac SurgEXTRACORPOREAL LIFE SUPPORT 2005;79:178–84
CA
RD
IOV
AS
CU
LA
R
y blade septostomy, was performed under combinedadioscopy and echocardiographic guidance. Sequentialalloon inflations were carried out to achieve left heartecompression that was confirmed by echocardiography.A pulmonary artery catheter is positioned in patientsithout any evidence of coagulopathy. Depending onyocardial contractility, inotropic support is reduced to
ecrease myocardial oxygen demand and facilitate recov-ry. A low tidal volume (5 to 6 mL/kg) of mechanicalentilation with positive end-expiratory pressure at 8 to0 cm H2O is applied to fight hydrostatic pressure that isue to postcapillary hypertension and to minimize pul-onary hemorrhage. All patients are anticoagulated with
eparin as early as possible to reach an ACT of approx-mately 160 to 180 seconds at full flow assistance, and anCT of more than 200 seconds when flow is reducedelow 1.5 L/min during the weaning phase.
anagement in the Intensive Care Unithe perfusionist operates the Biomedicus pump initially,nd once the patient is stabilized, the intensive care unitICU) nurses monitor the pump function. At that time, aerfusionist is available for occasional monitoring visitsnd emergencies. Continuous venovenous hemofiltra-ion is used to regulate intravascular volume and overalluid balance and to enable the administration of blood
ig 1. Outcome and extracorporeal life support (ECLS) time in re-uscitated cardiac arrest population. (Tx � transplantation; VAD �entricular assist device.)
ig 2. Overall survival in both populations. (Dday; ECLS � extracorporal life support; M �
onth.) Surviving patients are indicated at dia-ond for ECLS and at black square for ECLS
ontinued after day 1.
ats.ctsnetjournDownloaded from
roducts without induction of volume overload. Trans-sophageal echocardiography is used serially to enablehe assessment of progressive myocardial recovery ando exclude intracardiac clot or other abnormalities.
Biomedicus pump heads are changed if there is evidencef high plasma free hemoglobin on ECLS. The entire circuit
s usually changed when an oxygenator change out isequired, for example, if plasma water drips from the
axima Carmeda-coated oxygenator into the gas phase.The postresuscitation neurologic state is assessed clin-
cally by awaking patients. In case of coma, serial elec-roencephalogram (EEG) and transcranial Doppler ultra-ound is performed.
eaninghe decision to discontinue ECLS support is based uponombined criteria. The weaning protocol is performed byhe assessment of hemodynamic profile and myocardialunction under echocardiography during the progressiveeduction of pump flow to 500 mL/min. During thiseriod, the anticoagulation is adapted to adequate valuesf ACT (250 to 300 seconds). Weaning of ECLS is per-ormed if the left ventricular ejection fraction is stable� 50%). Echocardiographic assessment of myocardialunction remains, in our experience, the most importantredictor of successful weaning from ECLS.ECLS is terminated in the absence of an efficient
ardiac function recovery or when cardiac transplanta-ion or bridge to another mechanical device cannot beonsidered (futility criteria). Termination of ECLS isonsidered when there is evidence of multiorgan failure,verwhelming sepsis, or profound neurologic impact.The neurologic outcome at hospital discharge was
ssessed according to the cerebral performance classCPC) categories of:
. good cerebral performance,
. moderate cerebral disability,
. severe cerebral disability,
. coma or vegetative state, or
. brain death or death.
ata Analysisalues of continuous variables are expressed as means �
tandard deviation. Comparison of means was per-ormed using the independent sample t test and Mann-
181Ann Thorac Surg MASSETTI ET AL2005;79:178–84 EXTRACORPOREAL LIFE SUPPORT
CA
RD
IOV
AS
CU
LA
R
hitney test for nonparametric data. In addition, awo-tailed p value of less than 0.05 was considered tondicate statistical significance.
esults
orty patients were treated after meeting the inclusionriteria in this study; patient characteristics are shown inable 1. All received ECLS during external cardiac mas-age, and the main indications for ECLS support areisted in Table 2. The PBS setup was used in all cases
ithin the hospital: for 16 patients in the operatingheater, 11 in the intensive care unit, 2 in the coronaryntensive care unit, 7 in the cath-lab, and 4 in themergency department. Before the initiation of ECLSupport, 2 patients were on intraaortic balloon pump forschemic cardiomyopathy.
Femoral cannulation and stable extracorporeal flowas established in all patients; no retroperitoneal bleed-
ng or aortic or iliofemoral dissection occurred. Fiveatients needed early surgical revision at the site ofannulation for related bleeding dissection. Early in ourxperience, one patient developed severe limb ischemiand required urgent revascularization and fasciotomy.ubstantial bleeding requiring massive transfusion waslso a problem in 3 patients among the early 18 survivors.Time from cardiac arrest to initiation of ECLS was
xtremely variable, and 35 of 40 patients experiencedardiac arrest within the hospital. Among the 18 patients45%) who survived this early phase of resuscitation, 630%) demonstrated the return of intrinsic cardiac func-ion and were weaned off support with an average ECLSuration of 91 � 57 hours (range, 20 to 240). Two patientsere bridged to transplantation after 82 � 68 hours
range, 30 to 180), and 9 patients (50%) were transferredo a ventricular assist device (VAD) after 37 � 26 hoursrange, 4 to 90 hours) (Fig 1). Eight patients survivedospital discharge, 4 (50%) of whom had required ECLS
or drugs intoxication and were weaned off the emer-ency support (Fig 2). All patients were in CPC 1.While on ECLS support, 14 patients developed renal
ailure and 8 required hemofiltration. None of the dis-harged patients required out-of-hospital hemodialysis.
Five patients presented with early pulmonary hemor-hage and 1 patient experienced pulmonary sepsis.
Twenty-two patients (55%) were withdrawn fromCLS before 24 hours because of severe neurologic injury
n � 15), multiorgan failure (n � 4), or futility (n � 3).
omment
urrent resuscitation attempts from cardiac arrest yielduboptimal results, with survival varying from 1% to 5% forhe out-of-hospital cardiac arrests and from 17% to 25% forhe in-hospital arrest [1, 2]. Spontaneous circulation isestored in 50% of individuals who receive CPR, amonghom 10% to 30% acquire permanent brain damage [12,
3]. The feasibility of emergent cannulation and bypass
upport for emergencies depends on a collaborative effort p
ats.ctsnetjournDownloaded from
etween in-house surgical, cardiology, anesthesia, emer-ency room, nursing, and perfusion personnel [14].Early reestablishment of hemodynamics during refrac-
ory cardiac arrest may facilitate survival by avoidingschemic neurologic insult while providing essential car-iac and renal perfusion. During cardiac arrest, cerebralxygen storages are consumed and consciousness is lostithin 20 seconds; adenosine triphosphate and glucoseisappear in 5 minutes. Restored circulation is followedy global and multifocal cerebral hypoperfusion [13].omplex chemical disturbance accounts for the death ofulnerable neurons, and encephalopathy develops over aeriod of 3 days or longer.After cardiac function was restored, patients under-ent serial neurologic evaluations within 12 hours of
nstitution of ECLS that allowed neurologic viability to bestablished. In most patients, the decision to terminateupport or to proceed to long-term mechanical assist wasade within 24 hours.This protocol allowed reasonable initial mechanical cir-
ulatory support for all patients. The cost associated with aomplete disposable circuit (cannulas, tubing set, pumpead, and long-term oxygenator) is approximately $2,800S. The immediate implantation of a uni-VAD or bi-VADould have led to unnecessary costs for the patients whoould not have survived or been viable transplantation
andidates. In addition, straightforward VAD therapyould have induced prolonged critical ischemic time until
eperfusion, affecting further viability.
CLS Indication and Patient Selectionlinical experience with emergency CPB during pro-
onged cardiac arrest is limited, and survival rates varyetween 0% to 64% [5–11]; the range depends on themall number of patients and selection as well as thexperience of the team. However, the most importantactor for survival seems to be related to the delay untilnset of mechanical circulatory support.Debate in the literature is ongoing regarding ECLS
ndication versus CPR duration. In a multi-institutionalxperience, Hill and colleagues [5] reported that the timerom witnessed arrest to ECLS had moderate predictivealue for mortality and had not been a major consider-tion for withholding ECLS if evidence of neurologicunction persisted. This contrasts with the experience of
artz and colleagues [8], who suggested 30 minutes as aut-off for bypass initiation. In our experience, the aver-ge 71 minutes of external cardiac massage before thenset of CPB in surviving patients, Table 3 shows that theffectiveness rather than the duration of CPR has to beonsidered in the decision process.
The ability to correct the underlying pathology iseported as a determinant factor in the survival of pa-ients by many authors [9, 10], and it becomes obvioushat ECLS was either therapeutic in itself or useful forupporting the patient until appropriate palliative ororrective interventions could be carried out. Intoxicationith cardiotoxic drugs and severe accidental hypother-ia are two widely accepted indications for ECLS in
atients presenting with prolonged cardiac arrest [15–17],
182 MASSETTI ET AL Ann Thorac SurgEXTRACORPOREAL LIFE SUPPORT 2005;79:178–84
CA
RD
IOV
AS
CU
LA
R
hereas its usefulness in all other groups of patients isighly controversial. In our experience, 4 of 8 long-termurviving patients were treated for cardiotoxic drug in-oxication and were weaned without bridging to otherystems or to transplantation.
CLS Cannulationespite occasional reports on the transthoracic institu-
ion of emergency ECLS, the femoral approach is thereferred method of cannulation [18–20]. Access to theessels through the groin is rapid and easy to accomplishnywhere in the hospital while patient is being resusci-ated by closed chest massage. Nevertheless, cannula-ion-related complications, limb ischemia, and majorleeding, are frequent problems that contribute to theignificant morbidity associated with emergency ECLS.
agovern [4] reported that 46% of all patients withemorofemoral cannulation for ECLS needed surgicalepair. Schwarz [11] reported 24% of patients needed aurgical procedure either to achieve cannula placementr to correct ischemia or bleeding after successfulannulation.
All these experiences are related to direct percutaneousannulation in which flow obstruction by the cannula, orechanical vessel wall trauma during insertion can cause
ife-threatening limb ischemia accentuated by the intenseasoconstriction that accompanies the low cardiac output.leeding complications are related to the unsuccessfulttempts, especially in the heparinized patient. To over-ome these problems, we have adopted a modifiedeldinger technique: surgical exposure of the femoral ves-els and puncture of their anterior wall under direct vision.istal limb perfusion follows the cannulation procedure.espite this technique, 3 out of 18 surviving patientseeded reintervention for groin hemostasis.
CLS Managementhe neurologic state of patients was clinically assessed byaking attempts. In case of postanoxic coma, acuteeurologic deterioration, or both, controlled ventilationnd ECLS were discontinued when brain death was
able 3. Relationship Between Outcome andardiopulmonary Resuscitation Duration
ssessed. In other circumstances in which a prediction of m
ats.ctsnetjournDownloaded from
ermanent vegetative state was possible and complica-ions such as bleeding and multiorgan failure developed,CLS termination using “letting die” protocols was con-idered. Adjunctive predictive evaluations included se-ial EEG, evoked potentials, and transcranial Dopplervaluations, although in our experience they failed tohow a reliable correlation with neurologic outcome.
In recent years, hypothermia treatment using the sur-ace cooling method has been performed successfullyith the aim of brain protection in patients with postan-xic encephalopathy. The results of two prospectiveandomized trials that compared mild hypothermia withormothermia in comatose survivors of out-of-hospitalardiac arrest were published in 2002 [21, 22]. Based onhe published evidence to date, official guidelines recom-
end mild hypothermia as a therapeutic tool in thenconscious adult patients, with return to spontaneousirculation after out-of-hospital cardiac arrest. Such cool-ng may also be beneficial in ECLS patients sufferingrom postanoxic encephalopathy. Nagao and colleagues7] reported an interesting experience in 23 patientsdmitted after out-of-hospital cardiac arrest for whomhe protocol consisted of ECLS combined with the inser-ion of an intraaortic balloon pump and induction of mildystemic hypothermia. After this encouraging experi-nce, future research should evaluate the usefulness andafety of ECLS to provide both systemic cooling andemodynamic support in cardiac arrest patients.Despite the use of heparin-coated circuits and a low
ose of heparin, bleeding that needed transfusions re-ained a major problem. Three (17%) out of 18 survivors
equired massive transfusions. Although Yamashita andolleagues [23] identified bleeding and consecutive mul-iorgan failure as one main cause of mortality, 1 of the 3atients with massive bleedings was a long-term survivor
n our population.Peripheral ECLS configuration with right atrial-to-
emoral artery bypass can lead to insufficient leftentricular unloading, left ventricular distension, andlood stagnation. Insufficient unloading of the lefteart is therefore an important issue, especially foratients with asystole or ventricular fibrillation inhom this condition can sustain pulmonary conges-
ion and edema that hinders left ventricular recovery.onsequently, it is fundamental to restore left ventric-lar ejection with inotropic agents even if systemicerfusion is adequately maintained by ECLS. In oureries the filling pressures were monitored with awan-Ganz catheter or transesophageal echocardiog-aphy. Although all precautions were applied, inap-ropriate unloading of the left heart was a commonomplication that justified the early bridging to otherechanical cardiac assist systems or, as in the last
atient, a percutaneous balloon atrial septostomy. Be-ause of the spectacular pulmonary edema regressionnd on the basis of other experiences in the literature24], we strongly recommend this mechanical unload-ng through percutaneous balloon technologies.
The decision to wean from ECLS, to bridge to other
echanical systems, or to transplant depended on many
actors. The underlying pathology leading to a correctiventervention or to a rapid recovery justified this short-term
echanical support, and weaning was legitimate afterome days of support. Unknown etiology of the cardiomy-pathy or failure to attempt satisfactory unloading of the
eft heart with persistent pulmonary edema led to theecision to proceed with an early bridge to a ventricularssist device with paracorporeal ventricles or to an urgenteart transplantation if the recovery potential was esti-ated to be unreasonable.
onclusion
ur multidisciplinary continuous ECLS experienceuggests that patients with hemodynamic collapse thats due to cardiac arrest clearly benefit from this methodhat leads to rapid restoration of hemodynamic condi-ions. Furthermore, its easy setup, safe management,nd efficiency associated with its low cost comparedith other mechanical systems make ECLS an attrac-
ive and popular therapeutic technology. In addition,herapeutic hypothermia should improve survival inhose patients presenting with cerebral ischemic dam-ge before restoration of circulation.
he authors thank those people who collaborated in thisultidisciplinary approach to cardiac arrest: operating room
nd intensive care unit personnel, the perfusionist’s team,nd Dr Eric Pondaven of the Emergency Care Department.
eferences
1. Andreasson AC, Herlitz J, Bang A et al. Characteristics andoutcome among patients with a suspected in-hospital car-diac arrest. Resuscitation 1998;39:23–31.
2. Engdahl J, Holmberg M, Karlson BW, Luepker R, Herlitz J.The epidemiology of out-of-hospital cardiac arrest. Resusci-tation 2002;52:235–45.
3. Safar P, Abramson NS, Angelos M et al. Emergency cardio-pulmonary bypass for resuscitation from prolonged cardiacarrest. Am J Emerg Med 1990;8:55–67.
4. Magovern GJ Jr, Simpson KA. Extracorporeal membraneoxygenation for adult cardiac support: the Allegheny expe-rience. Ann Thor Surg 1999;68:655–61.
5. Hill J, Bruhn P, Cohen S. Emergent applications of cardio-pulmonary support: a multi-institutional experience. AnnThorac Surg 1992;54:699–704.
6. Dembitsky WP, Moreno-Cabral RJ, Adamson RM, Daily PO.Emergency resuscitation using portable extracorporealmembrane oxygenation. Ann Thorac Surg 1993;5:304–9.
7. Nagao K, Hayashi N, Kanmatsuse K et al. Cardiopulmonarycerebral resuscitation using emergency cardiopulmonary
ospital discharge. These results are substantially worse
tiepapi
2005 by The Society of Thoracic Surgeonsublished by Elsevier Inc
ats.ctsnetjournDownloaded from
bypass, coronary reperfusion therapy and mild hypothermiain patients with cardiac arrest outside the hospital. J Am CollCardiol 2000;36:776–83.
8. Hartz R, LoCicero J 3rd, Sanders JH Jr, Frederiksen JW, JoobAW, Michaelis LL. Clinical experience with portable cardio-pulmonary bypass in cardiac patients. Ann Thorac Surg1990;50:437–41.
9. Phillips SJ, Zeff RH, Kongtahworn C, et al. Percutaneouscardio pulmonary by-pass: application and indication foruse. Ann Thorac Surg 1989;47:121–3.
0. Chen YS, Chao A, Yu HY, et al. Analysis and results ofprolonged resuscitation in cardiac arrest patients rescued byextracorporeal membrane oxygenation. J Am Coll Cardiol2003;41:197–203.
1. Schwarz B, Mair P, Margreiter J et al. Experience with percu-taneous venoarterial cardiopulmonary bypass for emergencycirculatory support. Crit Care Med 2003;31:758–64.
2. Safar P. Resuscitation from clinical death: pathophysio-logic limits and therapeutic potentials. Crit Care Med1988;16:923– 41.
3. Siesjo BK, Siesjo P. Mechanisms of secondary brain injury.Eur J Anesthesiol 1996;13:247–68.
4. Martin GB, Rivers EP, Paradis NA, Goetting MG, Morris DC,Nowak RM. Emergency department cardiopulmonary by-pass in the treatment of human cardiac arrest. Chest 1988;113:743–51.
5. Banner W. Risks of extracorporeal membrane oxygenation:is there a role for use in the management of the acutelypoisoned patients? J Toxicol Clin Toxicol 1996;34(4):365–71.
6. Massetti M, Bruno P, Babatasi G, Neri E, Khayat A. Cardio-pulmonary bypass and severe drug intoxication. J ThoracCardiovasc Surg 2000;120:424–5.
7. Behringer W, Sterz F, Domanovits H. Percutaneous cardio-pulmonary bypass for therapy resistant cardiac arrest fromdigoxin overdose. Resuscitation 1998;37:47–50.
8. Read R, St Cyr J, Tornabene S, Whitman G. Improved cannu-lation method for extracorporeal membrane oxygenation. AnnThorac Surg 1990;50:670–1.
9. Greason KL, Hemp JR, Maxwell JM, Fetter JE, Moreno-Cabral RJ. Prevention of distal limb ischemia during cardio-pulmonary support via femoral cannulation. Ann ThoracSurg 1995;60:209–10.
0. Hendrickson SC, Glower DD. A method for perfusion of theleg during cardiopulmonary bypass via femoral cannulation.Ann Thorac Surg 1998;18:984–7.
1. Safar PJ, Kochanek PM. Therapeutic hypothermia after car-diac arrest. N Engl J Med 2002;346:612–3.
2. Bernard SA, Gray TW, Buist MD, et al. Treatment of coma-tose survivors of out-of-hospital cardiac arrest with inducedhypothermia N Engl J Med 2002;346:557–63.
3. Yamashita C, Ataka K, Azami T, Nakagiri K, Wakiyama H,Okada M. Usefulness of postoperative percutaneous cardio-pulmonary support using a centrifugal pump: Retrospectiveanalysis of complications. Artif Organs 1999;23:360–5.
4. Seib PM, Faulkner SC, Erickson CC, et al. Blade andballoon atrial septostomy for left heart decompression inpatients with severe ventricular dysfunction on extra
assetti and colleagues report the use of extracorporealife support (ECLS) for the treatment of adults who sufferardiac arrest refractory to conventional medical therapy.CLS had to be discontinued within 24 hours due torain death or multiorgan failure in the majority ofatients (55%) while only 8 patients (20%) survived to
han reports of survival rates exceeding 50% when ECLSs used as a resuscitation tool for children [1, 2]. Thextended duration of conventional resuscitation in theresent study (105 minutes vs 12 to 55 minutes in pedi-tric studies) emphasizes the need for a systematic ap-roach to emergency ECLS that addresses the following
ssues: organization of a “rapid resuscitation” ECLS
