Extracorporeal circulation: Practicalconsiderations for the anesthetistH. LYNN THOMAS, CRNA, BSWausau, Wisconsin

The author provides a comprehensiveoverview of the basic principles andpractical considerations involved withextracorporeal circulation, otherwiseknown as cardiopulmonary bypass. Theintent of the article is to afford theanesthetist a basic understanding of notonly the procedure itself, but its effectson the blood elements, oxygen transport,and tissue perfusion.

Extracorporeal circulation (ECC) or cardiopul-monary bypass (CPB) evolved over the years andhas contributed to great advances and a rapidgrowth in cardiac surgery. Keeping abreast of therapid advances in management and therapy ofpatients with cardiac disease has necessitated ateam concept. And, the anesthetist is a vital partof that team which provides a complex array ofservices of a pharmacologic, surgical, biochemical,anesthetic, and physical nature. A considerableamount of knowledge and judgment is expected ofthe anesthetist, ranging from an understanding ofthe physiology of the particular lesion, the surgicalprocedure, the physiology of CPB, to the normaland pathologic circulatory, respiratory and renalphysiology.

This article is designed to give a brief, yetspecific, overview of the basic elements involvedwith ECC so that the anesthetist will have a betterunderstanding of ECC. Not only is it essential for

the anesthetist to understand these particularmechanisms, but a knowledge of the effects ECCplaces on blood elements, oxygen transport, andtissue perfusion is of the utmost importance inproviding safe and optimal care to the patient.

HistoryThe earliest ideas and investigations into arti-

ficial oxygenation of the blood, both intra- andextra-corporeal, probably dates back to 1812 whenLegallois 1 wrote, "If one could substitute for theheart a kind of injection of artifical blood, eithernaturally or artificially made, one would succeedin maintaining alive indefinitely any part of thebody whatsoever." Nearly 50 years later in 1858,this became a near reality when the physiologistBrown-Seguard attempted, with moderate success,to perfuse the decapitated head of a dog. He wasone of the first researchers to emphasize the im-portance of cerebral blood flow and showed that afive-minute period of ischemia to the brain wassufficient to cause death. Nevertheless, these futileattempts where only the beginning of a long lineof intensive investigations to oxygenate bloodartificially.

It was not until the 1930's that other breakthroughs occurred. Gibbon 2 in 1937 gave an initialreport on experimental total cardiopulmonary by-pass using a mechanical heart and lung. Thismethod, however, was not perfected until the1950's. In 1951, Dennis8 and associates were thefirst to attempt the repair of an intracardiac defect

December/1981 565

in a human patient with the aid of a totallymechanical heart-lung apparatus. Though theirefforts failed, the knowledge that resulted ledGibbon 2 in 1953 to a successful repair of an atrialseptal defect using ECC for total CPB. Further in-vestigations have enabled ECC to become a pro-cedure which has achieved both a high degree ofprfection and which has given hope to manypatients requiring open heart surgery.

PumpOnce serious research began into ways of pro-

ducing CPB, it became evident that some type ofmechanical pump would have to be utilized fortissue perfusion purposes. Early thoughts and re-search were directed toward imitating nature, thatis, producing a pump with pulsatile flow whichwould give reasonably physiological pulse contours.Research 4 has demonstrated that flow should bepulsatile for optimal organ function and certainlyfor perfusion periods lasting six hours or longer.

In 1911, Hooker 5 designed a pump that wouldproduce just such a flow. However, the small size ofthe arterial inflow connectors found in the con-ventional heart-lung machine compared to thelarger size of the aorta with which it connected,dampened these physiological contours, renderingthem less physiological. Also, this pump was verycomplex in design, employing reciprocatingpositive displacement effected by a plastic dia-phragm and ball valves to control inflow and out-flow. Due to the mechanical action of such a pump,considerable damage to the blood and plasmaproteins occurred. Developmental efforts in thisarea later led to the adoption of the double rollerpump, as there was no convincing evidence thatserious disruption of organ function occurs fromcontinuous flow non-pulsatile pumps lasting lessthan six hours. 6

Roller pumps belong to the class of positivedisplacement pumps, that is, the roller progressesalong a resilient tube in order to "milk" blood outof it. Such pumps should have the capacity toprovide the equivalent of a basal cardiac outputunder the circumstances of perfusion, workingagainst a moderate pressure gradient. They shouldbe easily calibrated, controlled, and operable man-ually or by a self-contained power source if neces-sary. They should also provide a flow with minimaldevelopment of turbulence and shear forces as wellas minimal risk to blood integrity.

The double roller pump, which is most com-mon, has rollers at either end of an arm which ispierced in the center by a power drive shaft. Thispump utilizes a horseshoe-shaped casing and two

rollers revolving at the same axis, 1800 out ofphase. In this way when one roller finishes itsstroke, the other has already started and providesthe valving action. Backflow is impossible becausethe tube is always occluded by the roller at onepoint or another. These rollers can be adjusted toproduce significant negative pressure to aspirateblood from the surgical field and heart chambers.The average heart lung machine will have onepump for arterial inflow, one for venting theheart, and one or two for aspiration of free bloodfrom the surgical field.

There is evidence that hemolysis is quantita-tively related to the number of passages of theroller over the blood. The application of positivepressure to the cells being driven forward isgenerally low and well tolerated, therefore, theamount of trauma to the blood cells is relativelysmall. When the rollers are used to create a negativepressure for suctioning, this force may have themagnitude of several hundred millimeters. Forcesthis great will be associated with frothing, violentturbulence, acceleration and shear forces that mostcells will be unable to withstand. Therefore, therisk of hemolysis will depend more upon theamount and duration of suction used during aprocedure than on the rollers. 7

The tubing and interlinking connectors usedfor ECC by the roller pump should have lowthrombo-resistance and minimal effects on flow.Kaplan" gives a good outline of the characteristicsthat tubing should have. Normally, the arterialline diameter of the tubing used in a patientweighing above 40 kg is \. inch. However, re-sistance to flow and priming volumes will governthe diameter of the tubing used.

The structural integrity of pump tubing isthreatened to some degree by continuous use of aroller pu mp, and some fragmentation of the innersurface can occur. When this happens, plateletsand other blood components will be destroyedgiving rise to an increased chance of microemboli.The connectors linking the components of thesystem should be designed and finished so as tomaintain smooth laminar rather than turbulentflow. These connectors should also minimize thedevelopment of eddy currents, or cavitation causedby abrupt interior contour changes due to stenosisor ridges at tubing junctions. One problem of thistype is seen with the arterial connection.

An aortic arch cannula has a smooth, curvedright-angle bend leading to a narrowed segmentthat is inserted into the aorta. High pressures maybe required to deliver the pump output throughsuclh stenoses. This will cause marked flow ac-

Journal of the American Association of Nurse Anesthetists566

celeration of the stream of red cells and the conse-quent development of shear forces, convection andeddy currents, along with a fall in pressure. Thishas the potential for microbubble generation, redcell and platelet disruption, and may be a factorleading to the de-stabilization of the blood.9

Since it is evident that microemboli arepresent, filtration in the CPB circuit is essential.Very high levels of red cells, platelet and leukocyteaggregates have been demonstratd by Dopplercounting and filtration studies. 1" Also air, fat, sil-icone, and plastic particles have been found. Inone study the incidence of cerebral non-fat emboliwas reduced from 31.0 to 4.1% by filters. 11 Pre-sently, the Pall UltiporTM filter is the most com-monly used. It has resulted in less hematologicdisturbances and is associated with a continuingreduction in the risk of microemboli to all organs,though particularly the brain and lungs.

OxygenatorsThree types of oxygenators are currently

available for clinical use, and may be categorized interms of their interface with blood. In bubble anddisc oxygenators, oxygen and blood are in directcontact with each other. With membrane oxygen-ators, blood and oxygen are physically separated.

The rotating disc oxygenator was first de-veloped by Craford, Andersen and Bjork12 in1948, and later modified by Kay and Cross.13 Thisoxygenator consists of a horizontal glass cylindercontaining a row of flat steel discs mounted ver-tically on a horizontal steel rod and fitted betweentwo steel end plates which rotate about a centralshaft at a rate of 120 rpm. Venous blood entersthrough one endplate and fills the cylinder toabout one-third of its diameter. This then causesa thin film of blood to be forcefully spread overthe rotating disc. Oxygen, in ample supply, trans-fers to the blood in an amount dependent on thenumber of discs used, the area of the blood films,the rate of disc rotation, the 02 saturation of ven-ous blood, and the hemoglobin content of the redcells. With these conditions optimal, a siliconizedflat disc of 12.02 cm diameter effects a transfer of1.1 ml of 02 per minute.

In the early days of CPB, this oxygenator waswidely used and was the most popular oxygenator.However, this oxygenator was cumbersome toclean, pl)repare, and resterilize. Eventually, over-coming these disadvantages led to the developmentby DeWall"4 of an efficient bubble oxygenator.

The modern bubble oxygenator consists oftwo incompletely fused plastic sheets separated toform a channel for the inflow of unoxygenated

blood and the ventilating gas mixture. The venti-lating gases are passed through multiple perfora-tions in a diffusing plate giving rise to bubbleswhich exit into the venous blood at the bottom ofa columnal reservoir. This produces frothing andturbulence which assist in driving the now partlyarterialized blood forward into a debubblingchamber. Once here, coalescence of the foam andelimination of the last remaining bubbles areachieved by the action of antifoam surface activesubstances and by simple physical means such assettling, trapping, filtration, or vacuum.

Oxygenation in the bubbler is not membranelimited as in the normal lung, but depends uponthe gas-blood surface contact area, the thickness ofthe blood film, the mean red cell transit time, andthe partial pressure of O2 used. Normally, the gas-blood surface contact area is in the range of 15 m 2,and is governed by total gas flow and bubble size.It is well known that the smallest bubbles producethe greatest surface area, however, these smallbubbles are more stable and thus more difficult toremove. Therefore, bubbles 2-7 mm in diameterare used which result in an effectively thickerblood film. Since the time required for oxygenationis a function of the square of the film thickness, ared cell mean transit time of 1-2 sec is required.The result for most bubblers in current use is aflow of gas through the oxygenator column thatmust be at least equal to the flow of blood tomaintain both reasonable flow and gas exchange.

Since it is evident that current bubble oxygen-ators are not perfect, researchers have developedan oxygenator that more closely resembles thenatural lung-the membrane oxygenator.

As mentioned earlier, the membrane oxygen-ator operates with the blood and oxygen physicallyseparated. Its membranes are either solid or micro-porous. The membranes are arranged in flat,stacked, or coiled sheets, or in a tubular array.These membranes are constructed so that num-erous small capillary tubes in the device are com-pressed by external pressure, causing the blood inthe tubes to form a thin blood film layer over themembrane. Oxygen then flows between these layersof the membrane and diffuses into the blood. Thetype of material that composes tlhe membrane is ofutmost importance and usually consist of teflon,silastic or silicone. Silicone membranes have aCO2 :O., transfer capacity of 5:1 or 6:1, and thusmore closely resemble the situation in the lung.

Though there are advantages and disadvant-ages to each type of oxygenator, the membraneoxygenator is supl)erior when used for prolongedCPD support. However, since most procedures in-

December/1981 567

volving ECC have an average pump time of 60-90min, little benefits is derived from using the mem-brane oxygenator over the bubble oxygenator.

The fact that CO, is about 20 times more dif-fusible in body fluids than is 02 is well known.This same principle also applies to disc and bubbleoxygenators. In fact, with 02 flows sufficient toproduce acceptable arterial 02 tensions, the addi-tion of CO2 is required to avoid hypocarbia andmaintain normal arterial tensions. In many centers,3% CO2 and 97% 02 mixtures are used to main-tain the blood gases at more nearly normal levels.However, this approach may vary as CO solubilityvaries inversely and CO 2 output varies directlywith the temperature. As a consequence with cool-ing, more CO, will be needed for addition to themixture to maintain a normal PaCO2 ; with re-warming, less CO 2 will be needed.

The importance of precision in the control ofthe PaCO2 lies for the most part in its effects onthe metabolic state, the autonomic nervous system,catecholamine secretion, and thus on organ systemblood flow and function. Furthermore, alkalosisand hypothermia can cause a double leftward shiftin the oxyhemoglobin dissociation curve. Thisshift would then reduce the available 02 to ex-tremely low levels, thus making very little availablefor respiratory activity in tissue.

As for the membrane oxygenator, it is inter-esting to see that diffusion of CO 2 through mem-branes follows Graham's law, which is: "Therelative rapidities of diffusion of two gases varyinversely as the square roots of the densities."Therefore, most membrane oxygenators are unableto handle an increased CO. load, but are able tohandle CO 2 exchange without any problem whenPaCO2 is normal.

The heat exchanger is an essential componentof each bypass circuit. For many years, the Brown-Harrison and Sarns models were popular. As thesedevices are made of stainless steel, they are thusreusable; however, they do increase the primingvolume and may present problems related to thecleaning. They have sufficient resistance to flow tonecessitate their placement on the arterial sideand, with rewarming, pose the risk of bubble gen-eration caused by the reduction of gas solubilityin the blood being rewarmed.

Today, most modern oxygenators have incor-porated the heat exchanger into one of the cham-bers. This has the dual value of simplifying andreducing the internal volume of the circuit andreducing its priming volume. The risk of bubbleformation is still present and so a distal bubbletrap is essential. To enhance efficiency, the flows

of blood and water are usually arranged in acounter-current fashion. The temperature controland source of the water are the ordinary hot andcold water system, which can be augmented byactive water cooling if necessary.

In describing the oxygenators that are pre-sently used, we see that each involves man-mademechanical parts that are essential to the functionof the oxygenators. Undoubtedly, when bloodcomes into contact with these components, sometrauma to the blood and its constituents is boundto occur.

In actual use, oxygenators lead to a numberof undesirable and limiting functional problemssuch as the destruction of erythrocytes, leukocytes,and platelets,15 formation of microemboli, de-naturation of plasma proteins, coagulation defects,hyperoxia, hypoxia, hypocapnia, the need for large-volume pump primes, sterilizing difficulties andlastly, the inability to maintain the bypass safelyfor long surgical procedures.

When blood comes in contact with the plasticand the gas interface, it gives rise to protein de-naturation and the adhesion and agglutination ofplatelets. The denatured protein layer then at-tracts platelets, which also become adherent.Eventually, this becomes self-limiting. Once theplastic becomes coated, further extension of theprocess is largely halted. In contrast, the reactionsat the air-blood interface, which also alter thecoagulation proteins and consume platelets, con-tinue as long as the oxygenator is in use. Also, thedisc and bubble oxygenators tend to form largeamounts of free hemoglobin in the plasma due tored cell hemolysis, especially after more than twohours of bypass. 6

As stated earlier, the membrane of the mem-brane oxygenator contains a multitude of smallpores which, on initial contact with blood, inducesa mild degree of protein denaturation and plateletagglutination. This appears to combine with theprotein lying immediately above it to give arelatively fixed and non-moving layer that is es-sentially cell free. The protein layer, once formed,stabilizes the membrane reducing its subsequentreactivity with respect to blood. Therefore, lesstrauma to the blood is seen with the membraneoxygenator than with a disc or bubble oxygenator.

Currently, research is being conducted intothe use of anti-platelet drugs that can be added tooxygenators to decrease the loss of platelets in thesesystems. Addonizio 17 and associates have beenstudying thle effects of Prostaglandin E1 on plate-lets in various oxygenator systems. Their researchhas shown that the addition of this hormone to

Journal of the American Association of Nurse Anesthetists568

membrane oxygenators will decrease platelet loss.With bubble oxygenators, Prostaglandin E1 hasagain been shown to decrease platelet loss, how-ever, it must be used in a four-fold concentrationand at this strength Prostaglandin E1 becomes apotent vasodilator.

BypassIn total bypass, ties are simultaneously placed

tightly around the superior and inferior venae cavaand the caval catheters previously inserted. Thearterial pulsatile pressure tracing will disappearand become a mean perfusion pressure tracing,changes in the level which reflect the pressurefluctuations of the pump heads.

Although 5% of the total cardiac output sup-plies the coronary artieries, the majority of coron-ary blood flow returns to the right ventricle viathe coronary sinus. A small amount of blood fromthe bronchial vessels and thebesian veins contributeto the left ventricular venous return. Therefore, asump drain is needed to empty the left ventricleand prevent over-distention during bypass.8

When total bypass is started, many modalitiesmust be observed. One is the arterial blood pres-sure, usually monitored intraarterially by a pres-sure transducer. If venous return is inadvertentlystarted before arterial infusion, loss of blood fromthe patient can cause severe hypotension. If thearterial side is allowed to pump too long, severehypertension, circulatory overload, and evencardiac failure will occur. Most perfusionistsusually give the patient 100-200 cc of bloodthrough the arterial side before starting venousoutflow. If the blood pressure should fall pre-cipitously while the patient is going on bypass anddoes not respond to increased flow, the possibilityof arterial dissection must be considered.

Adequate venous return must also be watched.If a catheter is placed too far into the superior venacava, either the internal jugular vein or the sub-clavian vein could be occluded. When this occurs,a precipitous increase in the central venous pres-sure may occur and immediate corrective action isneeded. Obstruction of vena caval catheters willresult in a marked decrease in venous return anda sharp drop in the reservoir level. Obstruction ofthe jugular vein may produce a severe decrease inthe effective cerebral perfusion pressure. Sincevenous return is usually determined by gravity,the operating table may need to be raised to in-crease gravity flow if the return is inadequate.

As for partial bypass, the name is indicativeof the procedure, because some of the blood is stillpumped by the heart. More specifically, some

blood is pumped by both right and left ventriclescreating pulsatile flow. This blood flows throughthe lungs, back into the left heart, and out throughthe aorta. The rest of the circulating blood iscollected from the vena cava, sent through thepump and then back into the patient. Partial by-pass may be used to provide partial support in theimmediate pre- and post-operative period in whichlow cardiac output states may develop.

Femoral-femoral bypass is the conventionalform of partial bypass, though recently left atrial-to-femoral arterial bypass has become popular.These methods are most commonly used for repairof aneurysms of the descending thoracic aorta andare seldom seen with procedures involving openheart or coronary artery bypass grafts.

HemodilutionNow that we have described the pump and

the various oxygenators used in ECC, we mustlook closely at the circulatory changes broughtabout by hemodilution and the pump primes. Itis essential that the anesthetist understand theseconcepts and the effects they have on the patient.

Rheology of blood refers to the flow propertiesof blood. For most clinical purposes, only tworheological parameters of blood need to be con-sidered: (1) the viscosity function and (2) theyield stress. The viscosity function is a measure ofthe resistance of blood to flow and is defined asthe ratio of shear stress, or force per area thatcauses flow, to the velocity gradient, or shear rate.Practically speaking, the viscosity of blood is pro-portional to the pressure gradient divided by thespeed of flow. For simple liquids such as water,the pressure gradient is a constant. This theory isknown as Newtonian behavior. With blood, therelationship between pressure and flow is non-linear and the viscosity depends on shear rate:this is known as non-Newtonian behavior.

The yield stress of blood is a measure of theminimal force required to cause blood to flow.Below a certain force, or pressure, the attractionbetween adjacent red cells is too strong to beovercome, hence, the blood does not "yield." Assoon as the yield stress is exceeded, flow begins;but the large aggregates of red blood cells at thislow level of stress highly resist flow and thus havea high viscosity. As the pressure gradient and thespeed increase, the aggregates disperse and bloodviscosity decreases. At high shear rates, the de-formability of the cells becomes important. Factorswhich increase the tendency of red blood cells toaggregate affect the yield stress and the low shearrate properties, while factors which increase the

December/1981 569

rigidity of red blood cells affect flow at high shearrates.

A variety of factors increases viscosity and yieldstress, but the more important ones are hematocrit,the concentration of plasma proteins, and temper-ature.

The viscosity of blood highly depends onhematocrit (Hct) for two reasons: (1) the yieldstress or the tendency of blood to aggregate in-creases substantially with Hct and (2) the viscosityincreases rapidly when Hct is above 40%. If onelooks at the clinical application of these theories,we also see that other parameters come into view.The mean arterial pressure (MAP) is determined bythe product of cardiac output and total peripheralresistance (TPR).

Total peripheral resistance in turn, dependson two factors: (1) the local caliber of thearterioles, venules, and capillaries making up thebulk of the resistance to flow and (2) the viscosityof blood. Studies have shown that the TPR isdirectly proportional to the high shear rate bloodviscosity. This implies that marked changes inhematocrit during ECC at a constant flow may sub-stantially decrease MAP secondary to decreasedviscosity and TPR. Also, acute isovolemic hemodi-lution during anesthesia can occur with a dec-rement in the total transport of 02. This isbecause at a constant MAP, cardiac output is in-versely proportional to the viscosity of blood.Studies by Crowell, Gordon, LaVeen, and Mal-berg18-21 show that an Hct of 25-30% leads to theoptimal combination of flow and 02 carryingcapacity of blood.

If we further investigate viscosity, we see thatit is also affected by changes in the concentrationof plasma proteins. Of all the plasma proteins,only fibrinogen and the gamma globulins signifi-cantly influence viscosity (1) because the yieldstress of blood depends on fibrinogen concentra-tion, and (2) because both fibrinogen and thegamma globulins have the greatest effect on plasmaviscosity.22

With hemodilution, the level of concentrationof serum proteins falls, thus reducing viscosity.This fall, however, is not without other effects.Unless there is some substitute for its oncoticpressure, the transcapillary shift of water to tissuewill be increased in all organ systems. 28 Part or allof this extracellular fluid (ECF) increase may benonfunctional and represents a third space. Excessfluid shifts to the tissue will be made evident bythe need to continuously add fluids to the CPBcircuit in order to maintain the same blood levelin the oxygenator.

The final factor affecting viscosity is temper-ature. Temperature affects the viscosity of bloodprimarily by influencing plasma viscosity. Gener-ally as a rule, viscosity increases 5% for each 1o Cdrop in temperature. However, with hemodilutionand a decrease in TPR, this change in viscosity isusually insignificant.

Many solutions have been used to prime theoxygenator, which normally has a volume rangingbelow 1.5 and 2.2 L. In the early days of ECC,blood was thought to be the most physiological fluidto prime the oxygenator, but the poor results anddifficulties with all blood primes soon became ap-parent. 24 Later, Zudhi 25 did extensive work withpump primes and was a strong advocate of D,W.

The basic principle governing the manage-ment of the connection of a patient to the CPBcircuit is to do so in a fashion requiring minimalmetabolic adaptation by the patient. This meansthat the pump prime should be as close as possibleto the composition of the patient's extracellularfluid. For most purposes, this has been interpretedto mean the use of a balanced salt solution with orwithout supplemental Ca++ and Mg++. Calciumhas been omitted by many groups in hope of re-ducing 02 demand in the myocardium and the O.2debt associated with ischemia.

We have already stated that a shift of fluidto a "third" space occurs when we hemodilute theblood volume and lower oncotic pressure. It mustbe assumed that due to the stress of anesthesia andsurgery, increased levels of glucocorticoids andmineralocorticoids will be seen. Despite isotonicexpansion of the blood and total ECF at the endof bypass and the maintenance of adequate urinaryoutput, the net effect of the development of a thirdspace and the activation of these mechanisms is toproduce moderate degrees of sodium and waterretention.

The changes, primarily involving aldosterone,also tend to significantly increase the perioperativerate of potassium loss and may contribute to lowserum K+ levels. Other factors that may contributeto this situation are the use of diuretics, hyper-ventilation, and ECF expansion with fluids isotonicwith respect to sodium at a time in which sodiumretaining mechanisms are active. This lattermechanism might be expected to lead to increasedrates of intrarenal Na+ for K+ exchange and thusincreased K+ loss. Therefore, careful monitoringof K+ levels is essential.

Body compositional changes entailing reduc-tions in Ca++, Mg++, P04=, and Zn may also beseen. In the absence of Ca++ and Mg++ in theprime, total values will decrease with hemodilution,

Journal of the American Association of Nurse Anesthetists570

paralleling the decrease in serum proteins. Hypo-calcemia is usually seen during bypass but seemsrelatively insignificant. Nevertheless, occasional ad-ministration of Ca++ may be a useful adjunct inthe management of the hypotension often seen atthe end of bypass.

The role of Mg++ is not clear. In the clinicalsetting, reductions in the total Mg++ and theionized faction occur with hemodilution. This de-ficiency may make the heart more prone to ar-rhythmias, especially those associated with the useof digitalis.

Metabolic oxygen demandWe have now looked at many components of

ECC, but none are more important than the abilityto provide adequate tissue perfusion and oxygen-ation. The blood flow requirements for patientsduring CPB that any oxygenator should be able tosatisfy may be described in terms of milliliters perkilogram of body weight or milliliters per squaremeter of body surface area. If one examines the02 delivery equations , it indicates that 02 deliveryis a function of cardiac output and 02 carryingcapacity and that there is a certain basal require-ment of O that must be satisfied. Normally thisis in the range of 300 cc of O per min. With redblood cell dilution and hematocrit reduction,lower 02 carrying and delivery capacity results.Therefore, cardiac output must be increased or the02 demand reduced by hypothermia.

At the onset of bypass, with the patient nor-mothermic, 02 carrying capacity will be low, butmay be compensated for by a rightward shift inthe oxyhemoglobin dissociation curve resulting inincreased peripheral O2 extraction. In addition,the metabolic acidosis resulting from the 02 debtwill lead to a reduction in red cell On affinity.However, reduction in 0O. demand by hypothermiais the essential factor used to deal with this prob-lem. Once hypothermia is induced and the patienthemodiluted, the decreased capacity of blood tocarry 02 will be more than offset by the increasedflow resulting from a reduced TPR.

In the early days of ECC, oxygenators wereinadequate and a low flow of blood was used. To-day's modern oxygenators are capable of adding300 cc and greater, of O per min to the blood flow,thus diminishing the need for lower flows. As wepointed out earlier, certain basal metabolic de-mands are required, which means certain flow ratesmust be met. In the average adult, flow rate rangesfrom 3.2-3.6 L/m/M 2. Once hypothermia and he-modilution have been initiated, the metabolic 02demands can be met with flows in the range of

2.2-2.5 L/m/M 2 or 50-80 ml/kg/min. If flowsgreater than this occur in the oxygenator, thechances of increased red blood cell destruction in-variably rises.

Originally, hypothermia was induced by sur-face cooling. It was soon learned that this tediousprocess cooled the vital organs less effectively thanfat or muscle. Today internal cooling using theheat exchanger is much more effective both coolingand rewarming vital organs more readily than fatand muscle.

The cooling-rewarming process is not a steadystate and is non-uniform in terms of its temper-ature distribution from one organ system to theother in the body. Therefore, supplemental surfaceheating and cooling using a K-thermia blanket isoften used. Because of the difference in timecourse, organ blood flow, and resulting organ tem-peratures, core or average temperatures can onlybe estimated.

The fall in 02 consumption with reduction intemperature is not a linear process and is difficultto assess. Oxygen consumption normally will bereduced to 50% at 30°C, to 25% at 25°C, to 15%at 20°C, and to 10% at 15°C. It is also importantto realize that as temperature decreases, the dis-solved O, content of plasma increases. The amountof dissolved O, is approximately 0.3 volume % at38°C, 1.5 volume % at 30°C, 2 volume % at 25°C,2.5 volume % at 20°C, and over 4 volume %at O°C. As stated earlier, the normothermic, he-modiluted patient will have a low 02 carryingcapacity. After rewarming, if a low cardiac outputstate develops, this induced 0, capacity could becritical. Therefore, alkalosis and decreases in theHgb concentration should be avoided. If necessary,packed red cells can be infused.

Hypothermia, in addition to reducing 02 con-sumption, may have other effects. Like hemodilu-tion, it enhances the stability of blood and serumlipids, and reduces the risk of subsequent hemato-logical disruption after bypass. By constriction ofthe body's capacitance vessels, hypothermia alsoleads to complex changes in the size and distribu-tio()n of thle blood volume. The blood volume isusually redutced in a non-uniform fashion that isundoubl)tedly related to the temperature and thereduced On demand of each particular organ sys-tem.

When rewarming the patient, care must betaken that the temperature of the infused blooddoes not exceed 40°C, because gross hemolysiswith subsequent development of a bleeding di-athesis may occur. Care should also be given to thepatient shivering in the p)ostoperative period. The

December/1981 571

increase in total body O0 consumption that mayresult varies between 100 and 400% of normal.This is significant in that it must be matched bycorresponding increases in cardiac output in orderto maintain 02 delivery at the normal rate. In thepresence of significant hemodilution, the increasedperipheral 02 extraction may result in markedvenous desaturation which may contribute to sys-temic arterial hypoxemia.19 Also, with cardiac out-put (CO) increased, the possibility of a shortenedpulmonary red blood cell mean transit time exists.Together, these two defects could easily lead to amarked reduction in arterial 02 tension.

Myocardial preservationProtection of the myocardium during ECC de-

pends upon the extent to which it is possible tocontrol surgical trauma and the effects of ischemia.This is of the utmost importance to the anesthetistbecause it is our responsibility to see that the heartfunctions as well, if not better, at the end of bypass.

Cardiac surgical procedures often require abloodless, relaxed and motionless field during theoperation, which is easily accomplished by ischemicarrest induced by aortic cross-clamping. Any periodof ischemia is accomplished by metabolic andstructural changes which determine the functionalrecovery of the heart in the postoperative period.

The safe period of ischemia is not well de-fined, but 20-30 min is generally considered theupper limit. If this time period is extended, sub-stantial subendocardial necrosis may occur withlow output syndrome in the postoperative period.The need for protection of the myocardium duringischemic arrest has been well recognized and anumber of methods have been employed.26 Ofthese, hypothermia and pharmacological arrestwith cold cardioplegia solutions have now gainedwide acceptance in clinical practice.

Hypothermia and surface cooling of the heartwas first used by Shamway.27 It provides a blood-less, arrested heart, and lower energy require-ments; delays the depletion of high-energy phos-phate reserves, and lactic acid accumulation; andretards the morphological and functional dete-rioration associated with ischemic arrest. Moderatehypothermia and surface cooling have been suf-ficient for 20 min, but profound cooling of themyocardium is known to cause myocardial damagedue to crystallization of the membrane lipids andpoor ventricular performance on perfusion. 28

The concept of pharmacological arrest wasinitially introduced by Hooker 29 in 1929, with ex-tensive research by Melrose s30 in 1955. These initialsolutions contained high concentrations of K+ and

were normothermic. Since that time, a number ofother constituents have been added and the solu-tions cooled. Cold cardioplegia solutions are def-initely advantageous and protect the myocardiumby virture of their cooling effect and added com-ponents. When injected into the aortic root aftercross clamping, the cold cardioplegic solutions aredistributed throughout the myocardium by naturalpathways, thus producing homogenous cooling ofthe myocardium. The myocardial energy require-ments, 02 debt, and accumulation of metaboliteswhich inhibit the anaerobioses are prevented. Thesubendocardium is well protected and the hemo-dynamic instability as a result of poor myocardialcontractility and reduced compliance associatedwith ischemic arrest is considerably reduced.

In recent years, a number of constituents havebeen added in various concentrations to differentcardioplegic solutions, and an understanding oftheir mechanisms of action is essential.

Potassium has been known to arrest the heartin diastole for many years. 31 Potassium causes ionicparalysis of the cellular membrane transport systemby blocking the initial fast phase of myocardialcellular depolarization and thus preserves theenergy reserves for the maintenance of integrity ofcell membranes and active transport of Na-Kpump. Initially, the Melrose solution containedover 200 mEq of K+/L with a tonicity greaterthan 400 mOsm/L that produced many side effectsand proved to be unbeneficial. Today, K+ in aconcentration of 10-30 mEq/L has been consideredoptimal for good postoperative recovery.

Magnesium and procaine are also effective car-dioplegic agents and augment the arrest inducedby K+. Magnesium blocks the initial depolariza-tion by its effect on transmembrane ion move-ments and also has the benefit of blocking in-tracellular metabolism. Magnesium aspartate hasbeen considered more advantageous as it acceler-ates the regeneration of adenosine triphosphate(ATP) from inosine phosphate. Procaine hydro-chloride, a local anesthetic, binds to the proteinsin the membrane itself and reduces transmem-brane ion movement.

The cellular metabolic process (such as themembrane pump) are pH dependent, therefore,slight alkalosis would appear beneficial. Bicarbon-ate, inorganic phosphates, and tromethamine(THAM) have been used as buffering agents tominimize the ischemic injury and to raise the pHbetween 7.4 and 7.8. Energy substrate support maybe conferred by the inclusion of glucose and in-sulin. Similarly, a slight increase in osmolarityappears to be useful in minimizing intracellular

Journal of the American Association of Nurse Anesthetists572

water accumulation and has led to the use of sor-bitol, mannitol or plasma proteins to reduceedema.

Now that we have identified some of the con-stituents of cold cardioplegic solutions and under-stand how they work, we see that multidose ad-ministration of cold cardioplegia with moderatehypothermia and surface cooling has been foundmost satisfactory for prolonged aortic cross clamp-ing and has undoubtedly improved the prognosisof a number of patients undergoing surgical cor-rection of complex cardiac lesions.

Heparin and protamineAn inability to prevent clotting of the blood

was one of the principal causes of failure in theearly experiments with ECC. Defibrinogenatedblood was later used, but also proved to be afailure. In 1916, a young medical student namedJay McClean discovered a mucopolysaccharide ex-tracted from the liver of a dog that preventedcoagulation. Heparin was investigated over thenext several decades and was introduced clinicallyin the 1940's.

Heparin is the strongest organic acid foundin the body; it is most commonly prepared frombovine lung and porcine or bovine intestinalmucosa. Heparin, in the presence of its co-factors,leads to the almost instantaneous production of athrombin-antithrombin complex deactivating thethrombin. This heparin-antithrombin co-factorcomplex also appears to inactivate factors IX, X,XI, and XII.8

Once administered, heparin becomes proteinbound and is distributed in the plasma. It has ahalf-life of 60-100 min in normothermic man butmay be extended with increasing dosage and de-creasing temperature. Its elimination pattern isunclear, but may involve biodegradation in thereticuloendothelial system.

After entering the circulation, the anticoag-ulant response will vary depending upon such fac-tors as temperature, age, body composition, andmost particularly the lean body mass. For totalheparinization, the usual dose is 2-3 mg/kg or 200-300 units/kg. To assess precisely the initial effectand rate of decline in each patient, a number oftests have been suggested.8 One method that is easyand convenient to use in the operating room is theCelite® activated clotting time (ACT) described byHattersley.32

A baseline value is determined prior to hepar-inization and repeated just after giving heparinand at 30-min intervals. With the dose of heparinin mg/kg on the vertical axis, a dose response

curve can be constructed. Bull and associates33 haveadvocated that heparin be given initially in a doseof 2 mg/kg, an ACT determined, and the line ex-trapolated to give a value between 400-480 sec.They suggest using the line drawn from this pointthrough the original value to calculate the remain-ing dose of heparin required. The remaining dos-ages should maintain values of at least twice controland not less than 350 sec.

The use of protamine as a heparin antagonistwas suggested in 1937. It has been found that pro-tamine binds almost mg/kg with heparin. In theabsence of heparin, protamine has anticoagulationproperties of its own. Probably this is a result of itscapacity to activate the coagulation mechanism andproduce a consumption coagulopathy involvingplatelets.

To calculate the dosage of protamine re-quired, residual heparin activity should bemeasured. The ACT test, as noted, can be usedto estimate the biological half-life of heparin andthe amount remaining. It can then serve as a guidefor the amount of protamine required for heparinreversal. Normally, this is about 2-3 mg/kg ofprotamine. When protamine is administered, careshould be taken to inject the dose slowly and overseveral minutes. Protamine given as a bolus hasbeen associated with direct myocardial depressionfollowed by hypotensive episodes. The exact etio-logic mechanisms for its hypotensive effects are notentirely clear, but it is thought that protamine willact directly on the smooth muscle of the vessels andcause vasodilatation.

MonitoringAs with all major procedures, adequate mon-

itoring is essential. It is the responsibility of theanesthetist to understand and precisely interpretthe results attained from all measurements and actaccordingly-in the least amount of time.

During bypass, blood gases are frequentlymeasured using an intra-arterial catheter, to checkPaOs, PaCO2 , pH, 02 saturation, C0 2 , and baseexcess. These parameters are monitored usuallyevery 30 min, unless circumstances suggest theneed for more frequent determinations. Increasingmetabolic acidosis usually suggests inadequatetissue perfusion, and must be defined and correctedimmediately. Central venous PO s's are also mea-sured and should be kept in the range of 40 torr.If metabolic acidosis persists, sodium bicarbonatecan be administered using the Astrup equation.This is calculated by the following: taking bodyweight in kilograms x base deficit x 0.3 to equalmilliequivalents of bicarbonate to be administered.

December/1981 573

Hematocrit and electrolytes should also be mea-sured with each blood gas sample taken. Normally,while on bypass, the hematocrit should be keptbetween 25-30% while the K+ level should bekept around 4.5 mEq/L.

In recent years, pulmonary artery pressure(PAP) and pulmonary artery wedge pressure(PAWP) have been monitored by the use of aSwan-GanzTM catheter. The PAP and PAWP shouldall be low or zero while on bypass. If these indicesbecome elevated, it means the left ventricle isbecoming distended as a result of inadequateventing.

The arterial pressure needs to be watched veryclosely to maintain a mean pressure of 50-100 torr.If the mean pressure falls below 50 torr, bloodvolume is increased either by adding fluid to thepump or directly to the patient. If a reasonablevolume of fluid is given with no response, then analpha agonist such as phenylephrine may be in-dicated. If the mean blood pressure increases to100 torr or greater, a vasodilator such as sodiumnitroprusside or one of the inhalational agents maybe administered.

During ECC, a urinary output of 1 cc/kg/hris desirable. When urinary output is inadequate,renal hypoperfusion may be the cause. Since therenal threshold for hemoglobin is approximately100-150 mg/100 ml, red urine is a sign of massivehemolysis. Metabolic acidosis and a low urinaryoutput combined create an ideal condition for thedevelopment of acute tubular necrosis. Therefore,urine output should be increased with a diureticand alkalization of the urine.

As with all surgical procedures, the ECGshould be monitored continuously. Specific atten-tion should be placed on the S-T segment for thedevelopment of ischemia or conduction distur-bances. It is generally accepted that a modified V-5lead will show changes of this type best. Onceasystole has occurred by the use of a cold cardio-plegic solution, the anesthetist must watch for thereturn of ventricular fibrillation. Once this isnoticed, the surgeon should be notified so thatmore cardioplegic solution can be used.

Pulmonary complicationsTheoretically, during total bypass the only

blood entering the lungs is 0.5-1.0% of the pumpoutput that enters the bronchial arteries. At highflow rates, bronchial arterial flow may be normalor even increased. Consequently, pulmonary venouscongestion may develop. While the patient is ontotal bypass, the lungs are often held in theirdeflated state, which predisposes to a progressive loss

of surfactant and decreased distensibility. Thesetend to return toward normal when normal circu-lation is restored.

Hypoxia of lung tissue itself, due to the ab-sence of ventilation, may also contribute to alveolardamage. To avoid these complications, somegroups recommend maintaining the lungs in a con-stant state of inflation with 02, C0 2 , or helium.Others have shown that pulmonary complicationsare decreased when the lungs are intermittentlyventilated throughout the course of bypass, orwhen positive end-expiratory pressure is used.

Several additional factors related to ECC cancontribute to pulmonary dysfunction. Fluid over-load, unrecognized during bypass, may becomeevident almost immediately after the patientcomes off the pump. Left ventricular failure withlow cardiac output secondary to poor myocardialfunction is a classic cause of lung impairment.High 0, concentrations for prolonged periods inan already functionally or anatomically compro-mised lung will accelerate the onset of pulmonarydamage characteristic of 02 toxicity. Vasoactivesubstances released by various tissues or damagedcells during bypass may induce pulmonary vaso-constriction sufficient to produce ischemic damage.

Methylprednisolone sodium succinate has beenrecommended for both therapeutic 85 and prophy-lactics" use for pulmonary changes following ECC.However, the still unsettled questions of effec-tiveness and high cost make its routine use de-batable.37

Renal effects of ECCAs stated earlier, a pulsatile flow to the kidneys

is of great benefit since perfusion is better and therenin-angiotension activity and aldosterone secre-tion are reduced. However, long-term pulsatileflow appears to be clearly superior for the preser-vation of organ function.

Acute renal failure is a complication of opera-tions for the correction of cardiac abnormalities,especially when such correction involves the use ofECC. The mortality due to this acute disease pro-cess in postoperative cardiac patients approaches75%. 38 Acute tubular necrosis may occur as a directclinical entity in a patient who is otherwise en-joying a satisfactory postoperative course or it mayoccur as part of a generalized organ failure in apatient who is approaching a terminal state. Ineither case, acute tubular necrosis is a diseaseprocess of which the anesthetist must be aware,with particular attention given to its prevention.

Patients undergoing cardiac surgery are sub-jected to a period during which decreased renal

Journal of the American Association of Nurse Anesthetists574

perfusion is likely to occur. Many of these patientshave low cardiac outputs which will result in lowrenal perfusion. In addition, if the cardiac diseaseis the result of a generalized vascular disorder suchas atherosclerosis, then the same disease processwhich produced the cardiac abnormality is likelyto also involve the kidneys. Whether the onset ofrenal ischemia is triggered by the pre-existing lowflow state or whether it results from a further de-crease in perfusion in association with ECC, thenet effect appears to be intense vasoconstriction inthe superficial cortex with a reduction in glom-erular filtration.39

A second factor which is associated with acutetubular necrosis is the presence of free hemoglobinin the urine. Earlier pumps and oxygenators wereassociated with a high degree of red cell hemolysisthat resulted in the release of free hemoglobin intothe circulation. Initially this hemoglobin will beprotein bound, largely due to haptoglobin, andwill subsequently be removed by the reticuloendo-thelial system. As the binding sites become oc-cupied, more hemoglobin will remain free in thecirculation. Above a threshold of 100-150 mg/100ml, hemoglobin will be excreted by the kidneywith little trouble. However, with low rates ofrenal tubular flow and aciduria, significantamounts of hemoglobin may be converted intoacid hematin crystals resulting in acute tubularnecrosis.

SummaryThe anesthetist involved with cardiac pro-

cedures which require ECC must have a vast arrayof knowledge to be properly prepared to deal withany complications that may arise in the care of thepatient. This article has focused only on the basicprinciples and practical considerations involvingECC. For the anesthetist who will make cardiac an-esthesia his or her specialty, a more in-depth under-standing that is beyond the scope of this articlewould be required.

REFERENCES

(1) LeGallois, JJC. 1812. Experiences sur le principe de la vie,notammelat sur celui des mouvements du coeur, et sur le siegede ce principe, Paris.(2) Gibbon, JH Jr. 1937. Artificial maintenance of circulationduring experimental occlusion of pulmonary artery, Arch. Surg.34:1105.(3) Dennis, C, Karlson, KE, Eder, WP, et al. 1951. Pump

oxygenator to supplant the heart and lungs for brief periods;A method applicable to dogs, Surgery 29:697.(4) Milnor, WR. 1972. Pulsatile blood flow, N. Eng. J. Med.

287:27.(5) Hooker, DR. 1910. A study of the isolated kidney-the in-fluence of pulse pressure upon renal function, Am. J. Physiol.27:27.

(6) Goodyear, AVN, Glenn, WWL. 1951. Relation of arterialpulse pressure to renal function, Am. J. Physiol. 167:689.(7) Okies, JE, Goodnight, SH, Litchford, B, et al. 1977. Effects

of infusion of cardiotomy suction blood during extracorporealcirculation for coronary bypass surgery, J. Thorac. Cardiovasc.Surg. 74:440.(8) Kaplan, JA. 1979. Cardiac Anesthesia, 2nd edition, Grune& Stratton, pp. 393-441.(9) Bass, RM, Longmore, DB. 1969. Cerebral damage duringopen heart surgery, Nature 222:30.(10) Dutton, RC, Edmunds, LH. 1973. Measurement of emboliin extracorporeal perfusion systems, J. Thorac. Cardiovasc. Surg.65:523.(11) Osborn, JJ, Swank, RL. 1970. Clinical use of a Dacron®wool filter during perfusion for open heart surgery, J. Thorac.Cardiovasc. Surg. 60:575-581.(12) Bjork, VO. 1948. Brain perfusion in dogs with artificially

oxygenated blood, Acta Chir. Scand. 96:1.(13) Cross, FS, Berne, RM, Hirose, Y, et al. 1956. Evaluation of

a rotating disc oxygenator, Proc. Soc. Exp. Biol. Med., 93:210.(14) DeWall, RA, Warden, HE, Read, RC, et al. 1956. A sim-ple expendable artificial oxygenation for open heart surgery,Surg. Clin. North Am., 36:1025.(15) Kvarsten, B, Cappelen, C, Jr, Ostreud, A. 1974. Bloodplatelets and leukocytes during cardiopulmonary bypass, Scand.J. Thorac. Cardiovasc. Surg., 8:142-145.(16) Anderson, MN, Kerchika, K. 1969. Blood trauma pro-duced by pump oxygenators: Comparison of 5 different units,J. Thorac. Cardiovasc. Surg. 57:238-244.(17) Addonizio, VP, et al. 1979. Effects of prostaglandin E, onplatelet loss during in vivo and in vitro extracorporeal circula-tion with a bubble oxygenator, 1. Thorac. Cardiovasc. Surg.77:119-126.(18) Crowell, JW, Smith, EE. 1967. Determinant of optimalhematocrit, J. Appl. Physiol. 22:501-504.(19) Gordon, RJ, Snyder, GK, Tritel, H, et al. 1974. Potential

significance of plasma viscosity and hematocrit variations inmyocardial ischemia, Am. Heart J. 87:175-182.(20) I.aVeen, HH, Moon, I, Nafis, A. 1980. Lowering bloodviscosity to overcome vascular resistance, Surg. Gynecol. Obstet.150:139-149.(21) Malmberg, PO, Woodson, RD. 1979. Effect of anemia onoxygen transport in hemorrhagic shock, J. Appl. Physiol. 47:882-888.(22) Gordon, RJ. 1980. Why is knowing about blood rheologyimportant in anesthesia? ASA Refresher Course Iectures 111:1-9.(23) Staub, NC. 1974. Pulmonary Edema, Physiol. Rev. 54:678.(24) Gadboys, HL, Slonim, R, Litwak, RS. 1962. The homo-logous blood syndrome: Preliminary observations of its relation-ship to clinical cardiopulmonary bypass, Ann. Surg. 156-797.(25) Zuhdi, N, McCollough, B, Carey, J, and Greer, A. 1961.Double helical reservoir heart lung machine designed for hypo-thermic perfusion primed with five percent glucose in waterinducing hemodilution, Arch. Surg. 82:320.(26) Chatrath, RR, Kaul, TK, and Walker, DR. 1980. Myocar-dial protection during cardioplegia in open-heart surgery, Canad.Anaesth. Soc. J. 27 (4) :381-387.(27) Shumway, NF, Lower, RR, and Stoffer, RC. 1959. Selectivehypothermia of the heart in anoxic cardiac anesthesia, Surg.Gynae. and Obst. 129:750-754.(28) Angell, WW, Rikkers, L, Dong. E. and Shumway, NE.1969. Organ viability with hypothermia J. Thorac. Cardiovasc.Surg. 57:619-624.(29) Hooker, DR. 1930. On recovery of the heart in electricshock, Am. J. of Physiol. 91:305-328.(30) Melrose, DG, Dreyer, B, Bentall, AJ, et al. 1955. Electivecardiac arrest, Lancet 2:21.(31) Ringer's. 1883. 1. Physiol. 4:29-44.(32) Hattersley, DG. 1966. Activated coagulation time of wholeblood, ]AMA 196:435.(33) Bull, BS, Huse, WM, Brauer, FS, et al. 1975. Heparintherapy during extracorporeal circulation; Use of a dose responsecurve to individualize heparin and protamine dosage, J. Thorac.Cardiovasc. Surg. 69:785.

December/1981 575

(34) Cargaff, F, Olson, KB. 1937-38. Studies in the chemistryof blood coagulation VI Studies on the action of heparin andother anticoagulants. The influence of protamine on the anti-coagulent effects in vivo, 7J. Biol. Chem. 122:153.(35) Wilson, JW. 1972. Treatment or prevention of pulmonarycellular damage with pharmacologic doses of corticosteroids,Surg. Gyne. and Obstet. 134:675-681.(36) Dietzman, RH, Leenseth, JB. Goott, B, et al. 1975. Theuse of methylprednisolone during cardiopulmonary bypass, .Thorac. Cardiovasc. Surg. 69:870-873.(37) Coffin, LH, Shinozaki, T, DeMeules, JE, et al. 1975. Inef-

fectiveness of methylprednisolone during cardiopulmonary by-pass. Am. J. Surg. 130:555-559.(38) Norman, JC. 1972. Cardiac Surgry, 2nd edition, Appleton.Century-Crofts, pp. 133-183, 545-555.(39) Stahl, WM and Stone, AM. 1970. Prophylactic diuresiswith ethacrynic acid for prevention of postoperative renalfailure, Ann. Surg. 172:361.

ADDITIONAL READINGI. Glenn, WWL, Liebow, AA, and Lindskog, GE. 1975. ThoracicCardiova~scular Surgery with Related Pathology, 3rd edition,Appleton-Century-Croft, pp. 1163-1225.2. Wylie, WD and Churchill-Davidson, HG. 1978. A Practice ofAnesthesia, 4th edition, W.B. Saunders Co., pp. 608-637.

AUTHORH. Lynn Thomas, CRNA, BS, attended Oklahoma State

University where he received a BS in Microbiology. He receivedhis diploma in nursing from St. Anthony Hospital School ofNursing in Oklahoma City, Oklahoma. Mr. Thomas wrote thispaper while a senior student at Wausau Hospital Center Schoolof Anesthesia in Wausau, Wisconsin. Mr. Thomas currently servesas a Certified Registered Nurse Anesthetist in Austin, Texas.

Journal of the American Association of Nurse Anesthetists