随着对危重症患者抢救成功率的提高.docx

随着对危重症患者抢救成功率的提高.docx

《随着对危重症患者抢救成功率的提高.docx》由会员分享，可在线阅读，更多相关《随着对危重症患者抢救成功率的提高.docx（13页珍藏版）》请在冰豆网上搜索。

随着对危重症患者抢救成功率的提高

IntensityofContinuousRenal-ReplacementTherapyinCriticallyIllPatients

TheRENALReplacementTherapyStudyInvestigators

NEnglJMed2009;361:

1627-1638October22,2009DOI:

10.1056/NEJMoa0902413

Share:

Abstract

Article

References

CitingArticles （243）

Letters

Acutekidneyinjuryisassociatedwithsubstantialmorbidityandmortality.1 Itisacommonfindingamongpatientsintheintensivecareunit（ICU）2 andisanindependentpredictorofmortality.3 Acutekidneyinjurysevereenoughtoresultintheuseofrenal-replacementtherapyaffectsapproximately5%ofpatientsadmittedtotheICUandisassociatedwithamortalityrateof60%. 4 Theoptimalapproachtorenal-replacementtherapy,aswellastheoptimalintensityandtimingofsuchtherapy,incriticallyillpatientsremainsunclear.Inonesingle-center,randomized,controlledstudyinwhichcontinuousrenal-replacementtherapywasthesoletreatmentapproach,survivalimprovedwhentheintensityoftherapywasincreasedfromanassignedeffluentrateof20mlperkilogramofbodyweightperhourtoeither35or45mlperkilogramperhour.5 However,subsequentsingle-centerstudieshavehadconflictingresults.6-8

TherecentlyreportedVeteransAffairs/NationalInstitutesofHealthAcuteRenalFailureTrialNetworkStudy（ClinicalTrials.govnumber,NCT00076219）9 showedthatincreasingtheintensityofrenal-replacementtherapydidnotdecreasemortalityamongpatientswithacutekidneyinjury.Incontrasttootherstudies,whichusedcontinuousrenal-replacementtherapyexclusively,thisstudyassignedpatientstoaprotocolofeitherintermittentorcontinuousrenal-replacementtherapyaccordingtowhethertheywerehemodynamicallystableorunstable,respectively.ThisdesignreflectsclinicalpracticeintheUnitedStatesandelsewherebutmakesitdifficulttocarryoutaformalcomparisonoftreatmentintensitiesthatwouldbeindependentoftheparticulartreatmentapproach.Weconductedarandomized,controlledstudytotestthehypothesisthatincreasingtheintensityofcontinuousrenal-replacementtherapywouldreducemortalityat90days.

METHODS

StudyDesign

TheRandomizedEvaluationofNormalversusAugmentedLevel（RENAL）ReplacementTherapyStudywasaprospective,randomized,parallel-grouptrialdesignedtoassesstwolevelsofintensityofcontinuousrenal-replacementtherapyincriticallyillpatientswithacutekidneyinjury.ThestudywasconductedbetweenDecember30,2005,andNovember28,2008,in35ICUsinAustraliaandNewZealand.Thestudyprotocolisoutlinedinthe SupplementaryAppendix,availablewiththefulltextofthisarticleatNEJM.org.ItwasapprovedbythehumanresearchethicscommitteesoftheUniversityofSydneyandallparticipatinginstitutions.TheintegrityofdatacollectionwasverifiedbytheGeorgeInstituteforInternationalHealthmonitoringteam.Anindependentdataandsafetymonitoringcommitteereviewedsafetydataandinterimresultswiththeaimofprovidingadvicetothetrialmanagementcommitteeshouldsuchanalysesprovebeyondareasonabledoubtthataugmentedcontinuousrenal-replacementtherapyledtoanetbenefitorharmintermsofmortality.

StudyPopulation

Patientswereeligibleforenrollmentiftheywerecriticallyill,were18yearsofageorolder,hadacutekidneyinjury,weredeemedbythetreatingcliniciantorequirerenal-replacementtherapy,andmetatleastoneofthefollowingcriteria:

oliguria（urineoutput<100mlina6-hourperiod）thatwasunresponsivetofluidresuscitationmeasures,aserumpotassiumconcentrationexceeding6.5mmolperliter,severeacidemia（pH<7.2）,aplasmaureanitrogenlevelabove70mgperdeciliter（25mmolperliter）,aserumcreatinineconcentrationabove3.4mgperdeciliter（300μmolperliter）,orthepresenceofclinicallysignificantorganedema（e.g.,pulmonaryedema）.Writteninformedconsentwasobtainedfromthepatientorresponsiblesurrogatebymeansofeitheraprioriordelayedconsent.（Foradetaileddescriptionofdelayedconsent,seethe SupplementaryAppendix.）

Patientswhohadreceivedanypreviousrenal-replacementtherapyduringthesamehospitaladmissionorwhowereonmaintenancedialysisforend-stagekidneydiseasewereineligibleforthestudy.（Foradetailedlistofinclusionandexclusioncriteriaandthecriteriafordiscontinuingthestudytreatment,seethe SupplementaryAppendix.）

Intervention

Thepatientsinbothgroupsweretreatedwithcontinuousvenovenoushemodiafiltration.Replacementfluidwasdeliveredintotheextracorporealcircuitafterthefilter（i.e.,postdilution）,witharatioofdialysatetoreplacementfluidof1:

1.Theeffluentflowprescribedwasbasedonthepatient'sbodyweightatthetimeofrandomizationandwaseither40mlperkilogramperhour（forthehigher-intensitygroup）or25mlperkilogramperhour（forthelower-intensitygroup）.Bloodflowwaskeptabove150mlperminute.Fluidwasremovedbydecreasingtheflowofthereplacementfluidandofthedialysateinequalproportion,sothateffluentexceededthembothbyanyamountprescribedbytheclinician.FilterswiththeAN69membrane（Gambro）wereused.HemosolBOfluid（Gambro）wasusedasthedialysateandreplacementfluid.Gambrohadnoroleintheinitiation,design,analysis,orreportingofthestudy.

StudyOutcomes

Theprimarystudyoutcomewasdeathfromanycausewithin90daysafterrandomization.Secondaryandtertiaryoutcomesincludeddeathwithin28daysafterrandomization,deathintheICU,in-hospitaldeath,cessationofrenal-replacementtherapy,durationofICUandhospitalstays,durationofmechanicalventilationandrenal-replacementtherapy,dialysisstatusatday90,andanyneworganfailures.

StatisticalAnalysis

Allstatisticalanalyseswereconductedaccordingtoapredefinedplan.10,11 Thetargetenrollmentwas1500patients,whichprovided90%powertodetectan8.5%absolutereductionin90-daymortalityfromabaselineof60%（alphalevel,<0.05）.Twointerimanalyseswereperformedandreviewedbyanindependentdataandsafetymonitoringcommittee.SincetheHaybittle–PetorulewithamaximumofthreeanalyseswasusedtolimittheoverallprobabilityofatypeIerrorto0.05,thefinalanalysiswasconductedatanalphalevelof0.048.

Allanalyseswereperformedaccordingtotheintention-to-treatprinciple,withnoimputationformissingvalues.Datafrompatientswhowerelosttofollow-upwerenotanalyzed.Proportionswerecomparedwiththeuseofthechi-squaretest,andcontinuousvariableswereanalyzedwiththeuseofStudent'st-test.Mantel–Haenszeladjustedoddsratiosandtheircorresponding95%confidenceintervalswerecalculated.Analysisoftheprimaryoutcomeforthetwogroupswasalsoperformedbymeansofthelog-ranktest,withtheresultspresentedasaKaplan–Meiercumulative-incidenceplot.

Prespecifiedsubgroupanalyseswereperformedaccordingtothepresenceorabsenceofsepsis;failureofoneormorenonrenalorgans;aSequentialOrganFailureAssessment（SOFA）cardiovascularscoreof3or4atbaseline（onascalerangingfrom0to4,withahigherscoreindicatingmoresevereorgandysfunction）;andanestimatedglomerularfiltrationrateoflessthan60mlperminutewithin6monthspriortorandomization.Weassessedsubgroupsforheterogeneityoftreatmenteffectaccordingtoacceptedclinicalguidelines.12

Statisticalanalyseswereperformed,independentlychecked,andreplicatedwiththeuseofSASsoftware,version9.1.

RESULTS

Enrollment

BetweenDecember1,2005,andAugust31,2008,weenrolled1508patients,ofwhom747wereassignedtothehigher-intensitytreatmentgroupand761tothelower-intensitytreatmentgroup（Figure1FIGURE1

NumbersofPatientsEnrolledintheStudy,RandomlyAssignedtoaTreatmentGroup,andIncludedintheAnalysis.）.Consentwassubsequentlywithheldorwithdrawnfor43patients（2.9%）,25ofwhomhadbeenassignedtohigher-intensitytherapyand18tolower-intensitytherapy;only1patientwaslosttofollow-up,thustheprimaryoutcomewasavailablefor1464patients（97.1%）.

BaselineCharacteristics

Allbaselinecharacteristicsweresimilarbetweenthetwogroups（Table1TABLE1

BaselineCharacteristicsoftheStudyPatients.）.Theserumcreatinineconcentrationsbeforerandomizationinthehigher-intensityandlower-intensitytreatmentgroupswere3.8mgperdeciliter（338μmolperliter）and3.7mgperdeciliter（330μmolperliter）,respectively.Inall,73.9%ofpatientswerereceivingmechanicalventilation,49.4%hadseveresepsis,and82.5%werereceivingvasoactivedrugs.

StudyandSupportiveTreatments

Table2TABLE2

CharacteristicsofStudyTreatmentsandSubsequentUseofRenal-ReplacementTherapy. liststhecharacteristicsofthestudytherapy.Themeandurationoftreatmentinthetwogroupswassimilar,butduringtherapy,theyhadsignificantlydifferentmeandailyserumcreatinineconcentrations（1.9mgperdeciliter[170μmolperliter]inthehigher-intensitygroupvs.2.3mgperdeciliter[204μmolperliter]inthelower-intensitygroup,P<0.001）andbloodureanitrogenlevels（35.6mgperdeciliter[12.7mmolperliter]vs.44.5mgperdeciliter[15.9mmolperliter],P<0.001）.Thesedifferenceswereconsistentwiththedifferenceintheintensityofthedeliveredtreatment（meaneffluentrate,33.4mlperkilogramofbodyweightperhourinthehigher-intensitygroupvs.22.0inthelower-intensitygroup;P<0.001）.Patientsreceivinghigher-intensitycontinuousrenal-replacementtherapyweremorelikelytoreceiveregionalextracorporeal-circuitanticoagulationwithheparinandprotamine（P=0.007）andrequiredmorefiltersperday（0.93vs.0.84,P<0.001）.Only7.6%and7.0%ofthepatientsinthehigher-in