REVISTA
BRASILEIRA
DE
ANESTESIOLOGIA
OfficialPublicationoftheBrazilianSocietyofAnesthesiologywww.sba.com.br
SCIENTIFIC
ARTICLE
Monitorization
of
the
effects
of
spinal
anaesthesia
on
cerebral
oxygen
saturation
in
elder
patients
using
near-infrared
spectroscopy
Aysegul
Kusku
a,
Guray
Demir
b,∗,
Zafer
Cukurova
b,
Gulay
Eren
b,
Oya
Hergunsel
baDepartmentofAnesthesiologyandReanimation,AksehirStateHospital,Aksehir,Konya,Turkey
bDepartmentofAnesthesiologyandReanimation,BakirkoyDr.SadiKonukTrainingandResearchHospital,Istanbul,Turkey
Received2April2013;accepted10June2013 Availableonline18October2013
KEYWORDS
Cerebraloxygen saturation; Spinalanaesthesia; Near-infrared spectroscopy
Abstract
Objective: Centralblockageprovidedbyspinalanaesthesiaenablesrealizationofmanysurgical procedures,whereashemodynamicandrespiratorychangesinfluencesystemicoxygendelivery leadingtothepotentialdevelopmentofseriesofproblemssuchascerebralischemia, myocar-dialinfarctionandacuterenalfailure.Thisstudywasintendedtodetectpotentiallyadverse effectsofhemodynamicandrespiratorychanges onsystemicoxygendeliveryusingcerebral oxymetricmethodsinpatientswhounderwentspinalanaesthesia.
Methods:Twenty-fiveASAI---IIGrouppatientsaged65---80yearsscheduledforunilateralinguinal herniarepairunderspinalanaesthesiawereincludedinthestudy.Followingstandard moni-torizationbaselinecerebraloxygenlevelsweremeasuredusingcerebraloximetricmethods. Standardized Mini Mental Test (SMMT)was applied before andafter the operationso as to determine the levelof cognitivefunctioning of thecases. Using a standardtechnique and equal amountsofalocalanaesthetic drug(15mgbupivacaine 5%)intratechalblockadewas performed.Meanbloodpressure(MBP),maximumheartrate(MHR),peripheraloxygen satura-tion(SpO2)andcerebraloxygenlevels(rSO2)werepreoperativelymonitoredfor60min. Pre-andpostoperativehaemoglobinlevelsweremeasured.Thevariationsindataobtainedandtheir correlationswiththecerebraloxygenlevelswereinvestigated.
Results:Significantchangesinpre-andpostoperativemeasurementsofhaemoglobinlevelsand SMMTscoresandintraoperativeSpO2levelswerenotobserved.However,significantvariations were observed inintraoperative MBP, MHRandrSO2 levels. Besides,a correlationbetween variationsinrSO2,MBPandMHRwasdetermined.
∗Correspondingauthor.
E-mail:[email protected](G.Demir).
Conclusion:Evaluationofthedataobtainedinthestudydemonstratedthatpost-spinaldecline inbloodpressureandalsoheartratedecreasessystemicoxygendeliveryandadverselyeffects cerebral oxygen levels. However, this downward change did not result in deterioration of cognitivefunctioning.
©2013SociedadeBrasileiradeAnestesiologia.PublishedbyElsevier EditoraLtda.Allrights reserved.
PALAVRAS-CHAVE
Saturac¸ãodeoxigênio cerebral;
Raquianestesia; Espectroscopiadeluz próximaao
infravermelho(NIRS)
Monitoramentodosefeitosdaraquianestesiasobreasaturac¸ãodeoxigêniocerebral
empacientesidososcomousodeespectroscopiadeluzpróximaaoinfravermelho
Resumo
JustificativaeObjetivo:o bloqueio central proporcionado pela raquianestesia possibilita a realizac¸ãodemuitosprocedimentoscirúrgicos,enquantoasalterac¸õeshemodinâmicase respi-ratóriasinfluenciamaofertadeoxigêniosistêmico,levandoaodesenvolvimentoempotencial deumasériedeproblemas,comoisquemiacerebral,infartodomiocárdioeinsuficiênciarenal aguda.Oobjetivodesteestudofoidetectarpotenciaisefeitosadversosdasalterac¸ões hemod-inâmicas e respiratóriassobre a oferta deoxigênio sistêmico,usando métodos oximétricos cerebraisempacientessubmetidosàraquianestesia.
Métodos: vinteecincopacientes, 65---80anosdeidade,estadofísicoASAI---II,programados paracorrec¸ãodehérniainguinalunilateralsobraquianestesiaforamincluídosnoestudo.De acordocomomonitoramentopadrão,osníveisdeoxigêniocerebralforammedidosnoinício doestudousandométodosoximétricoscerebrais.OMiniTestePadronizadodoEstadoMental (StandardizedMiniMentalTest---SMMT)foiaplicadoantesedepoisdaoperac¸ãoparadeterminar oníveldefuncionamentocognitivodoscasos.Usandoumatécnicapadrãoequantidadesiguais deum fármacoanestésico local(15mgdebupivacaínaa5%),obloqueiointratecalfoi real-izado.Pressãoarterialmédia(PAM),frequênciacardíacamáxima(FCM),saturac¸ãoperiférica deoxigênio(SpO2)eníveiscerebraisdeoxigênio(rSO2)forammonitoradosnopré-operatório por60min.Osníveispré-epós-operatóriosdehemoglobinaforammedidos.Asvariac¸õesnos dadosobtidosesuascorrelac¸õescomosníveiscerebraisdeoxigênioforaminvestigadas.
Resultados: nãoobservamosalterac¸õessignificativasnasmensurac¸õesdehemoglobina,escores doSMMTeníveisdeSpO2nosperíodospré-epós-operatório.Noentanto,variac¸õessignificativas foramobservadasnos níveis de PAM,FCM erSO2 noperíodo intraoperatório.Alémdisso, a correlac¸ãoentreasvariac¸õesderSO2,PAMeFCMfoideterminada.
Conclusão:aavaliac¸ãodosdadosobtidosnoestudodemonstrouqueaquedadapressãoarterial pós-raquianestesiaetambémdafrequênciacardíacadiminuiaofertadeoxigêniosistêmicoe afetanegativamenteosníveiscerebraisdeoxigênio.Contudo,essaalterac¸ãonãoresultouem deteriorac¸ãodafunc¸ãocognitiva.
©2013SociedadeBrasileiradeAnestesiologia.PublicadoporElsevierEditoraLtda.Todosos direitosreservados.
Introduction
Thoughmain objectiveof spinalanaesthesiais toprovide sensory and motor blockade, sympathetic denervation is regardedasasideeffectinducingthedevelopmentof sys-temicalterations.1 Spinalanaesthesia-related hypotension isthemostfrequentlyencounteredcomplication.Systemic vascular resistance decrease, as a result of decreased arterialpressureandheartratedecreasesecondaryto sym-pathetic blockade resulting in decline in cardiac output. Systemicdeliveryofoxygendecreasesinproportionwitha decreaseincardiacoutputleadingtoanonsetofmany prob-lemssuchasdevelopmentofcerebralischemia,myocardial infarction,acuterenalfailureandcardiacarrestbecauseof tissuehypoxia.1,2
Elder population increases globally at an extremely higherrateinparallelwithimprovementsinthequalityof life.Elder andvery old peopleare considered tobe ≥65 and≥80yearsofage,respectively.3When comparedwith
generalanaesthesia,applicationofspinalanaesthesiaoffers elderpeoplesomeadvantagesduringandafterthe opera-tionaspreservationofcognitivefunctioning,lesseramount of intraoperative bleeding, decreased risk of thromboem-bolismandprovisionofeffectiveanalgesia.However,ithas also some disadvantages as hypotension, bradycardia and delayedambulation.4---10
Evenifdecreasedcardiacoutputcausedbyspinal anaes-thesiadoesnotimpairhemodynamicprocessesandsystemic delivery of oxygen at an extreme rate or induce clinical symptoms, it especially exerts a certain impacton cere-bral blood flow. Even though markedlydepressing effects ofhypotensiononcerebralcirculationofparticularlyelder patients hasbeen notedin variousstudies,this subjectis stilldebatable.11,12
is a monitor used for the measurement of regional cere-braloxygensaturation(rSO2).Thismonitordoesnotrequire pulserateandbloodflowmeasurements.Itmainly demon-strates the balance between cerebral oxygen deliver and demandforoxygen,inadditiontorSO2intargetorgans.13
Systemicdelivery ofoxygen(DO2)isusedtodetermine the amount of oxygen required bytissues. Cardiac stroke volumeisindirectproportionwithheartrate,haemoglobin levelandarterialbloodoxygencontent.Incaseswhereone or morethan oneofthese parametersdecrease quantita-tively,systemicdelivery ofoxygenis alsoreducedleading to tissue hypoxia. In practice, each of the more easily measured parametres suchas blood pressure, heart rate, peripheral oxygen saturation and haemoglobin levels is a predictorofsystemicoxygendelivery.Cerebraloxygenation is also affected by levels of haemoglobin, MBP, MHR and SpO2.14,15 Inthis study,theseparametreshave been mon-itoredinpatientsover65yearsofagewhohadundergone spinalanaesthesia.
In thisstudy,our aimwastoinvestigate theimpactof thechangesdevelopedsecondarytotheapplicationofspinal anaesthesiaoncerebraloxygenationinelderpeople accord-ingtoastandardprotocol.
Materials
and
methods
Afterapprovaloftheinstitutionalethicscommittee,atotal of25casesaged65---80yearsinASAI---II(AmericanSocietyof Anaesthesiologists)riskgroupswhowerescheduledfor uni-lateral hernia repair to be performed by the department of general surgery under preferred mode of anaesthesia (i.e.spinalanaesthesia)wereincludedinthestudy.Patients withmentaland/orneurologicaldiseases,congestiveheart failure,anemia,hematologicdisorderswerenotincludedin thestudy.Spinalanaesthesiaadministeredcaseswhose lev-elsofsensoryblockadeareinadequatefortherealizationof theoperationoraboveT10asdetectedbyapin-pricktest wereexcludedfromthestudy.
Thepatientswerebroughtintothepreoperative prepa-ration room and their baseline blood pressures (MAP), maximum heart rates (MHR), peripheral (SpO2) and cerebral oxygen saturation (Somanetics Invos Oxymeter 5100C:SomaneticsComp.,1653 EastMapleRoad, Troy,MI, USA)valuesweremonitoredandrecorded.Avenousaccess route was opened using an 18---20G peripheral vein can-nulaand RLsolution wasgiven intravenouslyat a rateof 10mLkg−1. Then the patients were taken into the oper-ating room and values of MAP, MHR, rSO2 and SpO2 were recorded.While thepatients intheerect sittingposition, the skin of the entry site was disinfected and covered with a sterile drape in compliance with an antiseptic technique. Witha 22---25 Quincke needle,L3---L4 interver-tebral space was entered and the needle was advanced into subarachnoidal space and intratechal blockade was realized using a standard technique and equal amounts of a local anaesthetic drug (15mg, 0.5% bupivacaine) for all patients. After spinal anaesthesia, the patients were placed in an appropriate supine position. Sensory block levelsof thecases were determined usingpin-prick tests performed within postspinal 10min (3, 5 and 10 min) andevaluatedbased onresponsive dermatomes.Pre- and
postoperative haemoglobin levels were determined. Any medicationeffective oncognitivefindingswasnot admin-isteredduringtheperioperativeperiod.StandardizedMini MentalTest (SMMT)wasappliedonpatientssoasto eval-uatecognitivefunctioning beforeandwithin6h afterthe operation.Thescoresobtainedwererecorded.Achievement ofadequatesensoryblockadewasawaitedfor30minafter inductionof spinalanaesthesia(atthe startof the opera-tion,VAS<20mm), then surgical intervention wasstarted after approval of the anaesthesiologist. After spinal tap, hemodynamicparametreswerefollowedupat5-min inter-vals for 60min. Administration of ephedrine wasplanned incases where actual blood pressuremeasurementswere 30% lower than the baseline values or decline in systolic ordiastolicbloodpressuresbelow90mmHgand40mmHg, respectively,whileatropinsulphateinjectionswere consid-eredinpatientswithmaximumheartrates below40bpm. Thedataretrievedwereevaluatedstatisticallyand correla-tionsbetweensignificantlyalteredvariablesandvariations incerebraloxygenationwereevaluated.
Statisticalanalysis
Forthe statistical analysis of study findings, NCSS (Num-berCruncherStatisticalSystem)2007&PASS2008Statistical Software(Utah,USA)programwasused.Asanoutcomeof Poweranalysis,inevaluationsbasedontheaverageofthe leftrSO2measurements,whenweassumeadifferenceof12 andstandarddeviationof10,inthedefinedgroupswitha statistical power of 0.90, ˇ: 0.10 and ˛:0.10, the samp-ling size was determined as 23 patients for each group. In order to determine the correlation between rSO2 and MBPmeasurements, measurementtimepointsweretaken as covariants and generalized linear regression analyses wereperformed.Fortheevaluationofstudydata descrip-tive statistical methods (means±standard deviation) and fortheanalysisofcorrelationsbetweenparametersinthe comparison of quantitative data Spearman’s rho correla-tioncoefficientswereused.Significancewasevaluatedat ap<0.05level.
Results
The study was performed on 19 (76%) male and 6 (24%) femalepatients. Ages of the patients ranged between 65 and77yearswithameanageof69.80±4.38years.
Statisticallysignificantvariationswereobserved among MBP measurements performed at baseline and at 5, 10, 15, 20, 30, 40, 50 and 60 min (p=0.0001) (Table 1). In ordertodeterminethecorrelationbetweenmeasurements ofrSO2andMBP,measurementtimepointswereassumedas covariantsandgeneralizedlinearregressionanalyseswere performed.Asignificantcorrelationwasobservedbetween rSO2 andMBPvaluesandamongmeasurementtimepoints (p=0.006 and p=0.001, respectively). Adjusted R2 value which ascertained the level of correlation was estimated as0.269(Table2).
Table1 LevelsofMBP,MHR,SpO2,rightandleftrSO2,SMMTandHb.
Base 5min 10min 15min 20min 25min
MBP 114.88±16.10 106.64±13.38 100.16±14.23 97.32±14.61 92.52±13.66 89.88±13.38 MHR 78.20±11.32 75.92±11.06 74.80±10.69 73.92±10.22 74.04±9.57 72.56±9.82 SpO2 98.53±1.04 98.52±1.15 98.36±0.99 98.24±1.05 98.36±1.18 98.52±0.96 RightrSO2 65.04±6.75 64.08±6.23 62.84±6.28 61.64±6.08 60.56±5.81 58.88±5.41 LeftrSO2 64.96±5.74 63.84±5.85 62.44±5.73 61.36±5.68 59.80±5.42 57.84±5.34
Beforeoperation
SMMT 18.44±2.53
Hb 12.56±1.78
30min 40min 50min 60min p
MBP 88.60±13.20 86.40±12.90 84.88±11.13 83.80±11.71 0.0001
MHR 72.08±9.15 72.20±9.35 72.40±9.21 72.04±8.76 0.0001
SpO2 98.32±1.03 98.44±1.04 98.40±1.08 98.32±0.94 0.598
RightrSO2 57.68±5.03 56.96±5.00 56.32±4.96 55.76±5.01 0.0001
LeftrSO2 56.88±5.34 56.40±5.08 55.64±5.30 54.88±5.24 0.0001
Afteroperation p
SMMT 18.48±2.55 0.664
Hb 11.80±1.56 0.132
Table2 MeasurementtimepointswereassumedascovariantsinthedeterminationofthecorrelationbetweenrSO2andMBP andMHRandgeneralizedlinearregressionanalysiswasperformed.Astatisticallysignificantcorrelationwasobservedbetween levelsofrSO2,MBPandMHRandalsoamongmeasurementtimepoints.AdjustedR2valuesascertainingthelevelofcorrelation werealsocalculated.
rSO2 Collectionofsquares Degreesofpermissiveness Meansquares F p
MBP 240.429 1 240.429 7.740 0.006
MHR 225.330 1 225.330 7.246 0.007
Measurementtimes
MBP 2853.631 9 317.070 10.207 0.001
MHR 5023.388 9 558.154 17.950 0.001
Residue
MBP 15190.591 489 31.065
MHR 15205.690 489 31.095
rSO2/MBP rSO2/MHR
R2 0.291 0.287
AdjustedR2 0.269 0.268
betweenrSO2andMHRvaluesandamongmeasurementtime points(p=0.007, andp=0.001,respectively).Adjusted R2 value which determined the level of the correlation was foundtobe0.268(Table2).
AsignificantchangewasobservedamongrightrSO2 mea-surements obtained at baseline, 5, 10, 15, 20, 25, 30, 40, 50 and 60min (p=0.0001) (Table 1). When compared withprespinalbaselinevalues,asteadydeclinewasnoted inrightrSO2 measurements.Besides, significantvariations were observed amongleft rSO2 measurements performed at baseline, 5, 10, 15, 20, 25, 30, 40, 50 and 60min (p=0.0001) (Table 1). A steady decline was also seen in leftrSO2measurementsrelativetoprespinalbaselinevalues
(p=0.0001)(Table1).Asignificantchangewasnotobserved among SPO2 measurements (p=0.598), pre- and postop-erative SMMT scores (p=0.664) and haemoglobin levels (p=0.794)(Table1).
Discussion
theelder population,acceptable levelofhaemoglobinfor cerebralperfusionwasindicatedas7g/dL.16---18Inthisstudy, any patient with severe bleeding episodes which would adversely affect haemoglobinlevelsand respiratory prob-lemswhichwouldunfavorablyaffectarterialbloodoxygen contentwerenotencountered.
Hypotension which occurs in patients operated under spinal anesthesia is the most frequently (33%) encoun-teredcomplicationofthespinalanaesthesia.Hypotensionis relatedtothelossofsystemicvascularresistancebecause ofsympatheticblockade.Venulardilationenhancesvenous capacityleadingtostagnationofbloodinveinsandreduced cardiacoutput.1,2 Hypotension mightbe severeenough to affectblood circulation invariousorgansystems.19 Inline withthisfinding, various studieshave been performed on theimpactofspinal anaesthesiaoncerebral oxygenation. Someofthesestudieshavedemonstratedthatspinal anaes-thesia effects cerebral oxygenationunfavourably which is moreprominent in the elder patients.11,12,20 In thisstudy, significant changes were observed in rSO2, MBP and MHR measurementsperformedatpostspinal5,10,15,20,25,30, 40,50,and60minwhencomparedwiththebaselinevalues. (p=0.0001). A correlation was noted between decreases incerebral oxygenation,MBPandMHRvalues.These data revealedthatdecreaseincerebraloxygenlevelsare associ-atedwithinabilitytocompensatehypotensiveattackswhich occur asa result of degeneration of cerebral vessels and deranged autoregulatory mechanisms and also decreased cardiacoutputbecauseofdecelerationofheartrateinthe elderly.
Respiratory functions in spinal anaesthesia are not adversely affected providedthat spinal blockade involves dermatomesinnervatedbyT7---T10spinalnerves.Pulmonary functionsarenotaffectedandrespiratoryrateperminute, end-tidalCO2,PaO2,PaCO2valuesdonotchangeinspinal blockade up to T4 level thanks to compensatory mecha-nisms exerted by diaphragma which is innervated by N. phrenicus.17Inthisstudy,thelevelofspinalblockadedidnot exceedT8levelinpatientswhohadundergonespinal anaes-thesia, accordingly significant respiratory changes which wouldotherwiseadverselyaffectSpO2andcerebraloxygen deliverywerenotobserved(p=0.598).
Cognitivedysfunctioncanbedefinedaslossofmemory andintellectual talents.Mostfrequently, hypotensionand hypothermia increase the risk of postoperative cognitive dysfunction.Standardized minimental test is ascreening test which quantitatively evaluates cognitivefunctions. It is preferred in various studies thanks to its established validity, reliability, brevity, and easy applicability during pre-andpostoperativeperiods.21Theseverityof deteriora-tionincognitivefunctions inpatients whoreceivedspinal anaesthesia is dependent on the degree and duration of hypoxiaandhypotension.22 Thoughanyalterationin respi-ratory functions was not observed in our study and the levelsofhemodynamic parametersdecreased without any significantdifferencebetweenpre-andpostspinalcognitive functionsasassessedbySMMTscores (p>0.05).Decreases in blood pressure and heart rates observed in patients included in our study did not adversely affect cognitive functions.
In conclusion, hemodynamic changes occurring during intratechal anaesthesiain elderlypatients affectcerebral
oxygenation adversely. This effect is directly correlated tothe developmentof hypotensionanddecreasein heart rate.Although in thisstudy hemodynamic changes devel-opingduringspinalanaesthesiadidnotleadtoaclinically symptomaticstate, literature datahave emphasizedtheir potential risk in the elder patients. Owing to evolving innovativeimagingtechniques, it ispossible todetermine disordered cerebral oxygen balance. Among them, NIRS is a reliable and easily applicable method. This study has demonstrated that this imaging modality is helpful in the management of anaesthesia in the elderly and it can beroutinely usedin risky patients thanks toits easy applicability.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
References
1.Albright G, Forster R. Spinal analgesia-physiolgic effects. In: Collins VJ, editor. Principles of anesthesiology. 3rd ed. Philedelphia:Lea&Febiger;1993.p.1445---570.
2.Atkinson RS. Spinal analgesia. In: Atkinson RS, Rushmman GB, DaviesNJH,editors. Lee’s synopsisof anaesthesia.11th ed.Oxford:Buttenvort-HeinemannInternationalEdition;1993. p.691---719.
3.AbramsWB,BeersMH,BerkowR,editors.Acrossnational per-spective.TheMerckmanualofgeriatrics.WhitehouseStation, NJ:MerckResearchLaboratories;1996.p.587---95.
4.˙Ichley LA. Hypotension, subarachnoid block and the elderly patient.Anaesthesia.1996;51:1139---43.
5.Rooke GA, Freund PR, Jacobson AF. Hemodynamic response and change in organ blood volume during spinal anesthesia in elderlymen withcardiac disease. AnesthAnalg. 1997;85: 99---105.
6.˙IchleyLA, StuartJC, ShortTG,et al.Haemodynamiceffects of subarachnoid block in elderly patients. Br J Anaesth. 1994;73:464---70.
7.˙Ishikawa K, Yamakage M, Omote K, et al. Prophylactic IM smalldose phenylephrine blunts spinal anesthesia induced hypotensiveresponseduringsurgicalrepairofhipfracturein theelderly.AnesthAnalg.2002;95:751---6.
8.Eupre LA, Jones CA, Saunders LD, et al. Best practices for elderlyhipfracturepatients.Asystematicoverviewofthe evi-dence.JGenInternMed.2005;20:1019---25.
9.Sheehan E, Neligan M, Murray P. Hip arthroplasty, changing trends in a national tertiary referral centre. Ir J Med Sci. 2002;171:13---5.
10.CarpenterRL,CaplanRA,BrownDL,etal.Incidenceandrisk factors for side effects ofspinal anesthesia. Anesthesiology. 1992;76:906---16.
11.HoppensteinD,ZoharE,RamatyE,etal.Theeffectsofgeneral vsspinal anesthesia onfrontalcerebraloxygensaturation in geriatricpatientsundergoingemergencysurgicalfixationofthe neckoffemur.JClinAnesth.2005;17:431---5.
12.Nishikawa K, Hagiwara R, Nakamura K, et al. The effects of the extentof spinal block on theBIS score and regional cerebraloxygensaturationinelderlypatients:aprospective, randomized,and double-blindedstudy.JClinMonitComput. 2007;10:109---14.
14.BoumaGJ,MuizelaarJP.Cerebralbloodflowinsevereclinical headinjury.NewHoriz.1995;3:384---94.
15.PaulsonOB,WaldemarG,SchmidtJF,etal.Cerebral circula-tionundernormalandpathologicalconditions. AmJCardiol. 1989;2:2---5.
16.Cook DJ, OliverJr WC, OrszulakTA, etal. Cardiopulmonary bypasstemperature,hematocrit,andcerebraloxygendelivery inhumans.AnnThoracSurg.1995;60:1671---7.
17.Hino A, Ueda S, MizukawaN, et al. Effect of hemodilution on cerebral hemodynamics and oxygen metabolism. Stroke. 1992;23:423---6.
18.SungurtekinH,CookDJ,OrszulakTA,etal.Cerebralresponse tohemodilutionduringhypothermiccardiopulmonarybypassin adults.AnesthAnalg.1999;89:1078---83.
19.McCrae AF, Wildsmith JAW. Prevention and treatment of hypotension during central block. Br J Anaesth. 1993;70:672---80.
20.MinvilleV,Asehnoune K,SalauS,etal. Theeffectsofspinal anesthesiaoncerebralbloodflowinveryelderly.Anesth Anal-gesia.2009;108:1291---4.
21.YaoFS,TsengCC,HoCY, etal.Cerebraloxygendesaturation isassociatedwithearlypostoperativeneuropsychological dys-functioninpatientsundergoingcardiacsurgery.JCardiothorac VascAnesth.2004;18:552---8.
