Mobile

A FABP-ulous 'rule out' strategy? Heart fatty acid binding protein and troponin for rapid exclusion of acute myocardial infarction

Description
Many Emergency Departments (EDs) utilise 'triple marker' testing with CK-MB, myoglobin and troponin I (cTnI) to exclude acute myocardial infarction (AMI) within hours of presentation. We evaluated the ability of 8 biomarkers to rapidly
Categories
Published
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Resuscitation 82 (2011) 1041–1046 ContentslistsavailableatScienceDirect Resuscitation  jo   u   rnalhomepage:   www.elsevier.com/locate/resuscitation Clinical   paper A   FABP-ulous   ‘rule   out’   strategy?   Heart   fatty   acid   binding   protein   and   troponinfor   rapid   exclusion   of    acute   myocardial   infarction  Richard   Body a , ∗ ,GarryMcDowell a ,SimonCarley b ,Christopher   Wibberley c ,Jamie   Ferguson b ,Kevin   Mackway-Jones b a UniversityofManchester,OxfordRoad,Manchester,M139WL,UnitedKingdom b EmergencyDepartment,ManchesterRoyalInfirmary,OxfordRoad,Manchester,M139WL,UnitedKingdom c ManchesterMetropolitanUniversity,Facultyof    Health,PsychologyandSocialCare,HathersageRoad,Manchester,M13   0JA,   UnitedKingdom a   r   t   i   c   le   i   nf   o  Articlehistory: Received18January2011Receivedinrevisedform10March2011Accepted15March2011 Keywords: AcutemyocardialinfarctionAcutecoronarysyndromesTroponinsHeartfattyacidbindingproteinMyoglobinCK-MBMyeloperoxidaseBrainnatriureticpeptide d -DimerDiagnosisSensitivityandspecificity ab   s   t   ra   ct Objective:   Many   EmergencyDepartments   (EDs)   utilise‘triplemarker’testing   with   CK-MB,   myoglobin   andtroponin   I   (cTnI)   toexclude   acute   myocardial   infarction   (AMI)   within   hoursof    presentation.   Weevaluatedthe   abilityof    8   biomarkers   to   rapidly   excludeAMI   at   the   pointof    presentationand   investigated   whether‘triple   marker’   testing   representsthe   optimalmultimarker   strategy. Methods:   Werecruitedpatientswhopresentedto   the   ED   with   suspectedcardiac   chest   painoccurringwithin   24h.Bloodwasdrawn   at   the   time   of    presentation.Diagnostic   value   was   assessedby   calculatingthe   areaunderthe   ROC   curve(AUC)andamultivariate   model   was   constructedby   logisticregression.   Theprimary   outcome   wasa   diagnosis   of    AMI,   established   by   ≥ 12-htroponin   testing   inallpatients. Results:   705includedpatientsunderwent   venepuncture   a   medianof    3.5   hafter   symptom   onset.Heartfattyacid   binding   protein(H-FABP)had   an   AUC   of    0.86(95%CI0.82–0.90),   which   was   significantly   higherthan   anyother   biomarker   includingcTnI.   While   no   single   biomarker   could   enable   exclusion   of    AMI,multivariate   analysis   identified   cTnI   andH-FABP   as   the   optimal   biomarker   combination.   Combinedwithclinicalrisk   stratification,   thisstrategy   had   a   sensitivityof    96.9%,   specificity   of    54.7%,   PPV   32.4%   and   NPV98.8%. Conclusions:   We   have   derivedanalgorithm   that   would   enable   AMItobe   immediatelyexcluded   in315(44.7%)patients   atthe   cost   of    missing6AMIs   per1000   patients   treated.   While   the   risk   islikelytobeunacceptable   for   clinicalimplementation,   we   have   highlighted   anareaforfuture   developmentusingserial   testing   and   increasingly   sensitive   assays. © 2011 Elsevier Ireland Ltd. All rights reserved. 1.Background Theuseofbiochemicalmarkersisalreadyanessentialpartof theriskstratificationof    patientswithsuspectedacutecoronarysyndromes(ACS)intheEmergencyDepartment(ED).Measure-mentofcardiactroponinsisanestablishedstandardofpracticeforthesepatientsandformsanimportantpart   of    thegoldstan-dardforthediagnosisofacutemyocardialinfarction(AMI). 1–3 Howevertheinsufficientsensitivityof    troponinsatthetimeof    pre-sentationmandatesthatthemajorityof    patientsareadmittedtohospitalpendingtheirevaluation. 1,2,4 Followinginvestigation,only  ASpanishtranslatedversionof    theabstractofthisarticleappearsasAppendixin   thefinalonlineversionatdoi:10.1016/j.resuscitation.2011.03.015. ∗ Correspondingauthorat:CardiovascularSciencesResearchGroup,3rdFloor,Core   TechnologyFacility,TheUniversityof    Manchester,46,GraftonStreet,Manch-ester,M13   9WL,UnitedKingdom.Tel.:+4407880712929;fax:+441612766925. E-mailaddress: rbody@doctors.org.uk(R.Body). a   minorityof    thepatientsadmittedareactuallydiagnosedwithACS,suggestingthatthereistremendouspotentialtoreduceunneces-saryhospitaladmissions(andunnecessaryempiricaltreatment)inthispatientgroupbyimprovingdiagnostictechnology. 5 Inrecentyearstherehasthereforebeena   growinginterestinidentifyingcombinationsof    biomarkersthatmay   facilitateearlyexclusionof ACSatthetimeofpresentation.To   date,however,noneofthemul-timarkerpanelsinvestigatedhavebeenrecommendedforthisuseinpractice. 6–9 Creatinekinase-MB(CK-MB)andmyoglobinarebiomarkersof myocardialnecrosis,levelsofwhichareknowntoriseearlyaftertheonsetofmyocardialnecrosis.Usedinisolationthesebiomark-erslacksufficientsensitivitytoenablesafeexclusionof    AMI   in   theED. 10 However,‘triplemarkertesting’usingthecombinationof CK-MB,myoglobin(‘earlymarkers’)andtroponin(a‘latemarker’),improvessensitivity. 11,12 Indeed,studieshavesuggestedthatthisstrategyissafeandeffectiveandclinicalprotocolsincorporatingtriplemarkertestingcontinuetobeusedin   clinicalpracticeatmanyinstitutions. 13,14 Serialsamplingremainsnecessary,how- 0300-9572/$–seefrontmatter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2011.03.015  1042  R.Bodyetal./    Resuscitation 82 (2011) 1041–1046 ever,andtheprotocolhasmainlybeenvalidatedagainstwhatarenowoutdatedreferencestandardsforAMI. 15 Inrecentyearsanumberof    promisingnovelbiomarkershavebeenidentifiedthatmay   enhanceourabilitytoexcludeAMI   atthepointofpresentationtotheED.Theseincludebrainnatriureticpeptide(BNP),whichhasbeenshowntoincreasethediagnosticaccuracyoftriplemarkertesting; 13,16 heartfattyacidbindingpro-tein(H-FABP),anadditionalmarkerof    myocardialnecrosisthatrisesveryearlyaftertheonsetof    AMI; 17 d -dimer,a   markerof    fibrinclotbreakdownthatisusedwidelyforsuspectedvenousthrom-boembolismalthoughelevatedlevelshavealsobeendemonstratedinpatientswithACS; 18–20 andmyeloperoxidase(MPO),amarkerofinflammationandoxidativestressthatplaysanimportantroleinplaquedestabilization. 21,22 Finally,neutrophilgelatinaseasso-ciatedlipocalin(NGAL)isaglycoproteinfoundinthegranulesofneutrophils.NGALisa   sensitiveearlymarkerof    acutekidneyinjury 23 butitmay   alsoplaya   roleinmodulatingtheinflam-matoryresponse,actingasascavengerof    bacterialproductsatsitesofinflammation.NGALhasbeendemonstratedin   coronaryatheroscleroticplaqueinmice,withenhancedlevelsofexpressioninmicethatdevelopedmyocardialinfarction. 24 Further,NGALhasbeenshowntoaugmenttheactivityofmatrixmetalloproteinase-9(MMP-9),aproteolyticenzymethatmay   destabilisecoronaryatheromabythinningtheprotectivefibrouscap. 25 Therearethere-foreseveralreasonswhyNGALmay   havepromiseasanearlymarkerofAMI,althoughtodatetherehavebeennoreportsof    itsvalueinthisregard.We   aimedtoevaluatetheaccuracyof    triplemarkertesting(CK-MB,   myoglobinandtroponinI)fordiagnosingor   excludingAMIusingasingletestatthepointofpresentationto   theED.Fur-ther,weaimedtoinvestigatewhetherthisbiomarkercombinationisindeedoptimalorwhetherotherbiomarkers(H-FABP,BNP, d -dimer,myeloperoxidaseandNGAL)wouldhavegreaterbenefitandenablesafe,immediateexclusionof    AMI. 2.Methods Thisisasubstudyof    theEarlyVascularMarkersof    AcuteCoronarySyndromes(EVaMACS)project,a   prospectivediagnos-ticcohortstudyconductedatManchesterRoyalInfirmary,whichhasatotalannualEDcensusofapproximately145,000patients(including39,000‘majorcases’,19,000presentationstotheEmer-gencyEyeCentre,37,000presentationsto   theon-siteprimarycareservices,43,000presentationsto   theMinorInjuryUnitand7000others). 26–32 ThestudywasapprovedbytheLocalResearchEthicsCommittee.Allpatientsprovidedwritteninformedconsent.PatientswhopresentedtotheEDwithsuspectedcardiacchestpainoccurringwithintheprevious24hwereeligibleforinclusion.Subjectswithchesttraumacausingsuspectedmyocardialcontu-sion,renalfailurerequiringdialysis,anothermedicalconditionnecessitatinghospitaladmission,pregnancy,thosewho   couldnotspeakEnglishlanguageandprisonerswereexcluded.Patientswhodidnotconsenttoandprovideabloodsampleforanalysisbytheresearchteamwerealsoexcluded.Adesignatedresearcherwasphysicallypresentwithinthedepartmentforbetween10and24honover90%ofdaysandwasavailableon-call24hperday,7daysperweek(nocallsweredeclined).Inaddition,traininggiventoEDdoctorsensuredthatatleastonepersontrainedattakingconsentwasphysicallypresentintheED24hperday,7daysperweekthroughoutthestudyperiod.Finally,allEDdoctorsandnurseswereactivelyencouragedandremindedtocontacta   researcheronencounteringanypotentiallyeligiblepatient,regardlessof    thetimeofdayornight.Clinicaldatawererecordedatthetimeof    EDpresentationusinga   custom-designedcasereportform.Patientsunderwentvenepunctureatthetimeof    presentationtotheED.EDTA(ethylenediaminetetra-aceticacid)plasmasam-pleswerecentrifugedwithintheEDat10,000rpmfor10min   andimmediatelystoredat − 20 ◦ Cforupto48handat − 70 ◦ Cthere-after.Samplesweretransportedtoa   centrallaboratoryindryiceandwerethawedoncefortesting.LevelsofCK-MB(95thpercentile4.3   ng/ml),myoglobin(95thpercentile107ng/ml),troponinI(cTnI,99thpercentile<0.055ng/ml,limitofdetection0.055ng/ml),H-FABP(90thpercentile58ng/ml;95thpercentile68ng/ml),MPO(95thpercentile510pM), d -dimer(95thpercentile600ng/ml),BNP(95thpercentile73ng/ml)andNGAL(99thpercentile201ng/ml)weremeasuredin   automatedmannerbyfluorescenceimmunoas-say(AlereSanDiego,Inc.,SanDiego,CA)).Thereported90th,95thand99thpercentilesweredeterminedbythemanufacturerin   ahealthyreferencepopulation.Theprimaryoutcomeof    thisstudywasa   diagnosisof    acutemyocardialinfarction(AMI).All   patientsunderwentreferencestan-dardtroponinTtestingatleast12haftertheonsetof    themostsignificantsymptoms(Rochediagnostics,elecsysfourthgenera-tion,99thpercentile0.01ng/ml,co-efficientofvariation ± 10%at0.035ng/ml).Finaldiagnosiswasadjudicatedbytwo   independentphysicians,withreferencetoathirdindependentphysicianintheeventofanydisagreement,who   wereblindedto   levelsoftheinves-tigationalbiomarkersbuthadaccesstoreferencestandardtroponinTlevelsaswellasadditionalrelevantclinicalandimagingfindings,andwithreferenceto   theuniversaldefinitionofAMI. 3 Therewasonlyonedisagreementbetweentheindependentphysicians,whichwasaccountedforbyhumanerrorinreadinga   troponinlevel.  2.1.Statisticaltechniques As   biomarkerlevelsdidnotconformtothenormaldistribution(  p <   0.05,Kolmogorov–Smirnovtest),thedatawere   analysedusingnon-parametricmethods.Biomarkerlevelsweresummarisedbyreportingthemedianandinterquartilerange(IQR)andcompar-isonsbetweengroupsweremadewiththeMannWhitney U  test.Receiveroperatingcharacteristic(ROC)curveanalysiswas   usedto   comparetheoveralldiagnosticaccuracyof    thebiomarkers.TheareaundertheROCcurve(AUC)and95%confidenceintervals(95%CI)werecalculatedinSPSSversion16.0usinganon-parametricmethod.Sensitivity,specificity,positivepredictivevalue(PPV)andnegativepredictivevalue(NPV)werecalculatedusingthe90th,95thor99thpercentilecut-offsasspecifiedbythemanufacturerandcomparedusingMcNemar’stestforpairedproportions.Levelsofeachbiomarkerwerealsoenteredintoamultivariatestepwise(forwardconditional)logisticregressionanalysisinorderto   derivetheoptimalmodelforearlydiagnosisof    AMI.Basedonprevioussimilarstudies,weestimatedthatthepreva-lenceof    AMI   wouldbeapproximately20%. 33 Wethereforerequireda   samplesizeofapproximately730patientsto   detectof    atestwith100%sensitivitywithsufficientpowertoproducea   lower95%con-fidenceinterval(CI)>97.5%.StatisticalanalyseswereundertakeninSPSSversion16.0orMedCalcversion11.4.3.0.  2.2.Clinicalriskstratification AccordingtoBayesianprinciples,aninvestigationthataimsto‘ruleout’diseaseis   morelikelytobeeffectiveamongpatientswithalowpre-testprobabilityofdisease.WethereforecalculatedthediagnosticaccuracyofthemultivariatemodelifAMI   couldonlybeconsideredtobe‘ruledout’in   patientsat‘verylow’clinicalriskof AMIaccordingtotheGoldmanalgorithm, 34 whichhasbeenprevi-ouslyreportedto   beeffectivewhencombinedwithtriplemarkertesting. 13  R.Bodyetal./    Resuscitation 82 (2011) 1041–1046 1043 804 paents enrolled and consented AMI n=129 Included in final analysis n=705 Index test and reference standard n=705 No AMI n=576 Excluded, n=99: • No plasma sample available, n=91 • Withdrew consent, n=3 •   Renal dialysis, n=2 • Age <25 years, n=1 • Pain >24h ago, n=1 • Other medical condion mandang admission, n=1 Lost to follow up, n=0 Fig.1. Patientenrolmentandoutcomes. 3.Results Thestudycommencedat09:00on16thJanuary2006andendedat17:00on3rdFebruary2007.Atotalof    804patientswereincludedinthestudyduringthisperiod,705of    whomconsentedtoandpro-videdabloodsampleforthepurposesof    thisresearch.Allpatients(705)underwentreferencestandardtroponintesting ≥ 12h   aftersymptomonset.Intotal129(18.3%)patientshadAMI   ontheirindexadmission(Fig.1).Thebaselinecharacteristicsof    includedpatientsareshownin   Table1.   Mediantimefromsymptomonsettovenepuncturewas3.5   h(IQR1.8–7h).TheROCcurvesdemonstratingtheoverallabilityof    eachindi-vidualbiomarkertoaidthediagnosisof    AMIareshowninFig.2.H-FABPwasfoundto   havethelargestAUCat0.87,whichwas   signif-icantlygreaterthantheAUCforeachotherindividualmarker.Thesensitivity,specificity,PPVandNPV   of    eachindividualbiomarkerisdemonstratedin   Table2.Onmultivariateanalysis,a   combinationof    H-FABP(oddsratio18.2,95%CI   10.9–30.4)andcTnI(oddsratio11.9,95%CI   5.9–24.0)wasfoundtobeoptimalforearlydiagnosisof    AMI.Comparedtothetriplemarkerpanel(CK-MB,myoglobinandcTnI),thiscombina-tionhadsuperiorsensitivity(82.2%vs.   69.8%,  p <0.0001),specificity  Table1 Baselinecharacteristicsofthestudypopulation.Variable Total( N  =705)HadAMI   ( N    =   129)DidnothaveAMI( N    =   576)Ageinyears,mean(standarddeviation)58.6(14.3)Men   (%)430(61.0)87   (67.4)343(59.5)Previousangina(%)   220(31.2)28   (21.7)192(33.3)Previousmyocardialinfarction(%)165(23.4)28   (21.7)137(23.8)Hypertension(%)343(48.7)63   (48.8)280(48.6)Hyperlipidaemia(%)342(48.5)55   (42.6)287(49.8)Diabetesmellitus(%)126(17.9)26   (20.2)100(17.4)Smoking(%)218(30.9)57   (44.2)161(28.0)Familyhistoryof    ischaemicheartdisease(%)341(48.4)55   (42.6)286(49.7)Previouscoronaryintervention(%) 139(19.7)20(15.5)119(20.7)Peripheralvasculardisease(%)   13(1.8)2(1.6)11(1.9)Cerebrovasculardisease(%)70(9.9)11   (8.5)59(10.2)Chronicrenalimpairment(%)7(1.0)2(1.6)5(0.9)Currentaspirinuse(%)298(42.3)41   (31.8)257(44.6)Currentclopidogreluse(%) 86(12.2)13   (10.1)73(12.7)Currentstatinuse(%)311(44.1)48   (37.2)263(45.7)CurrentACEinhibitoruse(%)   165(23.4)27   (20.9)138(24.0)Currentbetablockeruse(%)168(23.8)22   (17.1)146(25.3)Timefromsymptomonset0–3h323(45.8)75   (58.1)248(43.1)3–6h169(24.0)29   (20.5)140(24.3)6–12h152(21.6)22   (17.1)130(22.6)>12h 61(8.7)3(2.3)58(10.1)  1044  R.Bodyetal./    Resuscitation 82 (2011) 1041–1046  Table   2 Diagnosticperformanceof    individualbiomarkersforAMI   atthetimeof    presentation.Biomarker(cut-off)Sensitivity,%   (95%CI)Specificity,%(95%CI)   PPV,%(95%CI)   NPV,%(95%CI)   LR+   (95%CI)LR  −   (95%CI)H-FABP(58ng/ml)75   (67–82)89   (86–91)60(52–68)94(92–96)6.8(5.3–8.7)0.3(0.2–0.4)CK-MB   (4.3ng/ml) 31(23–40)96   (94–97)63   (50–75)86(83–89)7.8(4.8–12.5)0.7(0.6–0.8)Myoglobin(107ng/ml) 59(50–68)82   (79–85)42   (35–50)90(87–92)3.3(2.6–4.1)0.5(0.4–0.6)cTnI   (0.055ng/ml)42   (33–51)96   (94–97)69(58–79)88   (85–90)10.1(6.5–15.6)0.6(0.5–0.7)BNP   (72ng/ml) 35(26–44)85   (82–88)34(26–43)85(82–88)2.3(1.7–3.1)0.8(0.7–0.9) d -Dimer(600ng/ml)35   (27–44)83   (79–86)31   (24–40)85(82–88)2.0   (1.5–2.7)0.8(0.7–0.9)NGAL    (201ng/ml)15(9–22)94   (91–95)34(22–48)83   (80–86)2.3(1.4–3.9)0.9(0.9–1.0)MPO   (510pM)60(51–68)58   (54–62)24   (20–29)87   (83–90)1.4(1.2–1.7)0.7(0.6–0.9) (85.8%vs.78.8%,  p <0.0001),PPV(56.4%vs.   42.5%,  p <0.0001)andNPV(95.6%vs.92.1%,  p =   0.02)forAMI.  3.1.Combinationwithclinicalriskstratificationfor    rapidexclusionofAMI  UsingBayesianprinciples,if    cTnIandH-FABPwereonlyusedtoexcludeAMIin‘verylowrisk’patientsfollowingclinicalriskstratification( n =   383,54.3%),thesensitivityof    themodelwouldincreaseto96.9%(95%CI92.3–99.2%)withaspecificityof    54.7%(50.5–58.8),PPV32.4%(27.7–37.3%)andNPV98.8%(96.8–99.7%).ThisstrategywouldhaveenabledAMI   to   beexcludedatthepointofpresentationin315(44.7%)ofpatientsatthecostof    missing4AMIs(i.e.6AMIsmissedper1000patientstreated). 4.Discussion InthisstudywehaveinvestigatedwhetheritispossibletosafelyexcludeAMIusingacombinationofbiomarkersmeasuredatthepointofpresentationin   patientswhopresentto   theEDwithpos-siblecardiacchestpain.We   comparedthevalueoftriplemarkertesting,whichmeasuresestablishedmarkersof    myocardialnecro-sisandisacommonlyusedstrategyatmanyinstitutions,tonovelmarkersofvariousprocessesin   thepathophysiologicalevolutionof  Fig.2. ROCcurvesdemonstratingtheperformanceof    biomarkers,measuredatthetime   of    presentation,forthediagnosisof    AMI.Abbreviations:AUC,areaunderthecurve. ACS.Previousstudieshavesuggestedthatsomeof    thesemarkers,particularlyBNPandMPO,couldpotentiallybeusefulaspartofamultimarkerstrategyto   enableearlyexclusionof    AMI. 22,35 NGAL hasnotpreviouslybeeninvestigatedbutpreliminarydatahadiden-tifieditasa   potentiallyvaluableearlymarkerof    AMI. 24,25 H-FABP,meanwhile,is   (likeCK-MB,myoglobinandcTnI)predominantlya   markerof    myocardialnecrosis,butavailabledatasuggestthatplasmalevelsriseveryearlyaftertheonsetofAMI,thusidentifyingitspotentialvalueintheemergencysetting. 36 OurfindingssuggestthatBNP, d -dimer,NGALandMPO   cannotbeusedfortheearlyexclusionofAMI,evenaspartofamultimarkerstrategy.However,H-FABPwas   foundto   be   superiortoallotherbiomarkersinvestigatedandhadearlydiagnosticvalueforAMI   thatwasindependentof    cTnI.Acombinationof    H-FABPandcTnIis   supe-riorto   themoreconventional‘triplemarkertesting’withCK-MB,myoglobinandcTnI.Further,combiningH-FABP,cTnIandclinicalriskstratificationtoexcludeAMI   in   lowriskpatientsyieldsa   highsensitivityandNPV.Thisstrategywouldallowjustunderhalfof    allpatientswithsuspectedcardiacchestpaintobeimmediatelyreas-suredthattheydonothaveAMI   and,in   theappropriateclinicalcontext,dischargedfromhospital.Witha   sensitivityof    96.9%itisperhapsunlikelythatthisprotocolwillbeclinicallyimplementedasthereisevidencetosuggestthata3.1%riskofmissingAMI(ifpresent)istoohighforclinicianstofeelcomfortable. 37 However,ourfindingsdohaveimplicationsforfutureresearchandforassaydevelopment.As   tro-poninassaysbecomeincreasinglymoresensitive,itis   likelythattheperformanceofourproposedalgorithmwillimprove.Indeed,ourdatawarrantfurtherexplorationwithuseof    highlysensitiveassaysthatarebecomingavailable. 38,39 Inthefuture,increasinglysensitiveH-FABPassaysmayalsoimprovetheaccuracyof    ouralgorithm. 4.1.Limitations Thisstudywasrobustlydesignedtoachieveitsobjectivesandincludeda   relativelylargenumberof    patientswithnover-ificationbiasandnolosstofollowup.Howeverourstudydoeshavelimitations.We   investigatedthediagnosticvalueof bloodsamplestakenatthetimeofinitialpresentationforfacilitatingearlydiagnosisandexclusionofAMI.Whilethisstrategywouldbeidealforenablingtimelydecisionmakingandjudicioususeofresources,serialbiomarkertestingovera   numberofhoursislikelyto   enhancethediagnosticvalueof    thebiomarkersweinvestigated.Givenour   findings,futureresearchshouldinvestigatetheaddedvalueof    serialsam-plingusingthecombinationofH-FABPandtroponin.Furtherresearchshouldalsoincorporatehighlysensitivetroponinassaysin   order   toevaluatethepotentialtoimproveearlydiagnosticvalue.Finally,oursamplesizecalculationsuggeststhat,if    ourmulti-markerpanelhadachieveda   sensitivityof100%,wewouldhaveneeded730patientsin   ordertodemonstratea   lower95%con-
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks