Approach to Acute Upper Gastrointestinal Bleeding in Adults

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 1/16

OfficialreprintfromUpToDate www.uptodate.com2015UpToDate

AuthorJohnRSaltzman,MD,FACP,FACG,FASGE,AGAF

SectionEditorMarkFeldman,MD,MACP,AGAF,FACG

DeputyEditorAnneCTravis,MD,MSc,FACG,AGAF

ThecontentontheUpToDatewebsiteisnotintendednorrecommendedasasubstituteformedicaladvice,diagnosis,ortreatment.Alwaysseektheadviceofyourownphysicianorotherqualifiedhealthcareprofessionalregardinganymedicalquestionsorconditions.TheuseofthiswebsiteisgovernedbytheUpToDateTermsofUse2015UpToDate,Inc.

Approachtoacuteuppergastrointestinalbleedinginadults

Alltopicsareupdatedasnewevidencebecomesavailableandourpeerreviewprocessiscomplete.Literaturereviewcurrentthrough:Apr2015.|Thistopiclastupdated:Jan20,2015.

INTRODUCTIONPatientswithacuteuppergastrointestinal(GI)bleedingcommonlypresentwithhematemesis(vomitingofbloodorcoffeegroundlikematerial)and/ormelena(black,tarrystools).TheinitialevaluationofpatientswithacuteupperGIbleedinginvolvesanassessmentofhemodynamicstabilityandresuscitationifnecessary.Diagnosticstudies(usuallyendoscopy)follow,withthegoalofbothdiagnosis,andwhenpossible,treatmentofthespecificdisorder.

Thediagnosticandinitialtherapeuticapproachtopatientswithclinicallysignificant(ie,thepassageofmorethanascantamountofblood)acuteupperGIbleedingwillbereviewedhere.Thisapproachisconsistentwithamultidisciplinaryinternationalconsensusstatementupdatedin2010,a2012guidelineissuedbytheAmericanSocietyforGastrointestinalEndoscopy,anda2012guidelineissuedbytheAmericanCollegeofGastroenterology[14].ThecausesofupperGIbleeding,theendoscopicmanagementofacuteupperGIbleeding,andthemanagementofactivevaricealhemorrhagearediscussedseparately.(See"Majorcausesofuppergastrointestinalbleedinginadults"and"Overviewofthetreatmentofbleedingpepticulcers"and"Generalprinciplesofthemanagementofvaricealhemorrhage"and"Methodstoachievehemostasisinpatientswithacutevaricealhemorrhage".)

Atableoutliningtheemergencymanagementofacutesevereuppergastrointestinalbleedingisprovided(table1).

INITIALEVALUATIONTheinitialevaluationofapatientwithasuspectedclinicallysignificantacuteupperGIbleedincludesahistory,physicalexamination,laboratorytests,andinsomecases,nasogastriclavage.Thegoaloftheevaluationistoassesstheseverityofthebleed,identifypotentialsourcesofthebleed,anddetermineifthereareconditionspresentthatmayaffectsubsequentmanagement.Theinformationgatheredaspartoftheinitialevaluationisusedtoguidedecisionsregardingtriage,resuscitation,empiricmedicaltherapy,anddiagnostictesting.

FactorsthatarepredictiveofableedcomingfromanupperGIsourceidentifiedinametaanalysisincludedapatientreportedhistoryofmelena(likelihoodratio[LR]5.15.9),melenicstoolonexamination(LR25),bloodorcoffeegroundsdetectedduringnasogastriclavage(LR9.6),andaratioofbloodureanitrogentoserumcreatininegreaterthan30(LR7.5)[5].Ontheotherhand,thepresenceofbloodclotsinthestoolmadeanupperGIsourcelesslikely(LR0.05).Factorsassociatedwithseverebleedingincludedredblooddetectedduringnasogastriclavage(LR3.1),tachycardia(LR4.9),orahemoglobinleveloflessthan8g/dL(LR4.56.2).

BleedingmanifestationsHematemesis(eitherredbloodorcoffeegroundemesis)suggestsbleedingproximaltotheligamentofTreitz.Thepresenceoffranklybloodyemesissuggestsmoderatetoseverebleedingthatmaybeongoing,whereascoffeegroundemesissuggestsmorelimitedbleeding.

Themajorityofmelena(black,tarrystool)originatesproximaltotheligamentofTreitz(90percent),thoughitmayalsooriginatefromthesmallbowelorrightcolon[6].Melenamaybeseenwithvariabledegreesofbloodloss,beingseenwithaslittleas50mLofblood[7].

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 2/16

Hematochezia(redormaroonbloodinthestool)isusuallyduetolowerGIbleeding.However,itcanoccurwithmassiveupperGIbleeding[8],whichistypicallyassociatedwithorthostatichypotension.(See'Physicalexamination'below.)

PastmedicalhistoryPatientsshouldbeaskedaboutpriorepisodesofupperGIbleeding,sinceupto60percentofpatientswithahistoryofanupperGIbleedarebleedingfromthesamelesion[9].Inaddition,thepatient'spastmedicalhistoryshouldbereviewedtoidentifyimportantcomorbidconditionsthatmayleadtoupperGIbleedingormayinfluencethepatient'ssubsequentmanagement.

Potentialbleedingsourcessuggestedbyapatient'spastmedicalhistoryinclude:

ComorbidillnessesmayinfluencepatientmanagementinthesettingofanacuteupperGIbleed.Comorbidillnessesmay:

MedicationhistoryAthoroughmedicationhistoryshouldbeobtained,withparticularattentionpaidtodrugsthat:

SymptomassessmentPatientsshouldbeaskedaboutsymptomsaspartoftheassessmentoftheseverityofthebleedandasapartoftheevaluationforpotentialbleedingsources.Symptomsthatsuggestthebleedingissevereincludeorthostaticdizziness,confusion,angina,severepalpitations,andcold/clammyextremities.

SpecificcausesofupperGIbleedingmaybesuggestedbythepatient'ssymptoms[6]:

Varicesorportalhypertensivegastropathyinapatientwithahistoryofliverdiseaseoralcoholabuse

Aortoentericfistulainapatientwithahistoryofanabdominalaorticaneurysmoranaorticgraft

Angiodysplasiainapatientwithrenaldisease,aorticstenosis,orhereditaryhemorrhagictelangiectasia

PepticulcerdiseaseinapatientwithahistoryofHelicobacterpylori,nonsteroidalantiinflammatorydrug(NSAIDs)use,orsmoking

Malignancyinapatientwithahistoryofsmoking,alcoholabuse,orH.pyloriinfection

Marginalulcers(ulcersatananastomoticsite)inapatientwithagastroentericanastomosis

Makepatientsmoresusceptibletohypoxemia(eg,coronaryarterydisease,pulmonarydisease).Suchpatientsmayneedtobemaintainedathigherhemoglobinlevelsthanpatientswithoutthesedisorders.(See'Bloodtransfusions'below.)

Predisposepatientstovolumeoverloadinthesettingoffluidresuscitationorbloodtransfusions(eg,renaldisease,heartfailure).Suchpatientsmayneedmoreinvasivemonitoringduringresuscitation.(See'Generalsupport'below.)

Resultinbleedingthatismoredifficulttocontrol(eg,coagulopathies,thrombocytopenia,significanthepaticdysfunction).Suchpatientsmayneedtransfusionsoffreshfrozenplasmaorplatelets.(See'Bloodtransfusions'below.)

Predisposetoaspiration(eg,dementia,hepaticencephalopathy).Endotrachealintubationshouldbeconsideredinsuchpatients.(See'Generalsupport'below.)

Predisposetopepticulcerformation,suchasaspirinandothernonsteroidalantiinflammatorydrugs(NSAIDs)(see"NSAIDs(includingaspirin):Pathogenesisofgastroduodenaltoxicity")

Areassociatedwithpillesophagitis(see"Medicationinducedesophagitis")

Promotebleeding,suchasantiplateletagents(eg,clopidogrel)andanticoagulants

Mayaltertheclinicalpresentation,suchasbismuthandiron,whichcanturnthestoolblack

Pepticulcer:Epigastricorrightupperquadrantpain

Esophagealulcer:Odynophagia,gastroesophagealreflux,dysphagia

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 3/16

PhysicalexaminationThephysicalexaminationisakeycomponentoftheassessmentofhemodynamicstability.Signsofhypovolemiainclude[6]:

Examinationofthestoolcolormayprovideacluetothelocationofthebleeding,butitisnotareliableindicator.Inaseriesof80patientswithseverehematochezia(redormaroonbloodinthestool),74percenthadacoloniclesion,11percenthadanupperGIlesion,9percenthadapresumedsmallbowelsource,andnositewasidentifiedin6percent[8].NasogastriclavagemaybecarriedoutifthereisdoubtastowhetherableedoriginatesfromtheupperGItract.(See'Nasogastriclavage'below.)

Thepresenceofabdominalpain,especiallyifsevereandassociatedwithreboundtendernessorinvoluntaryguarding,raisesconcernforperforation.Ifanysignsofanacuteabdomenarepresent,furtherevaluationtoexcludeaperforationisrequiredpriortoendoscopy.(See"Diagnosticapproachtoabdominalpaininadults",sectionon'Surgicalabdomen'.)

Finally,aswiththepastmedicalhistory,thephysicalexaminationshouldincludeasearchforevidenceofsignificantcomorbidillnesses.(See'Pastmedicalhistory'above.)

LaboratorydataLaboratoryteststhatshouldbeobtainedinpatientswithacuteuppergastrointestinalbleedingincludeacompletebloodcount,serumchemistries,livertests,andcoagulationstudies.Inaddition,serialelectrocardiogramsandcardiacenzymesmaybeindicatedinpatientswhoareatriskforamyocardialinfarction,suchasolderadults,patientswithahistoryofcoronaryarterydisease,orpatientswithsymptomssuchaschestpainordyspnea.(See"Criteriaforthediagnosisofacutemyocardialinfarction".)

TheinitialhemoglobininpatientswithacuteupperGIbleedingwilloftenbeatthepatient'sbaselinebecausethepatientislosingwholeblood.Withtime(typicallyafter24hoursormore)thehemoglobinwilldeclineasthebloodisdilutedbytheinfluxofextravascularfluidintothevascularspaceandbyfluidadministeredduringresuscitation.Itshouldbekeptinmindthatoverhydrationcanleadtoafalselylowhemoglobinvalue.Theinitialhemoglobinlevelismonitoredeverytwotoeighthours,dependingupontheseverityofthebleed.

Patientswithacutebleedingshouldhavenormocyticredbloodcells.Microcyticredbloodcellsorirondeficiencyanemiasuggestchronicbleeding.Becausebloodisabsorbedasitpassesthroughthesmallbowelandpatientsmayhavedecreasedrenalperfusion,patientswithacuteupperGIbleedingtypicallyhaveanelevatedbloodureanitrogen(BUN)tocreatinineorureatocreatinineratio(>20:1or>100:1,respectively)[10,11].Thehighertheratio,themorelikelythebleedingisfromanupperGIsource[10].

NasogastriclavageWhetherallpatientswithsuspectedacuteupperGIbleedingrequirenasogastrictube(NGT)placementiscontroversial,inpartbecausestudieshavefailedtodemonstrateabenefitwithregardtoclinicaloutcomes[12].Asanexample,aretrospectivestudylookedatwhethertherewereclinicalbenefitsfromNGTlavagein632patientsadmittedwithgastrointestinalbleeding[13].PatientswhounderwentNGTlavagewerematchedwithpatientswithsimilarcharacteristicswhodidnotundergoNGTlavage.NGTlavagewasassociatedwithashortertimetoendoscopy.However,therewerenodifferencesbetweenthosewhounderwentNGTlavageandthosewhodidnotwithregardtomortality,lengthofhospitalstay,surgery,ortransfusionrequirement.

Moreoften,NGTlavageisusedwhenitisunclearifapatienthasongoingbleedingandthusmightbenefitfrom

MalloryWeisstear:Emesis,retching,orcoughingpriortohematemesis

Varicealhemorrhageorportalhypertensivegastropathy:Jaundice,weakness,fatigue,anorexia,abdominaldistention

Malignancy:Dysphagia,earlysatiety,involuntaryweightloss,cachexia

Mildtomoderatehypovolemia:Restingtachycardia

Bloodvolumelossofatleast15percent:Orthostatichypotension(adecreaseinthesystolicbloodpressureofmorethan20mmHgand/oranincreaseinheartrateof20beatsperminutewhenmovingfromrecumbencytostanding)

Bloodvolumelossofatleast40percent:Supinehypotension

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 4/16

anearlyendoscopy.Inaddition,nasogastrictubelavagecanbeusedtoremoveparticulatematter,freshblood,andclotsfromthestomachtofacilitateendoscopy.(See"Nasogastricandnasoenterictubes",sectionon'Tubeplacement'.)

ThepresenceofredbloodorcoffeegroundmaterialintheaspiratealsoconfirmsanupperGIsourceofbleedingandpredictswhetherthebleedingiscausedbyalesionatincreasedriskforongoingorrecurrentbleeding[13,14].However,lavagemaynotbepositiveifbleedinghasceasedorarisesbeyondaclosedpylorus.ThepresenceofnonbloodybiliousfluidsuggeststhatthepylorusisopenandthatthereisnoactiveupperGIbleedingdistaltothepylorus[8].

WesuggestthatpatientsonlyundergoNGTlavageifparticulatematter,freshblood,orclotsneedtoberemovedfromthestomachtofacilitateendoscopy.

GENERALMANAGEMENT

TriageAllpatientswithhemodynamicinstability(shock,orthostatichypotension)oractivebleeding(manifestedbyhematemesis,brightredbloodpernasogastrictube,orhematochezia)shouldbeadmittedtoanintensivecareunitforresuscitationandcloseobservationwithautomatedbloodpressuremonitoring,electrocardiogrammonitoring,andpulseoximetry.

Atableoutliningtheemergencymanagementofacutesevereuppergastrointestinalbleedingisprovided(table1).

Otherpatientscanbeadmittedtoaregularmedicalward,thoughwesuggestthatalladmittedpatientswiththeexceptionoflowriskpatientsreceiveelectrocardiogrammonitoring.Outpatientmanagementmaybeappropriateforsomelowriskpatients.(See'Riskstratification'below.)

GeneralsupportPatientsshouldreceivesupplementaloxygenbynasalcannulaandshouldreceivenothingpermouth.Twolargecaliber(16gaugeorlarger)peripheralintravenouscathetersoracentralvenouslineshouldbeinsertedandplacementofapulmonaryarterycathetershouldbeconsideredinpatientswithhemodynamicinstabilityorwhoneedclosemonitoringduringresuscitation.(See"Pulmonaryarterycatheterization:Indications,contraindications,andcomplicationsinadults".)

Electiveendotrachealintubationinpatientswithongoinghematemesisoralteredrespiratoryormentalstatusmayfacilitateendoscopyanddecreasetheriskofaspiration.

FluidresuscitationAdequateresuscitationandstabilizationisessentialpriortoendoscopytominimizetreatmentassociatedcomplications[15].Patientswithactivebleedingshouldreceiveintravenousfluids(eg,500mLofnormalsalineorlactatedRinger'ssolutionover30minutes)whilebeingtypedandcrossmatchedforbloodtransfusion.Patientsatriskoffluidoverloadmayrequireintensivemonitoringwithapulmonaryarterycatheter.

Ifthebloodpressurefailstorespondtoinitialresuscitationefforts,therateoffluidadministrationshouldbeincreased.

BloodtransfusionsThedecisiontoinitiatebloodtransfusionsmustbeindividualized.Ourapproachistoinitiatebloodtransfusionsifthehemoglobinis

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 5/16

Arandomizedtrialsuggeststhatusingalowerhemoglobinthresholdforinitiatingtransfusionimprovesoutcomes.Inthetrial,921adultswithacuteupperGIbleedingwereassignedtoeitherarestrictivetransfusionstrategy(transfusiononlywhenthehemoglobinfellto

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 6/16

ofbleedingpepticulcers",sectionon'Acidsuppression'.)

Severalstudieshaveexaminedtheroleofacidsuppressiongivenbeforeorafterendoscopy(withorwithouttherapeuticintervention)[27].InthesettingofactiveupperGIbleedingfromanulcer,acidsuppressivetherapywithH2receptorantagonistshasnotbeenshowntosignificantlylowertherateofulcerrebleeding[2830].Bycontrast,highdoseantisecretorytherapywithanintravenousinfusionofaPPIsignificantlyreducestherateofrebleedingcomparedwithstandardtreatmentinpatientswithbleedingulcers[31].OralandintravenousPPItherapyalsodecreasethelengthofhospitalstay,rebleedingrate,andneedforbloodtransfusioninpatientswithhighriskulcerstreatedwithendoscopictherapy.(See"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Acidsuppression'.)

PPIsmayalsopromotehemostasisinpatientswithlesionsotherthanulcers.Thislikelyoccursbecauseneutralizationofgastricacidleadstothestabilizationofbloodclots[32].

ProkineticsBotherythromycinandmetoclopramidehavebeenstudiedinpatientswithacuteupperGIbleeding.Thegoalofusingaprokineticagentistoimprovegastricvisualizationatthetimeofendoscopybyclearingthestomachofblood,clots,andfoodresidue.Wesuggestthaterythromycinbeconsideredinpatientswhoarelikelytohavealargeamountofbloodintheirstomach,suchasthosewithseverebleeding.Areasonabledoseis3mg/kgintravenouslyover20to30minutes,30to90minutespriortoendoscopy.

Erythromycinpromotesgastricemptyingbaseduponitsabilitytobeanagonistofmotilinreceptors.Usingerythromycintoimprovegastricvisualizationhasbeenstudiedinatleastfourrandomizedcontrolledtrials[3336].Thestudiessuggestedthatasingledoseofintravenouserythromycingiven20to120minutesbeforeendoscopycansignificantlyimprovevisibility,shortenendoscopytime,andreducetheneedforsecondlookendoscopy.Treatmentappearstobesafe.

Ametaanalysisexaminedfivetrialswith316patientswhowereassignedtoerythromycin,metoclopramide,orplacebo[37].Theanalysisfoundthattheuseofaprokineticagentdecreasedtheneedforsecondlookendoscopy,butdidnotaffectthenumberofunitsofbloodtransfused,lengthofhospitalstay,orneedforsurgery.Insubgroupanalyses,erythromycincontinuedtoshowabenefitwithregardtotheneedforsecondlookendoscopy,butmetoclopramidedidnot.

Asecondmetaanalysisexaminedfourtrialswith335patientswhowereassignedtoeithererythromycinoracontrolgroup[38].Themetaanalysisfoundthatpatientswhoreceivederythromycinweresignificantlymorelikelytohaveanemptystomachatthetimeofendoscopycomparedwithpatientsinthecontrolgroup(69versus37percent).Patientstreatedwitherythromycinalsohadsignificantreductionsintheneedforsecondendoscopy,volumeofbloodtransfused,andlengthofhospitalstay.Finally,therewasatrendtowardshorterendoscopicproceduretimesanddecreasedmortalityforpatientstreatedwitherythromycin.

Erythromycinhasalsobeencomparedwithnasogastriclavage.Arandomizedtrialwith253patientsthatcomparederythromycinalonewithnasogastriclavagealoneandnasogastriclavagepluserythromycinfoundthatthequalityofvisualizationdidnotdiffersignificantlyamongthethreegroups[39].Inaddition,therewerenodifferencesamongthegroupswithregardtoprocedureduration,rebleedingrates,needforsecondendoscopy,numberoftransfusedunitsofblood,andmortality.

SomatostatinanditsanalogsSomatostatin,oritsanalogoctreotide,isusedinthetreatmentofvaricealbleedingandmayalsoreducetheriskofbleedingduetononvaricealcauses[40].Inpatientswithsuspectedvaricealbleeding,octreotideisgivenasanintravenousbolusof20to50mcg,followedbyacontinuousinfusionatarateof25to50mcgperhour.(See"Methodstoachievehemostasisinpatientswithacutevaricealhemorrhage",sectionon'Somatostatinanditsanalogs'.)

OctreotideisnotrecommendedforroutineuseinpatientswithacutenonvaricealupperGIbleeding,butitcanbeusedasadjunctivetherapyinsomecases.Itsroleisgenerallylimitedtosettingsinwhichendoscopyisunavailableorasameanstohelpstabilizepatientsbeforedefinitivetherapycanbeperformed.(See"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Somatostatinandoctreotide'.)

AntibioticsforpatientswithcirrhosisBacterialinfectionsarepresentinupto20percentofpatientswithcirrhosiswhoarehospitalizedwithgastrointestinalbleedinguptoanadditional50percentdevelopan

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 7/16

infectionwhilehospitalized.Suchpatientshaveincreasedmortality.

Multipletrialsevaluatingtheeffectivenessofprophylacticantibioticsincirrhoticpatientshospitalizedforbleedingsuggestanoverallreductionininfectiouscomplicationsandpossiblydecreasedmortality.Antibioticsmayalsoreducetheriskofrecurrentbleedinginhospitalizedpatientswhobledfromesophagealvarices.AreasonableconclusionfromthesedataisthatpatientswithcirrhosiswhopresentwithacuteupperGIbleeding(fromvaricesorothercauses)shouldbegivenprophylacticantibiotics,preferablybeforeendoscopy(althougheffectivenesshasalsobeendemonstratedwhengivenafterendoscopy).(See"Generalprinciplesofthemanagementofvaricealhemorrhage",sectionon'Infectionanduseofprophylacticantibiotics'.)

TranexamicacidTranexamicacidisanantifibrinolyticagentthathasbeenstudiedinpatientswithupperGIbleeding.AmetaanalysisthatincludedeightrandomizedtrialsoftranexamicacidinpatientswithupperGIbleedingfoundabenefitwithregardtomortalitybutnotwithregardtobleeding,surgery,ortransfusionrequirements[41].Whenonlystudiesthatusedantiulcerdrugsand/orendoscopictherapywereincluded,therewasnobeneficialeffect.ThissuggeststhatthereisnorolefortranexamicacidinthetreatmentofupperGIbleeding,sincethecurrentstandardofcareistotreatpatientswithprotonpumpinhibitorsandendoscopictherapy(ifindicated).

AnticoagulantsandantiplateletagentsWhenpossible,anticoagulantsandantiplateletagentsshouldbeheldinpatientswithupperGIbleeding.However,thethromboticriskofreversinganticoagulationshouldbeweighedagainsttheriskofcontinuedbleedingwithoutreversal,andthusthedecisiontodiscontinuemedicationsoradministerreversalagentsneedstobeindividualized.Insomecases(eg,stoppinganonsteroidalantiinflammatorydruginapatientwhoistakingitformildjointpain),thedecisiontostoptheseagentsmaybestraightforward.However,inmorecomplicatedcases,consultationwiththeproviderwhoprescribedtheanticoagulant/antiplateletmedicationshouldbeconsidered.(See"Managementofanticoagulantsinpatientsundergoingendoscopicprocedures",sectionon'Urgentproceduresinanticoagulatedpatients'and"Endoscopicproceduresinpatientswithdisordersofhemostasis"and"Managementofantiplateletagentsinpatientsundergoingendoscopicprocedures",sectionon'Urgentproceduresinpatientsonantiplateletagents'and"Correctingexcessanticoagulationafterwarfarin".)

Whentoresumethesemedicationsoncehemostasishasbeenachievedwillalsodependonthepatient'srisksforthrombosisandrecurrentbleeding.(See"Managementofanticoagulantsinpatientsundergoingendoscopicprocedures",sectionon'Resumptionofanticoagulants'and"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Riskfactorsforpersistentorrecurrentbleeding'.)

ConsultationsGastroenterologicalconsultationshouldbeobtainedinallpatientswithsuspectedclinicallysignificantacuteupperGIbleeding.Thedecisiontoobtainsurgicalandinterventionalradiologyconsultationspriortoendoscopyshouldbebaseduponthelikelihoodofpersistentorrecurrentbleeding,orrisks/complicationsstemmingfromendoscopictherapy(perforation,precipitationofmassivebleeding).

Asageneralrule,weobtainsurgicalandinterventionalradiologyconsultationifendoscopictherapyisunlikelytobesuccessful,ifthepatientisdeemedtobeathighriskforrebleedingorcomplicationsassociatedwithendoscopy,orifthereisconcernthatthepatientmayhaveanaortoentericfistula.Inaddition,asurgeonandaninterventionalradiologistshouldbepromptlynotifiedofallpatientswithsevereacuteupperGIbleeding[42].

DIAGNOSTICSTUDIESAlgorithmsprovidinganoverviewofthediagnosticapproachtopatientswithsuspecteduppergastrointestinalbleedingareprovided(algorithm1andalgorithm2).

UpperendoscopyUpperendoscopyisthediagnosticmodalityofchoiceforacuteupperGIbleeding[43,44].EndoscopyhasahighsensitivityandspecificityforlocatingandidentifyingbleedinglesionsintheupperGItract.Inaddition,onceableedinglesionhasbeenidentified,therapeuticendoscopycanachieveacutehemostasisandpreventrecurrentbleedinginmostpatients.Earlyendoscopy(within24hours)isrecommendedformostpatientswithacuteUGIbleeding,thoughwhetherearlyendoscopyaffectsoutcomesandresourceutilizationisunsettled.(See'Earlyendoscopy'belowand"Methodstoachievehemostasisinpatientswithacutevaricealhemorrhage",sectionon'Initialmanagement'and"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Endoscopictherapy'.)

EndoscopicfindingsinpatientswithpepticulcersmaybedescribedusingtheForrestclassification[45].

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 8/16

Findingsincludespurtinghemorrhage(classIa)(picture1),oozinghemorrhage(classIb),anonbleedingvisiblevessel(classIIa)(picture2),anadherentclot(classIIb)(picture3),aflatpigmentedspot(classIIc),andacleanulcerbase(classIII).Theendoscopicappearancehelpsdeterminewhichlesionsrequireendoscopictherapy.(See"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Endoscopictherapy'.)

Itmaybehelpfultoirrigatethestomachpriortoendoscopytohelpremoveresidualbloodandothergastriccontents.However,despiteirrigation,thestomachcanbeobscuredwithblood,potentiallymakingitdifficulttoestablishacleardiagnosisand/orperformtherapeuticmaneuvers.Inpatientsinwhombloodobscuresthesourceofbleeding,asecondendoscopymayberequiredtoestablishadiagnosisandtopotentiallyapplytherapy,butroutinesecondlookendoscopyisnotrecommended.(See'Nasogastriclavage'aboveand"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Secondlookendoscopy'.)

RisksofendoscopyRisksofupperendoscopyincludeaspiration,adversereactionstoconscioussedation,perforation,andincreasingbleedingwhileattemptingtherapeuticintervention.Patientsneedtobehemodynamicallystablepriortoundergoingendoscopy.

However,whilepatientsneedtobehemodynamicallystable,datasuggestthatmostpatientsdonotneedtohaveanormalhematocritinordertosafelyundergoendoscopy[46].Inaddition,endoscopyappearstobesafeinpatientswhoaremildlytomoderatelyanticoagulated[24].Inaretrospectivestudyof920patientswithupperGIbleedingundergoingupperendoscopy,patientswithlowhematocrits(30percent)withregardtocardiovascularcomplicationsandmortality[46].Inanotherretrospectivestudywith233patientswithupperGIbleedingwhoreceivedendoscopictherapy,anelevatedINRwasnotassociatedwithanincreasedriskofrebleeding,transfusionrequirement,surgery,lengthofstay,ormortality[24].TheINRwasbetween1.3and2.7in95percentofthepatients,sotheauthorscautionthattheresultsofthestudymayonlyapplytopatientswhoaremildlytomoderatelyanticoagulated.

Therisksversusbenefitsofupperendoscopyshouldbeconsideredinhighriskpatients,suchasthosewhohavehadarecentmyocardialinfarction.Inonestudy,forexample,200patientswhounderwentendoscopywithin30daysaftermyocardialinfarction(MI)werecomparedwith200controlsmatchedforage,sex,andendoscopicindication[47].Complications(includingfatalventriculartachycardia,nearrespiratoryarrest,andmildhypotension)occurredmoreofteninpatientswhohadarecentMI(8versus2percent).Complicationsoccurredmoreoften(21versus2percent)inpatientswhowereveryill(ApacheIIscore>16orhypotensionpriortoendoscopy).However,suchpatientsareatincreasedriskforcomplicationsevenwithoutendoscopyandmaybeparticularlyvulnerabletocomplicationsfromcontinuedbleedingwithoutendoscopy.(See"Predictivescoringsystemsintheintensivecareunit".)

OtherdiagnostictestsOtherdiagnostictestsforacuteupperGIbleedingincludeangiographyandataggedredbloodcellscan,whichcandetectactivebleeding[48,49].UpperGIbariumstudiesarecontraindicatedinthesettingofacuteupperGIbleedingbecausetheywillinterferewithsubsequentendoscopy,angiography,orsurgery[43].ThereisalsointerestinusingwirelesscapsuleendoscopyforpatientswhohavepresentedtotheemergencydepartmentwithsuspectedupperGIbleeding.Anesophagealcapsule(whichhasarecordingtimeof20minutes)canbegivenintheemergencydepartmentandreviewedimmediatelyforevidenceofbleeding.Confirmingthepresenceofbloodinthestomachorduodenummayaidwithpatienttriageandidentifypatientsmorelikelytobenefitfromearlyendoscopy[5053].(See"Angiographiccontrolofnonvaricealgastrointestinalbleedinginadults"and"Evaluationofobscuregastrointestinalbleeding"and"Wirelessvideocapsuleendoscopy",sectionon'Esophagealcapsuleendoscopy'.)

Acolonoscopyisgenerallyrequiredforpatientswithhematocheziaandanegativeupperendoscopyunlessanalternativesourceforthebleedinghasbeenidentified.Inaddition,patientswithmelenaandanegativeupperendoscopyfrequentlyundergocolonoscopytoruleoutarightsidedcolonicsourceforthebleeding,assuchlesionsmaypresentwithmelena.Inastudythatincluded1743colonoscopiesperformedfortheevaluationofmelenafollowinganondiagnosticupperendoscopy,asuspectedbleedingsourcewasidentifiedin5percentofpatients,aratethatwashigherthanthatseenin194,979averageriskscreeningcontrols(1percent).Despitetherelativelylowyieldinpatientswithmelena,weroutinelyperformacolonoscopyinpatientswithmelenaandanegativeupperendoscopy,aswellasinpatientswithhematochezia.(See"Approachtoacutelowergastrointestinalbleedinginadults",sectionon'Colonoscopy'.)

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=prin 9/16

RISKSTRATIFICATIONEndoscopic,clinical,andlaboratoryfeaturesmaybeusefulforriskstratificationofpatientswhopresentwithacuteupperGIbleeding(table2andpicture4)[5463],andtheuseofriskstratificationtoolsisrecommendedbytheInternationalConsensusUpperGastrointestinalBleedingConferenceGroup[2].Factorsassociatedwithrebleedingidentifiedinametaanalysisincluded[64]:

Severalinvestigatorshavedevelopeddecisionrulesandpredictivemodelsthatpermitidentificationofpatientswhoareatlowriskforrecurrentorlifethreateninghemorrhage[65].Suchpatientsmaybesuitableforearlyhospitaldischargeorevenoutpatientcare.Theeffectivenessofsuchruleshasbeenevaluatedinavarietyofclinicalsettings,withmoststudiessuggestingthatpatientsdeemedtobelowriskcansafelybedischargedearlyortreatedasoutpatients[5460,6573].Inaddition,thisapproachisassociatedwithreducedresourceutilizationcomparedwithuniversalhospitalizationofpatientswithacuteupperGIbleeding.

RiskscoresTwocommonlycitedscoringsystemsaretheRockallscoreandtheBlatchfordscore:

AIMS65isanotherscoringsystemthatusesdataavailablepriortoendoscopy.StudiessuggestithashighaccuracyforpredictinginpatientmortalityamongpatientswithupperGIbleeding[63,76].Thescorewasderivedusingdatafromadatabasethatcontainedinformationfrom187UnitedStateshospitals.Thederivationcohortuseddatafrom29,222hospitaladmissions.Thescorewasthenvalidatedusingaseparatedatasetcontaininginformationfrom32,504admissions.Thestudyfoundthatfivefactorswereassociatedwithincreasedinpatientmortality:

Hemodynamicinstability(systolicbloodpressurelessthan100mmHg,heartrategreaterthan100beatsperminute)

Hemoglobinlessthan10g/L

Activebleedingatthetimeofendoscopy

Largeulcersize(greaterthan1to3cminvariousstudies)

Ulcerlocation(posteriorduodenalbulborhighlessergastriccurvature)

TheRockallscoreisbaseduponage,thepresenceofshock,comorbidity,diagnosis,andendoscopicstigmataofrecenthemorrhage(calculator1)[54].Inonevalidationstudy,only32of744patients(4percent)whoscored2orless(outofamaximumof11)rebledandonlyonedied.

Ontheotherhand,inalaterstudyof247patientswhounderwentendoscopictherapyforbleedingpepticulcers,themodelperformedpoorlywhenpredictingrecurrentbleeding,underscoringtheneedforvalidationofthesemodels[74].

TheBlatchfordscore(alsoknownastheGlasgowBlatchfordscore),unliketheRockallscore,doesnottakeendoscopicdataintoaccountandthuscanbeusedwhenthepatientfirstpresents(calculator2)[59].Thescoreisbaseduponthebloodureanitrogen,hemoglobin,systolicbloodpressure,pulse,andthepresenceofmelena,syncope,hepaticdisease,and/orcardiacfailure.Thescorerangesfromzeroto23andtheriskofrequiringendoscopicinterventionincreaseswithincreasingscore.OnemetaanalysisfoundthataBlatchfordscoreofzerowasassociatedwithalowlikelihoodoftheneedforurgentendoscopicintervention(likelihoodratio0.02,95%confidenceinterval[CI]00.05)[5].

Asimplerversionofthescore,knownasthemodifiedGlasgowBlatchfordscore,iscalculatedusingonlythebloodureanitrogen,hemoglobin,systolicbloodpressure,andpulse.Thescorerangesfrom0to16.AprospectivestudyofthemodifiedscorefoundthatitperformedaswellasthefullBlatchfordscoreandthatitoutperformedtheRockallscorewithregardtopredictingtheneedforclinicalintervention,rebleeding,andmortality[75].

Albuminlessthan3.0g/dL(30g/L)INRgreaterthan1.5AlteredMentalstatus(Glasgowcomascorelessthan14,disorientation,lethargy,stupor,orcoma)Systolicbloodpressureof90mmHgorlessAgeolderthan65years

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 10/16

Inthevalidationcohort,themortalityrateincreasedsignificantlyasthenumberofriskfactorspresentincreased:

Inadditiontopredictingmortality,anincreasingscorewasalsoassociatedwithincreasedlengthofstay(from3.4daysforzeroriskfactorsto8.1daysforfiveriskfactors)andincreasedcost(averagecostof$5647USDwithzeroriskfactorsto$15,776USDwithfiveriskfactors).Prospectivestudiesareneededtoconfirmtheabilityofthescoretopredictmortality,lengthofstay,andcost.Inaddition,itisnotyetknownifthescorepredictsrebleedingfollowingendoscopictherapy.

EarlyendoscopyStudieshavereachedvariableconclusionswhendeterminingwhethertheapplicationofearlyendoscopyforriskstratificationandtreatmentreducesresourceutilizationoraffectspatientoutcomes[68,7780].Whereassomestudieshavedemonstratedreducedresourceutilizationandimprovedoutcomesfromearlyendoscopy[79,80],otherstudies,includingarandomizedtrial,didnot[68,77]:

ImplementationThedatapresentedabovesuggestthatriskstratificationisfeasibleandpermitsidentificationofpatientswhocanbemanagedsafelywithouthospitalization.However,forthesesystemstobesuccessful,theriskstratificationsystemmustbetieddirectlytodecisionsregardingpatientdischarge.Noneof

Zeroriskfactors:0.3percentOneriskfactor:1percentTworiskfactors:3percentThreeriskfactors:9percentFourriskfactors:15percentFiveriskfactors:25percent

Intherandomizedtrial,93outpatientswithacuteupperGIbleedingwereassignedtourgentendoscopy(beforehospitalization)orelectiveendoscopyafteradmission[68].Resultsoftheurgentendoscopyandarecommendationregardingpatientdispositionwereprovidedtotheattendingclinicianwhomadethefinaldecisionregardingpatientdisposition.

Thetimingofendoscopydidnotaffectresourceutilizationorpatientoutcomes.Lengthofstaywassimilar(fourversusfivedaysintheurgentanddelayedgroups,respectively),aswasthemeannumberofdaysintheintensivecareunit(1.2).Outpatientcarewasrecommendedfor19patients(40percent)intheurgentendoscopygroup.However,theattendingclinicianswhowereresponsibleformakingthedischargedecisionsonlyfollowedtherecommendationforoutpatientcareinfourpatients.

Thistrialsuggeststhatinorderforearlyendoscopytoreduceresourceutilization,stratificationneedstotranslateintochangesinpatientmanagement.Studiesshowingreducedutilizationhaveincorporatedprocessesbywhichpatientdispositionwaslinkeddirectlytotheriskstratificationsystem.

Abenefitforearlyendoscopy(definedasendoscopywithinonedayofadmission)wassuggestedbyalargeretrospectivestudyusingadatabaseofhospitalinpatientadmissions(NationwideInpatientSample)[80].Thestudylookedat35,747adultswithacutevaricealbleedingand435,765adultswithnonvaricealupperGIbleeding.Amongpatientswithacutevaricealhemorrhage,inpatientmortalitywas8.3percentforthosewhounderwentupperendoscopywithinonedayofadmissionandwas15.3percentforthosewhodidnot(adjustedoddsratio[OR]1.1895%CI1.081.31).ForpatientswithnonvaricealupperGIbleeding,thecorrespondingmortalityrateswere2.5and6.6percent,respectively(adjustedOR1.3295%CI1.261.38).

However,alimitationofthestudyisthatitdidnotdifferentiatepatientswhowereadmittedwithupperGIbleedingfromthosewhodevelopedupperGIbleedingwhilehospitalizedforotherreasons(mostofwhomwouldpresumablyundergoendoscopymorethanonedayfollowinghospitaladmission).Thiscouldskewtheresultstowardincreasedmortalityinthepatientswhodidnotundergoearlyendoscopysincepatientswhodevelopbleedingasinpatientsareknowntohavehighermortalityrates[81,82].

Anotherstudythatsuggestedabenefitwithregardtomortalityincluded8222patientswithupperGIbleeding[79].Patientswhodiedhadasignificantlylongerwaitingtimetoendoscopythanthosewhosurvived(1.65versus0.95daysadjustedOR1.10,95%CI1.061.14).

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 11/16

thepublishedriskscoreshasyetbeenadoptedwidely.

Asageneralrule,wedischargepatientswhomeetthefollowingcriteria:

However,thedecisiontodischargeapatientalsodependsuponindividualpatientfactors,suchasreliabilityforfollowupandconfidenceinthediagnosisinsomecases,weadmitpatientswhoappeartobelowriskforobservation.

Ifpatientsdonotmeetthesecriteriaweadmitthemtoamonitoredsettingorintensivecareunit(dependingupontheseverityofbleeding,comorbidities,andstabilityofvitalsigns).Mostpatientswhohavereceivedendoscopictreatmentforhighriskstigmatashouldbehospitalizedfor72hourstomonitorforrebleeding,sincemostrebleedingoccursduringthistime[2].

TREATMENTThetreatmentofpatientswithupperGIbleedingduetovariouscausesisdiscussedseparately.(See"Overviewofthetreatmentofbleedingpepticulcers"and"Contactthermaldevicesforthetreatmentofbleedingpepticulcers"and"Methodstoachievehemostasisinpatientswithacutevaricealhemorrhage".)

INFORMATIONFORPATIENTSUpToDateofferstwotypesofpatienteducationmaterials,"TheBasics"and"BeyondtheBasics."TheBasicspatienteducationpiecesarewritteninplainlanguage,atthe5 to6gradereadinglevel,andtheyanswerthefourorfivekeyquestionsapatientmighthaveaboutagivencondition.Thesearticlesarebestforpatientswhowantageneraloverviewandwhoprefershort,easytoreadmaterials.BeyondtheBasicspatienteducationpiecesarelonger,moresophisticated,andmoredetailed.Thesearticlesarewrittenatthe10 to12 gradereadinglevelandarebestforpatientswhowantindepthinformationandarecomfortablewithsomemedicaljargon.

Herearethepatienteducationarticlesthatarerelevanttothistopic.Weencourageyoutoprintoremailthesetopicstoyourpatients.(Youcanalsolocatepatienteducationarticlesonavarietyofsubjectsbysearchingon"patientinfo"andthekeyword(s)ofinterest.)

SUMMARYANDRECOMMENDATIONS

Havenocomorbidities

Havestablevitalsigns

Haveanormalhemoglobin

Havealikelybleedingsourceidentifiedonupperendoscopy

Haveasourceofbleedingthatisnotassociatedwithahighriskofrebleeding(eg,varicealbleeding,activebleeding,bleedingfromaDieulafoy'slesion,orulcerbleedingwithhighriskstigmata)(table2)

th th

th th

Basicstopics(see"Patientinformation:Upperendoscopy(TheBasics)"and"Patientinformation:GIbleed(TheBasics)")

BeyondtheBasicstopics(see"Patientinformation:Upperendoscopy(BeyondtheBasics)"and"Patientinformation:Pepticulcerdisease(BeyondtheBasics)")

Atableoutliningtheemergencymanagementofacutesevereuppergastrointestinalbleedingisprovided(table1).(See'Introduction'above.)

AcarefulhistoryshouldbeobtainedtoidentifypotentialsourcesoftheupperGIbleed,assesstheseverityofthebleed,andtoidentifycomorbidconditionsthatmayinfluencethepatient'ssubsequentmanagement.(See'Initialevaluation'above.)

Thephysicalexaminationshouldfocusonsignsthatindicatetheseverityofbloodloss,helplocalizethesourceofthebleeding,andsuggestcomplications.(See'Physicalexamination'above.)

Thepresenceofabdominalpain,especiallyifsevereandassociatedwithreboundtendernessorinvoluntaryguardingraisesconcernforperforation.Ifanysignsofanacuteabdomenarepresent,furtherevaluationtoexcludeaperforationisrequiredpriortoendoscopy.(See'Physicalexamination'above.)

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 12/16

Laboratorytestsobtainedinpatientswithacuteuppergastrointestinalbleedingincludeacompletebloodcount,serumchemistries,livertests,andcoagulationstudies.Inaddition,wesuggestrulingoutamyocardialinfarctioninolderadultpatientsandthosewithknowncardiovasculardiseasewhohaveseverebleeding,especiallyiftherehasbeenhemodynamicinstability.(See'Laboratorydata'above.)

WesuggestthatpatientsonlyundergoNGTlavageifparticulatematter,freshblood,andclotsneedtoberemovedfromthestomachtofacilitateendoscopy.(See'Nasogastriclavage'above.)

Patientswhorequirehospitalizationshouldbeadmittedtoamonitoredbedorintensivecareunitdependingupontheseverityofbleeding.(See'Triage'above.)

Wesuggestincorporationofavalidatedriskscoreforuppergastrointestinalbleedingintoroutineclinicalpracticetofacilitateoptimaltriagedecisions.(See'Riskscores'above.)

Generalsupportivemeasuresinclude:

Provisionofsupplementaloxygenbynasalcannula

Nothingpermouth

Twolargecaliber(16gaugeorlarger)peripheralcathetersoracentralvenousline

Placementofapulmonaryarterycathetershouldbeconsideredinpatientswithhemodynamicinstabilityorwhoneedclosemonitoringduringresuscitation

Forthemajorityofpatientswithacuteuppergastrointestinalbleedingwhodonothavesignificantcomorbidillnesses,werecommendgivingbloodtransfusionstomaintainthehemoglobinat7g/dL(70g/L)ratherthan9g/dL(90g/L)(Grade1B).However,patientswithactivebleedingandhypovolemiamayrequirebloodtransfusiondespiteanapparentlynormalhemoglobin.Forpatientsatincreasedriskofsufferingadverseeventsinthesettingofsignificantanemia,suchasthosewithunstablecoronaryarterydisease,wesuggesttransfusingtomaintainthehemoglobinat9g/dL(90g/L)ratherthan7g/dL(70g/L)(Grade2C).(See'Bloodtransfusions'aboveand"OverviewofthenonacutemanagementofunstableanginaandnonSTelevationmyocardialinfarction",sectionon'Redcelltransfusion'.).

Inpatientswithsuspectedvaricealbleeding,wesuggesttransfusingtoahemoglobinofnomorethan10g/dL(100g/L)(Grade2C).Itisparticularlyimportanttoavoidovertransfusioninpatientswithsuspectedvaricealbleeding,asitcanprecipitateworseningofthebleeding.(See'Bloodtransfusions'above.)

WesuggestthatpatientswithacuteupperGIbleedingbetreatedwithanintravenousPPIatpresentationuntilconfirmationofthecauseofbleeding,afterwhichtheneedforspecifictherapyandthedurationofPPIusecanbedetermined(Grade2B).(See'Acidsuppression'aboveand"Overviewofthetreatmentofbleedingpepticulcers",sectionon'Acidsuppression'.)

Wesuggestthaterythromycinbegivenpriortoendoscopyinpatientswhoarelikelytohavealargeamountofbloodintheirstomach,suchasthosewithseverebleeding.Areasonabledoseis3mg/kgintravenouslyover20to30minutes,30to90minutespriortoendoscopy.(See'Prokinetics'above.)

WerecommendthatpatientsknowntohavecirrhosiswhopresentwithacuteupperGIbleedingreceiveantibiotics,preferablybeforeendoscopy(Grade1A).(See"Generalprinciplesofthemanagementofvaricealhemorrhage",sectionon'Infectionanduseofprophylacticantibiotics'.)

Werecommendupperendoscopyfortheevaluationandmanagementofclinicallysignificant(ie,morethanascantamountofblood)acuteupperGIbleeding(Grade1A).Additionaldiagnostictestsmayberequiredinspecificcircumstances.Algorithmsprovidinganoverviewofthediagnosticapproachtopatientswithsuspecteduppergastrointestinalbleedingareprovided(algorithm1andalgorithm2).(See'Diagnosticstudies'above.)

ThetreatmentofpatientswithupperGIbleedingduetovariouscausesisdiscussedseparately.(See"Overviewofthetreatmentofbleedingpepticulcers"and"Contactthermaldevicesforthetreatmentofbleedingpepticulcers"and"Methodstoachievehemostasisinpatientswithacutevaricealhemorrhage".)

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 13/16

UseofUpToDateissubjecttotheSubscriptionandLicenseAgreement.

REFERENCES

1. BarkunA,BardouM,MarshallJK,NonvaricealUpperGIBleedingConsensusConferenceGroup.Consensusrecommendationsformanagingpatientswithnonvaricealuppergastrointestinalbleeding.AnnInternMed2003139:843.

2. BarkunAN,BardouM,KuipersEJ,etal.Internationalconsensusrecommendationsonthemanagementofpatientswithnonvaricealuppergastrointestinalbleeding.AnnInternMed2010152:101.

3. HwangJH,FisherDA,BenMenachemT,etal.TheroleofendoscopyinthemanagementofacutenonvaricealupperGIbleeding.GastrointestEndosc201275:1132.

4. LaineL,JensenDM.Managementofpatientswithulcerbleeding.AmJGastroenterol2012107:345.5. SrygleyFD,GerardoCJ,TranT,FisherDA.Doesthispatienthaveasevereuppergastrointestinal

bleed?JAMA2012307:1072.6. CappellMS,FriedelD.Initialmanagementofacuteuppergastrointestinalbleeding:frominitialevaluation

uptogastrointestinalendoscopy.MedClinNorthAm200892:491.7. Currentdiagnosisandtreatment:Surgery,13,DohertyG(Ed),McGrawHillCompanies,2010.p.493.8. JensenDM,MachicadoGA.Diagnosisandtreatmentofseverehematochezia.Theroleofurgent

colonoscopyafterpurge.Gastroenterology198895:1569.9. PalmerED.Thevigorousdiagnosticapproachtouppergastrointestinaltracthemorrhage.A23year

prospectivestudyof1,4000patients.JAMA1969207:1477.10. RichardsRJ,DonicaMB,GrayerD.Canthebloodureanitrogen/creatinineratiodistinguishupperfrom

lowergastrointestinalbleeding?JClinGastroenterol199012:500.11. MortensenPB,NhrM,MllerPetersenJF,BalslevI.Thediagnosticvalueofserumurea/creatinine

ratioindistinguishingbetweenupperandlowergastrointestinalbleeding.Aprospectivestudy.DanMedBull199441:237.

12. PallinDJ,SaltzmanJR.IsnasogastrictubelavageinpatientswithacuteupperGIbleedingindicatedorantiquated?GastrointestEndosc201174:981.

13. HuangES,KarsanS,KanwalF,etal.ImpactofnasogastriclavageonoutcomesinacuteGIbleeding.GastrointestEndosc201174:971.

14. AljebreenAM,FalloneCA,BarkunAN.NasogastricaspiratepredictshighriskendoscopiclesionsinpatientswithacuteupperGIbleeding.GastrointestEndosc200459:172.

15. BaradarianR,RamdhaneyS,ChapalamaduguR,etal.Earlyintensiveresuscitationofpatientswithuppergastrointestinalbleedingdecreasesmortality.AmJGastroenterol200499:619.

16. DugganJM.Gastrointestinalhemorrhage:shouldwetransfuseless?DigDisSci200954:1662.17. VillanuevaC,ColomoA,BoschA,etal.Transfusionstrategiesforacuteuppergastrointestinalbleeding.

NEnglJMed2013368:11.18. QaseemA,HumphreyLL,FittermanN,etal.Treatmentofanemiainpatientswithheartdisease:a

clinicalpracticeguidelinefromtheAmericanCollegeofPhysicians.AnnInternMed2013159:770.19. KravetzD,BoschJ,ArderiuM,etal.Hemodynamiceffectsofbloodvolumerestitutionfollowinga

hemorrhageinratswithportalhypertensionduetocirrhosisoftheliver:influenceoftheextentofportalsystemicshunting.Hepatology19899:808.

20. CerqueiraRM,AndradeL,CorreiaMR,etal.Riskfactorsforinhospitalmortalityincirrhoticpatientswithoesophagealvaricealbleeding.EurJGastroenterolHepatol201224:551.

21. KrigeJE,KotzeUK,DistillerG,etal.Predictivefactorsforrebleedinganddeathinalcoholiccirrhoticpatientswithacutevaricealbleeding:amultivariateanalysis.WorldJSurg200933:2127.

22. McCormickPA,JenkinsSA,McIntyreN,BurroughsAK.Whyportalhypertensivevaricesbleedandbleed:ahypothesis.Gut199536:100.

23. RestelliniS,KheradO,JairathV,etal.Redbloodcelltransfusionisassociatedwithincreasedrebleedinginpatientswithnonvaricealuppergastrointestinalbleeding.AlimentPharmacolTher201337:316.

24. WolfAT,WasanSK,SaltzmanJR.Impactofanticoagulationonrebleedingfollowingendoscopictherapyfornonvaricealuppergastrointestinalhemorrhage.AmJGastroenterol2007102:290.

25. MaltzGS,SiegelJE,CarsonJL.Hematologicmanagementofgastrointestinalbleeding.Gastroenterol

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 14/16

ClinNorthAm200029:169.26. ASGEStandardsofPracticeCommittee,AndersonMA,BenMenachemT,etal.Managementof

antithromboticagentsforendoscopicprocedures.GastrointestEndosc200970:1060.27. DorwardS,SreedharanA,LeontiadisGI,etal.Protonpumpinhibitortreatmentinitiatedpriorto

endoscopicdiagnosisinuppergastrointestinalbleeding.CochraneDatabaseSystRev2006:CD005415.

28. GisbertJP,GonzlezL,CalvetX,etal.ProtonpumpinhibitorsversusH2antagonists:ametaanalysisoftheirefficacyintreatingbleedingpepticulcer.AlimentPharmacolTher200115:917.

29. KavianiMJ,HashemiMR,KazemifarAR,etal.Effectoforalomeprazoleinreducingrebleedinginbleedingpepticulcers:aprospective,doubleblind,randomized,clinicaltrial.AlimentPharmacolTher200317:211.

30. LauJY,SungJJ,LeeKK,etal.Effectofintravenousomeprazoleonrecurrentbleedingafterendoscopictreatmentofbleedingpepticulcers.NEnglJMed2000343:310.

31. ChanWH,KhinLW,ChungYF,etal.Randomizedcontrolledtrialofstandardversushighdoseintravenousomeprazoleafterendoscopictherapyinhighriskpatientswithacutepepticulcerbleeding.BrJSurg201198:640.

32. GreenFWJr,KaplanMM,CurtisLE,LevinePH.Effectofacidandpepsinonbloodcoagulationandplateletaggregation.Apossiblecontributorprolongedgastroduodenalmucosalhemorrhage.Gastroenterology197874:38.

33. FrossardJL,SpahrL,QueneauPE,etal.Erythromycinintravenousbolusinfusioninacuteuppergastrointestinalbleeding:arandomized,controlled,doubleblindtrial.Gastroenterology2002123:17.

34. CoffinB,PocardM,PanisY,etal.ErythromycinimprovesthequalityofEGDinpatientswithacuteupperGIbleeding:arandomizedcontrolledstudy.GastrointestEndosc200256:174.

35. AltraifI,HandooFA,AljumahA,etal.Effectoferythromycinbeforeendoscopyinpatientspresentingwithvaricealbleeding:aprospective,randomized,doubleblind,placebocontrolledtrial.GastrointestEndosc201173:245.

36. CarbonellN,PauwelsA,SerfatyL,etal.Erythromycininfusionpriortoendoscopyforacuteuppergastrointestinalbleeding:arandomized,controlled,doubleblindtrial.AmJGastroenterol2006101:1211.

37. BarkunAN,BardouM,MartelM,etal.ProkineticsinacuteupperGIbleeding:ametaanalysis.GastrointestEndosc201072:1138.

38. BaiY,GuoJF,LiZS.Metaanalysis:erythromycinbeforeendoscopyforacuteuppergastrointestinalbleeding.AlimentPharmacolTher201134:166.

39. PateronD,VicautE,DebucE,etal.Erythromycininfusionorgastriclavageforuppergastrointestinalbleeding:amulticenterrandomizedcontrolledtrial.AnnEmergMed201157:582.

40. ImperialeTF,BirgissonS.SomatostatinoroctreotidecomparedwithH2antagonistsandplacebointhemanagementofacutenonvaricealuppergastrointestinalhemorrhage:ametaanalysis.AnnInternMed1997127:1062.

41. BennettC,KlingenbergSL,LangholzE,GluudLL.Tranexamicacidforuppergastrointestinalbleeding.CochraneDatabaseSystRev201411:CD006640.

42. KolkmanJJ,MeuwissenSG.Areviewontreatmentofbleedingpepticulcer:acollaborativetaskofgastroenterologistandsurgeon.ScandJGastroenterolSuppl1996218:16.

43. JutabhaR,JensenDM.Managementofuppergastrointestinalbleedinginthepatientwithchronicliverdisease.MedClinNorthAm199680:1035.

44. AdangRP,VismansJF,TalmonJL,etal.Appropriatenessofindicationsfordiagnosticuppergastrointestinalendoscopy:associationwithrelevantendoscopicdisease.GastrointestEndosc199542:390.

45. ForrestJA,FinlaysonND,ShearmanDJ.Endoscopyingastrointestinalbleeding.Lancet19742:394.46. BalderasV,BhoreR,LaraLF,etal.Thehematocritlevelinuppergastrointestinalhemorrhage:safetyof

endoscopyandoutcomes.AmJMed2011124:970.47. CappellMS,IacovoneFMJr.Safetyandefficacyofesophagogastroduodenoscopyaftermyocardial

infarction.AmJMed1999106:29.48. BarthKH.Radiologicalinterventioninupperandlowergastrointestinalbleeding.BaillieresClin

Gastroenterol19959:53.49. EmslieJT,ZarnegarK,SiegelME,BeartRWJr.Technetium99mlabeledredbloodcellscansinthe

investigationofgastrointestinalbleeding.DisColonRectum199639:750.

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 15/16

50. GralnekIM,ChingJY,MazaI,etal.Capsuleendoscopyinacuteuppergastrointestinalhemorrhage:aprospectivecohortstudy.Endoscopy201345:12.

51. MeltzerAC,AliMA,KresibergRB,etal.Videocapsuleendoscopyintheemergencydepartment:aprospectivestudyofacuteuppergastrointestinalhemorrhage.AnnEmergMed201361:438.

52. ChandranS,TestroA,UrquhartP,etal.RiskstratificationofupperGIbleedingwithanesophagealcapsule.GastrointestEndosc201377:891.

53. MeltzerAC,PinchbeckC,BurnettS,etal.Emergencyphysiciansaccuratelyinterpretvideocapsuleendoscopyfindingsinsuspecteduppergastrointestinalhemorrhage:avideosurvey.AcadEmergMed201320:711.

54. RockallTA,LoganRF,DevlinHB,NorthfieldTC.Selectionofpatientsforearlydischargeoroutpatientcareafteracuteuppergastrointestinalhaemorrhage.NationalAuditofAcuteUpperGastrointestinalHaemorrhage.Lancet1996347:1138.

55. CorleyDA,StefanAM,WolfM,etal.Earlyindicatorsofprognosisinuppergastrointestinalhemorrhage.AmJGastroenterol199893:336.

56. StanleyAJ,RobinsonI,ForrestEH,etal.Haemodynamicparameterspredictingvaricealhaemorrhageandsurvivalinalcoholiccirrhosis.QJM199891:19.

57. HayJA,MaldonadoL,WeingartenSR,EllrodtAG.Prospectiveevaluationofaclinicalguidelinerecommendinghospitallengthofstayinuppergastrointestinaltracthemorrhage.JAMA1997278:2151.

58. HayJA,LyubashevskyE,ElashoffJ,etal.Uppergastrointestinalhemorrhageclinicalguidelinedeterminingtheoptimalhospitallengthofstay.AmJMed1996100:313.

59. BlatchfordO,MurrayWR,BlatchfordM.Ariskscoretopredictneedfortreatmentforuppergastrointestinalhaemorrhage.Lancet2000356:1318.

60. CipollettaL,BiancoMA,RotondanoG,etal.OutpatientmanagementforlowrisknonvaricealupperGIbleeding:arandomizedcontrolledtrial.GastrointestEndosc200255:1.

61. MarmoR,KochM,CipollettaL,etal.Predictingmortalityinnonvaricealuppergastrointestinalbleeders:validationoftheItalianPNEDScoreandProspectiveComparisonwiththeRockallScore.AmJGastroenterol2010105:1284.

62. PangSH,ChingJY,LauJY,etal.ComparingtheBlatchfordandpreendoscopicRockallscoreinpredictingtheneedforendoscopictherapyinpatientswithupperGIhemorrhage.GastrointestEndosc201071:1134.

63. SaltzmanJR,TabakYP,HyettBH,etal.Asimpleriskscoreaccuratelypredictsinhospitalmortality,lengthofstay,andcostinacuteupperGIbleeding.GastrointestEndosc201174:1215.

64. GarcaIglesiasP,VilloriaA,SuarezD,etal.Metaanalysis:predictorsofrebleedingafterendoscopictreatmentforbleedingpepticulcer.AlimentPharmacolTher201134:888.

65. DasA,WongRC.PredictionofoutcomeofacuteGIhemorrhage:areviewofriskscoresandpredictivemodels.GastrointestEndosc200460:85.

66. BrulletE,CampoR,CalvetX,etal.Arandomizedstudyofthesafetyofoutpatientcareforpatientswithbleedingpepticulcertreatedbyendoscopicinjection.GastrointestEndosc200460:15.

67. GralnekIM,DulaiGS.IncrementalvalueofupperendoscopyfortriageofpatientswithacutenonvaricealupperGIhemorrhage.GastrointestEndosc200460:9.

68. BjorkmanDJ,ZamanA,FennertyMB,etal.Urgentvs.electiveendoscopyforacutenonvaricealupperGIbleeding:aneffectivenessstudy.GastrointestEndosc200460:1.

69. LongstrethGF,FeitelbergSP.Successfuloutpatientmanagementofacuteuppergastrointestinalhemorrhage:useofpracticeguidelinesinalargepatientseries.GastrointestEndosc199847:219.

70. LeeJG,TurnipseedS,RomanoPS,etal.EndoscopybasedtriagesignificantlyreduceshospitalizationratesandcostsoftreatingupperGIbleeding:arandomizedcontrolledtrial.GastrointestEndosc199950:755.

71. ImperialeTF,DominitzJA,ProvenzaleDT,etal.Predictingpooroutcomefromacuteuppergastrointestinalhemorrhage.ArchInternMed2007167:1291.

72. DasA,BenMenachemT,FarooqFT,etal.Artificialneuralnetworkasapredictiveinstrumentinpatientswithacutenonvaricealuppergastrointestinalhemorrhage.Gastroenterology2008134:65.

73. StanleyAJ,AshleyD,DaltonHR,etal.Outpatientmanagementofpatientswithlowriskuppergastrointestinalhaemorrhage:multicentrevalidationandprospectiveevaluation.Lancet2009373:42.

74. ChurchNI,DallalHJ,MassonJ,etal.ValidityoftheRockallscoringsystemafterendoscopictherapyforbleedingpepticulcer:aprospectivecohortstudy.GastrointestEndosc200663:606.

6/1/2015 Approachtoacuteuppergastrointestinalbleedinginadults

http://www.uptodate.com/contents/approachtoacuteuppergastrointestinalbleedinginadults?topicKey=GAST%2F2548&elapsedTimeMs=5&view=pri 16/16

Disclosures:JohnRSaltzman,MD,FACP,FACG,FASGE,AGAFNothingtodisclose.MarkFeldman,MD,MACP,AGAF,FACGNothingtodisclose.AnneCTravis,MD,MSc,FACG,AGAFEquityOwnership/StockOptions:Proctor&Gamble[Pepticulcerdisease,esophagealreflux(omeprazole)].Contributordisclosuresarereviewedforconflictsofinterestbytheeditorialgroup.Whenfound,theseareaddressedbyvettingthroughamultilevelreviewprocess,andthroughrequirementsforreferencestobeprovidedtosupportthecontent.AppropriatelyreferencedcontentisrequiredofallauthorsandmustconformtoUpToDatestandardsofevidence.Conflictofinterestpolicy

75. ChengDW,LuYW,TellerT,etal.AmodifiedGlasgowBlatchfordScoreimprovesriskstratificationinuppergastrointestinalbleed:aprospectivecomparisonofscoringsystems.AlimentPharmacolTher201236:782.

76. HyettBH,AbougergiMS,CharpentierJP,etal.TheAIMS65scorecomparedwiththeGlasgowBlatchfordscoreinpredictingoutcomesinupperGIbleeding.GastrointestEndosc201377:551.

77. SarinN,MongaN,AdamsPC.Timetoendoscopyandoutcomesinuppergastrointestinalbleeding.CanJGastroenterol200923:489.

78. TsoiKK,MaTK,SungJJ.Endoscopyforuppergastrointestinalbleeding:howurgentisit?NatRevGastroenterolHepatol20096:463.

79. TsoiKK,ChiuPW,ChanFK,etal.Theriskofpepticulcerbleedingmortalityinrelationtohospitaladmissiononholidays:acohortstudyon8,222casesofpepticulcerbleeding.AmJGastroenterol2012107:405.

80. WysockiJD,SrivastavS,WinsteadNS.Anationwideanalysisofriskfactorsformortalityandtimetoendoscopyinuppergastrointestinalhaemorrhage.AlimentPharmacolTher201236:30.

81. HearnshawSA,LoganRF,LoweD,etal.AcuteuppergastrointestinalbleedingintheUK:patientcharacteristics,diagnosesandoutcomesinthe2007UKaudit.Gut201160:1327.

82. MarmoR,KochM,CipollettaL,etal.PredictingmortalityinpatientswithinhospitalnonvaricealupperGIbleeding:aprospective,multicenterdatabasestudy.GastrointestEndosc201479:741.

Topic2548Version42.0

Disclosures