Parapneumonic Effusions and Empyema Richard W - PDF document

ParapneumonicEffusionsandEmpyemaRichardW.LightDivisionofAllergy,CriticalCare,PulmonaryDisease,andCriticalCareMedicine,VanderbiltUniversity,Nashville,TennesseeParapneumoniceffusionsoccurin20to40%ofpatientswhoarehospitalizedwithpneumonia.Themortalityrateinpatientswithaparapneumoniceffusionishigherthanthatinpatientswithpneu-moniawithoutaparapneumoniceffusion.Someoftheexcessmor-talityisduetomismanagementoftheparapneumoniceffusion.Characteristicsofpatientsthatindicatethataninvasiveprocedurewillbenecessaryforitsresolutionincludethefollowing:aneffusionoccupyingmorethan50%ofthehemithoraxoronethatislocu-lated;apositiveGramstainorcultureofthepleuralfluid;andapurulentpleuralfluidthathasapHbelow7.20oraglucosebelow ReceivedinoriginalformOctober21,2005;acceptedinfinalformOctober23,2005CorrespondenceandrequestsforreprintsshouldbeaddressedtoRichardW.Light,M.D.,VanderbiltUniversityMedicalCenter,T-1218MedicalCenterNorth,Nashville,TN37232-2650.E-mail:rlight98@yahoo.com 76PROCEEDINGSOFTHEAMERICANTHORACICSOCIETYVOL32006effusionsonthebasisoftheanatomiccharacteristicsofthe”uid(A),thebacteriologyofthepleural”uid(B),andthechemistryofthepleural”uid(C).Thisclassi“cationissomewhatanalogoustotheTMN(tumor-node-metastasis)classi“cationusedtoclas-sifytumors(9).Theclassi“cationsareshowninTable1.Theanatomy(A)ofthepleuraleffusionisbasedonthesizeoftheeffusion,whetheritisfree”owing,andwhethertheparietalpleuralisthickened.Aeffusions(thosewithapoorprognosis)occupymorethan50%ofthehemithorax,areloculated,and/orareassociatedwiththickeningoftheparietalpleural.Thebacte-riology(B)oftheeffusionisbasedonwhethersmearsorculturesarepositive.Ifeitherispositive,thebacteriologyisB,whichindicatesthatinvasiveproceduresareindicated.Ifthepleural”uidconsistsofpus,thebacteriologiccategoryisB,whichisalsoanindicationfordrainage.Thechemistry(C)oftheeffusionisbasedonthepHofthepleural”uid,andapHthatislessthan7.20indicatesthatinvasiveproceduresareindicated.Ifapleural”uidpHmeasurementwithabloodgasmachineisnotavailable,analternativemeasurementisapleural”uidglucoselevelwithacutofflevelof60mg/dl.OnthebasisoftheA,B,andCclassi“cation,theeffusioniscategorized.Thecategory1effusionisasmall(thicknessondecubitus,computedtomography[CT],orultra-soundstudies)free-”owingeffusion.Becausetheeffusionissmall,nothoracentesisisperformedandthebacteriologyandchemistryofthe”uidareunknown.Theriskofapooroutcomewithacategory1effusionisverylow.Thecategory2effusionissmalltomoderateinsize(thicknessandone-halfthehemithorax)andisfree”owing.TheGramstainandcultureofthepleural”uidarenegativeandthepleural”uidpHismorethan7.20.Itisimportanttoempha-sizethatthepleural”uidpHmustbemeasuredwithabloodgasmachine.NeitherapHmeternoranindicatorstripissuf“cientlyaccurate(8).Ifthepleural”uidpHisunavailable,thepleural”uidglucoselevelmustbemorethan60mg/dl.Theriskofapooroutcomewithacategory2effusionislow.Thecategory3effusionmeetsatleastoneofthefollowingcriteria:()theeffusionoccupiesmorethanone-halfthehemi-thorax,isloculated,orisassociatedwithathickenedparietalpleura;()theGramstainorcultureispositive;or()thepleural”uidpHislessthan7.20orthepleural”uidglucoseislessthan60mg/dl.Theriskofapooroutcomewithacategory3effusionismoderate.TABLE1.CATEGORIZINGRISKFORPOOROUTCOMEINPATIENTSWITHPARAPNEUMONICEFFUSIONSANDEMPYEMA PleuralSpacePleuralFluidPleuralFluidRiskofPoorAnatomyBacteriologyChemistryCategoryOutcomeDrainage :Minimal,free-flowingandB:cultureandandC:pHunknown1VerylowNoeffusion(10mmonGramstainlateraldecubitus)results:SmalltomoderateandB:negativeandC:pH7.202LowNofree-flowingeffusioncultureand10mmandGramstainone-halfhemithorax):Large,free-flowingorB:positiveorC:pH7.203ModerateYeseffusion(one-halfcultureandhemithorax)loculatedGramstaineffusion,oreffusionwiththickenedparietalpleura:pus4HighYes ReprintedbypermissionfromReference9.Thecategory4effusionischaracterizedbypleural”uidthatconsistsofpus.Theriskofapooroutcomewithacategory4effusionishigh.OPTIONSFORMANAGEMENTOFPLEURALFLUIDThereareseveraloptionsavailableforthemanagementofthepleural”uidinpatientswithparapneumoniceffusion:thesein-cludeobservation,therapeuticthoracentesis,tubethoracostomy,intrapleuralinstillationof“brinolytics,thoracoscopywiththebreakdownofadhesionsordecortication,thoracotomywiththebreakdownofadhesionsanddecortication,andopendrainageObservationisanacceptableoptionforcategory1pleuraleffu-sionsbecausetheriskofapooroutcomewithoutdrainageisverylow(6).Inpatientswithothercategoriesofparapneumoniceffusion,observationwithoutexaminationofthepleural”uidisnotac-ceptablebecauseexaminationofthepleural”uidisnecessarytoproperlycategorizetheeffusion(9).Althoughonlyabout10%ofpatientswithparapneumoniceffusionsrequiredrainageoftheireffusion,itisimportantnottodelaydrainageinthosewhorequireitbecauseaneffusionthatisfree-”owingandeasytodraincanbecomeloculatedanddif“culttodrainoveraperiodof12to24h(5,10).TherapeuticThoracentesisTherapeuticthoracentesiswasusedforthetreatmentofparap-neumoniceffusionsasearlyasthemiddleofthe19thcentury(11).Subsequently,in1962,theAmericanThoracicSocietyrecom-mendedrepeatedthoracentesisfornontuberculousempyemasthatwereintheearlyexudativephase(2).Then,in1968,SniderandSallehrecommendedthatpatientswithempyemabeman-agedwithtwotherapeuticthoracenteses,butif”uidaccumulatedafterthattime,thentubethoracostomyshouldbeperformed(12).However,therapeuticthoracentesisforparapneumoniceffusionhasbeenanoutdatedtreatmentforthepastcoupleofRecentstudiesinarabbitmodelofempyemahaveshownthatdailytherapeuticthoracentesisstarting48hafterempyemainductionisatleastaseffectiveastubethoracostomyinitiated Light:ParapneumonicEffusionsandEmpyemaatthesametime(13).Moreover,Stormandcoworkers(14)reporteddailythoracentesiseffectedtheresolutionofempyema(purulentpleural”uidorpositivemicrobiologicalstudiesonthepleural”uid)in48of51patients(94%).Simmersandassociatestreated29patientswithparapneumoniceffusionsthatconsistedofpus,hadpositivebacteriology,orhadpositivechemistrieswithalternate-dayultrasound-guidedthoracentesesandreportedthat24patients(86%)weresuccessfullytreated(15).Thedrawbacktothisstudywasthatthepatientsunderwentanaverageof7.73.5thoracentesesandtheaveragehospitalizationwas31d(15).Therehavebeennocontrolledstudiescomparingtherapeuticthoracentesiswithsmall-tubethoracostomyinthetreatmentofpatientswithcomplicatednonloculatedparapneumoniceffusions.TubeThoracostomyThemostcommonmethodbywhichparapneumoniceffusionshavebeeninitiallydrainedforthepastseveraldecadeshasbeentubethoracostomy.Thechesttubeshouldbepositionedinadependentpartofthepleuraleffusion.Althoughrelativelylarge(28…36F)chesttubeshavebeenrecommendedbymostduetothebeliefthatsmallertubeswouldbecomeobstructedwiththethick”uid,suchlargetubesareprobablynotnecessary.Inonestudy,103patientswithempyemaweretreatedwith8-to12-Fpigtailor10-to14-FMalecotcathetersinsertedwiththeSel-dingertechniqueundereitherultrasoundorCTscan(16).Thesesmallcathetersservedasthede“nitivetreatmentin80ofthe103patients(78%)(16).Theseresultsareatleastasgoodasthosereportedinsurgicalseriesinwhichmuchlargertubeswereused(17,18).Itislikelythatthecorrectpositioningofthechesttubeismoreimportantthanitssize(1).ThesmallcatheterswereplacedusingeitherultrasoundorCTscans,whereasnoimagingwasusedtoplacethelargecatheters.Theadvantagesofthesmallertubearethatitislesspainfultothepatientandiseasiertoinsert.Successfulclosed-tubedrainageofcomplicatedparapneu-moniceffusionsisevidencedbyimprovementintheclinicalandradiologicstatuswithin24h.Ifthepatienthasnotdemonstratedsigni“cantimprovementwithin24hofinitiatingtubethoracos-tomy,eitherthepleuraldrainageisunsatisfactoryorthepatientisreceivingthewrongantibiotic.Unsatisfactorypleuraldrainagecanbeduetothetubebeinginthewronglocation,loculationofthepleural”uid,ora“brinouscoatingofthevisceralpleura,whichpreventstheunderlyinglungfromexpanding.Ifdrainageisinadequate,ultrasonographyoraCTscanshouldbeobtainedtodelineatewhichoftheabovefactorsisresponsible.IntrapleuralFibrinolyticsIfthepleural”uidbecomesloculated,drainageofaparapneu-moniceffusionisdif“cult.Morethan50yago,Tilletandassoci-ates(19)reportedthattheintrapleuralinjectionofstreptokinase(a“brinolytic)andstreptodornase(aDNase)facilitatedpleuraldrainageinpatientswithempyemas.However,theuseofin-trapleuralstreptokinaseandstreptodornasewassubsequentlylargelyabandonedbecausetheirintrapleuralinjectionwasasso-ciatedwithsystemicsideeffects,includingfebrilereactions,gen-eralmalaise,andleukocytosis(20).However,startingwiththereportofBerghandcolleaguesinthelate1970s(21),therehavebeenseveraluncontrolledstudies(22…26),eachwithmorethan20patients,thatconcludedthat“brinolyticsareusefulinthemanagementofpatientswithloculatedparapneumoniceffu-sions.Bothstreptokinase(22…25)andurokinase(22,25,26)havebeenreportedtobeeffective.Bothagentsareadministeredintrapleurallyinatotalvolumeof50to100ml.Theusualdoseofurokinaseis100,000IUandthecostofonevialofurokinasethatcontains250,000IUis$490(1).Theusualdoseforstreptokinaseis150,000IU,butitisnolongeravailableintheUnitedStates.Recently,therehavebeenseveralreports(27…30)onthein-trapleuraluseoftissueplasminogenactivator(tPA)forloculatedparapneumoniceffusions.Allofthesehavereportedpositiveresultsbutnonehavebeencontrolled.Thedosesusedhavehadawiderange,butareasonabledoseis10mg.Thecostofavialcontaining50mgtPAis$1,042.Therehavenowbeenseveralcontrolledstudiesontheuseof“brinolyticsforcomplicatedparapneumoniceffusions(31…36).The“rststudywasnotrandomizedorblindedinthatthepatientsreceivedno“brinolyticsforthe“rstpartofthestudyandthenreceivedstreptokinaseforthelatterpartofthestudy(31).Thisstudy,whichincluded52patients,concludedthattherewasnosigni“cantdifferenceintheneedformoreinvasivesur-geryorinthemortalityrateinthetwogroups(31).Inasecondstudy,24patientswererandomizedtoreceivestreptokinase250,000IU/d,orsaline”ushesascontrols,forupto3d(32).Thestreptokinasegrouphadasigni“cantlygreaterreductioninthesizeofthepleural”uidcollectionandgreaterimprovementinthechestradiograph(32).Inathirdstudy,31patientswererandomlyassignedtoreceiveeitherintrapleuralurokinaseornormalsalinefor3d(33).Pleural”uiddrainagewascompletein13(86.5%)patientsintheurokinasegroupbutinonlyfour(25%)inthecontrolgroup.However,whenurokinasewassubse-quentlyadministeredtothe12withincompletedrainageinthesalinecontrolgroup,completedrainageoftheeffusionwasob-servedinonlysixpatients(50%)(33).InafourthstudyTun-cozgurandassociates(34)randomlyassigned49patientswithparapneumonicempyematoreceiveintrapleuralurokinaseornormalsalinedailyfor5consecutivedays.Patientswhoreceivedurokinaseinthisstudyhadashortertimefordefervescence(73vs.135d,p0.01),alowerneedfordecortication(60vs.29%,p0.01),andashorterhospitalization(144vs.214d,p0.01)(34).Ina“fthstudy,Diaconandassociates(35),inasingle-center,randomized,double-blindstudy,assigned44patientstoreceivedailypleuralwasheswithstreptokinaseorsaline.After3d,therewasnosigni“cantdifferencesinthegroups,butafter7d,streptokinase-treatedpatientshadahigherclinicalsuccessrate(82vs.48%,p0.01)andfewerreferralsforsurgery(43vs.9%,pHowever,themostrecentlypublishedstudyontheuseofintrapleural“brinolyticsforthetreatmentofcomplicatedparap-neumoniceffusions,whichwasthelargestandbeststudyeverperformed,wasnegative(36).Inthismulticenter,randomized,double-blindstudy,427patientswererandomizedtoreceiveintrapleuralstreptokinaseorplacebo.Inthisstudy,therewerenosigni“cantdifferencesbetweenthegroupsintheproportionofpatientswhodiedorneededsurgery(withstreptokinase:64of206patients[31%];withplacebo:60of221[27%];relativerisk,1.14[95%con“denceinterval,0.85…1.54];p0.43),aresultthatexcludedaclinicallysigni“cantbene“tofstreptokinase.Moreover,therewasnobene“ttostreptokinaseintermsofmortality,rateofsurgery,radiographicoutcomes,orlengthofthehospitalstay(36).Theresultsofthisrecentlypublishedmulticenterstudycastdoubtontheeffectivenessofintrapleural“brinolyticsforthetherapyofcomplicatedparapneumoniceffusions.Althoughthepreviouscontrolledstudiessupportedtheiruse,onlythestudiesofDiaconandcoworkers(35)andthemulticenterstudyfromtheUnitedKingdom(36)weredoubleblind.ItshouldbenotedthatthepleuraldiseaseinthestudyofDiaconandcoworkers(35)wasprobablymoreadvancedasindicatedbyalowerpleural”uidpHandahigherincidenceofsurgeryinthecontrolgroup.Itisalsopossiblethattheadministrationofnewer“brinolyticsaloneorinconjunctionwithDNasemayfacilitatethedrainageofcomplicatedparapneumoniceffusions.Indeed,atthepresenttime,thereisanothermulticentertrialunderwayintheUnited 78PROCEEDINGSOFTHEAMERICANTHORACICSOCIETYVOL32006Kingdominwhichpatientswithcomplicatedparapneumoniceffusionsarerandomizedtosaline,10mgtPA,1mgrecombinantDNase,orthecombinationoftPAandDNasetwiceaday.Untiltheresultsofthistrialareavailable,theuseofintrapleural“brinolyticsshouldbereservedforpatientsincenterswithoutaccesstovideo-assistedthoracicsurgeryandforpatientswhoarenotsurgicalcandidates.TheoriginalarticlesonenzymaticdebridementforloculatedparapneumoniceffusionsusedVaridase,whichconsistsofa“-brinolytic(streptokinase)andaDNase(streptodornase).ItisunclearhowmuchtheDNasecontributedtotheef“cacyofthepreparation.Wehaveshownthatwhenthickempyemicmaterialfromrabbitsisincubatedwitheitherstreptokinaseorurokinase,thereisnosigni“cantliquefactionofthe”uid(20).Incontrast,whenthe”uidisincubatedwithVaridase,the”uidbecomescompletelylique“edover4h.AlthoughVaridaseispresentlynotavailableintheUnitedStates,recombinanthumanDNase(Pulmozyme;Genentech,SanFrancisco,CA)isavailable.SimpsonandcoworkershaverecentlydemonstratedthatrecombinantDNasebyitselfisveryeffectiveatreducingtheviscosityofhumanempyema”uid(37).TheusefulnessofDNasewithorwithouta“brinolyticinthetreatmentofcomplicatedparapneu-moniceffusionsorempyemaisbeingevaluatedinamulticenterstudyasoutlinedinthepreviousparagraph.ThoracoscopywithLysisofAdhesionsOneoptionforthepatientwithanincompletelydrainedparap-neumoniceffusionisthoracoscopy.AchestCTscanshouldbeobtainedbeforethoracoscopytoprovideanatomicinformationaboutthesizeandextentoftheempyemacavity(38).Withthoracoscopy,theloculationsinthepleuralspacecanbedis-rupted,thepleuralspacecanbecompletelydrained,andthechesttubecanbeoptimallyplaced(38).Inaddition,thepleuralsurfacescanbeinspectedtodeterminethenecessityforfurtherintervention,suchasdecortication.Ifatthetimeofthoracos-copy,thepatientisfoundtohaveaverythickpleuralpeelwithalargeamountofdebrisandentrapmentofthelung,thethoracoscopyincisioncanbeenlargedtoallowfordecorticationiftheprocedurecannotbeaccomplishedviathoracoscopy(38).Thoracoscopyisveryeffectiveattreatingincompletelydrainedparapneumoniceffusions.Whenfourrecentstudieswithatotalof232patientsarecombined,thoracoscopywasthede“nitiveprocedurein178ofthepatients(77%)(39…42).Theoverallmortalitywas3%,andthemediantimeforchesttubedrainagepost-procedurerangedfrom3.3to7.1d.Themedianhospitalstaypost-thoracoscopyrangedfrom5.3to12.3d(39…42).Therewasonesmallstudythatrandomized20patientswitheitheraloculatedpleuraleffusionorapleural”uidpHoflessthan7.20toreceiveeitherchesttubedrainageplusstreptokinaseorthoracoscopy(43).Inthisstudy,thoracoscopywasthede“nitiveprocedurein10of11patients(91%),whereasstreptokinasewasde“nitiveinfourofninepatients(44%)(43).Theauthorsofthisstudyconcludedthat,inpatientswithloculatedparapneumoniceffusions,aprimarytreatmentstrategyofvideo-assistedthora-coscopicsurgeryisassociatedwithahigheref“cacy,shorterhospitalduration,andlesscostthanatreatmentstrategythatusescatheter-directed“brinolytictherapy(43).However,itshouldbenotedthattherewereasmallnumberofpatientsinthisstudyandthestudywasnotblinded.Decorticationinvolvestheremovalofall“broustissuefromthevisceralpleuraandparietalpleura,andtheevacuationofallpusanddebrisfromthepleuralspace(44).Decorticationeliminatesthepleuralsepsisandallowstheunderlyinglungtoexpand.Decorticationisamajorthoracicoperation,usuallyrequiringafullthoracotomyincisionandshouldthereforenotbeperformedonpatientswhoaremarkedlydebilitated.Eventhoughdecorticationisamajorprocedure,thepost-procedurehospitalizationisnotlong.Themedianpostoperativestayreportedinonestudyof71patientswasonly7d(45).Themortalityrateinthisserieswas10%,butallthepatientswhodiedhadotherseriousmedicalproblems(45).Thetimesforchesttubedrainageandforhospitalizationareshorterafterthoracoscopythanafterthoracotomywithdecortication(46).Whenmanagingpatientswithpleuralinfectionsintheacutestages,decorticationshouldonlybeconsideredforthecontrolofpleuralinfection.Decorticationshouldnotbeperformedjusttoremovethickenedpleurabecausesuchthickeningusuallyresolvesspontaneouslyoverseveralmonths(47).If,after6mo,thepleuraremainsthickenedandthepatientspulmonaryfunc-tionissuf“cientlyreducedtolimitactivities,decorticationshouldbeconsidered.Thepulmonaryfunctionofpatientswhoundergodecorticationcanincreasesigni“cantly(48).OpenDrainageChronicdrainageofthepleuralspacecanbeachievedwithopen-drainageprocedures.Twodifferenttypesofprocedurescanbeperformed.Withthesimplestprocedure,segmentsofonetothreeribsoverlyingthelowerpartoftheempyemacavityareremovedandoneormoreshort,large-boretubesareinsertedintotheempyemacavity.Thetubesaresubsequentlyirrigateddailywithamildantisepticsolution.Thedrainagefromthetubescanbecollectedinacolostomybagplacedoverthetubes.Alternatively,theempyemacavitycanbepackedwithgauze.Thisprocedureallowsthepatienttobefreedfromhisattachmenttothesuctionsystemandprovidesmorecompletedrainage(1).Asimilarbutmorecomplicatedprocedurelinesthetractbe-tweenthepleuralspaceandthesurfaceofthechestwithaskinandmuscle”apaftertwoormoreoverlyingribsareresected.Theadvantageofthisopen-”ap(Eloesser”ap)procedureisthatitcreatesaskin-lined“stulathatprovidesdrainagewithouttubes.ItcanthereforebemoreeasilymanagedbythepatientathomeandpermitsgradualobliterationoftheempyemaspaceItisimportanttonotperformanopen-drainageproceduretooearlyinthecourseofacomplicatedparapneumoniceffusion.Ifthevisceralandparietalpleuraadjacenttotheempyemacavityhavenotbeenfusedbythein”ammatoryprocess,exposureofthepleuralspacetoatmosphericpressurewillresultinapneumo-thorax.Beforeopen-drainageprocedures,thispossibilitycanbeevaluatedbyleavingthechesttubeexposedtoatmosphericpressureforashortperiodanddeterminingradiologicallywhetherthelunghascollapsed.Thehighmortalitywithpara-pneumoniceffusionsduringWorldWarIhasbeenattributedtoperformingopen-drainageprocedurestooearly(50).Patientstreatedwithopen-drainageprocedurescanexpecttohaveanopenchestwoundforaprolongedperiod.Inoneolderseriesof33patientstreatedbyopen-drainageprocedures,themediantimeforhealingthedrainagesitewas142d(5).Withdecortication,theperiodofconvalescenceismuchshorter(45),butdecorticationisamajorsurgicalprocedurethatcannotbetoleratedbymarkedlydebilitatedpatients.RECOMMENDEDMANAGEMENTOFPARAPNEUMONICEFFUSIONSItisrecommendedthatastepwiseapproachbetakenwithpa-tientswithparapneumoniceffusions.Thetreatmentoptionsaretherapeuticthoracentesis,tubethoracostomy,tubethoracostomywithintrapleural“brinolytics,thoracoscopy,andthoracotomy. Light:ParapneumonicEffusionsandEmpyemaThede“nitivetreatmentshouldbeperformedwithinthe“rst10dofhospitalization.Whenapatientwithpneumoniaisinitiallyevaluated,oneshouldaskifthepatienthasaparapneumoniceffusion.Thispossibilityshouldbeevaluatedwithdecubitusradiographsorultrasoundifthediaphragmsarenotvisiblethroughouttheirentirelengthonthelateralradiographsorifitappearsthereisloculatedpleural”uid.Ifpleural”uidispresentanditsthicknessbetweentheinsideofthechestwallandtheoutsideofthelungismorethan10mm,the”uidshouldbeanalyzedwithinashorttimeperiod.Thereasonforsamplingthepleural”uidinthesesituationsistodeterminewhetheranypoorprognosticfactorsarepresent(Table2).Thepresenceofpoorprognosticfactorsindicatesahigherlikelihoodfortheneedofaggressivedrainage.Ifthereisdoubtastohowmuchofthedensityinahemithoraxisparenchymalandhowmuchispleural,aCTscanofthechestshouldbeobtained.Ifapatienthassuf“cientpleural”uidtowarrantathoracente-sis,atherapeuticratherthanadiagnosticthoracentesisisrecom-mended(1).Thereasoningbehindthisrecommendationisasfollows.Ifno”uidreaccumulatesaftertheinitialtherapeuticthoracentesis,oneneednotworryabouttheparapneumoniceffusion.Ifthepleural”uidreaccumulatesandtherewerenopoorprognosticfactorsatthetimeoftheinitialthoracentesis,noadditionaltherapyisindicatedaslongasthepatientisdoingwell.Ifthe”uidreaccumulatesandtherewerepoorprognosticfactorspresentatthetimeoftheinitialthoracentesis,asecondtherapeuticthoracentesisshouldbeperformed.Ifthe”uidreac-cumulatesasecondtime,atubethoracostomyshouldbeper-formedifanyofthepoorprognosticfactorswerepresentatthetimeofthesecondtherapeuticthoracentesis(1).Performanceofthetherapeuticthoracentesiswillalsodelin-eatewhetherthepleural”uidisloculated.Ifthepleural”uidisloculated,andifanyoftheotherpoorprognosticfactorslistedinTable2arepresent,thenmoreaggressivetherapyisindicated.Thetwoprimaryoptionsatthistimearetubethoracostomywiththeinstillationof“brinolyticsorthoracoscopywiththelysisofadhesionsandanattemptatdecorticationifthelungdoesnotexpand.Thechoicebetweenthesetwoisdictatedsomewhatbylocalcircumstances.Ifthoracoscopyisunavailable,theobvi-ouschoiceis“brinolytics.Ifbothareavailable,onemaywanttotrytubethoracostomywith“brinolyticsinitially.However,ifcompletedrainageisnotobtainedwithoneortwoadministra-tionsofthe“brinolytics,oneshouldmovetothoracoscopy.If,withthoracoscopy,thelungdoesnotreexpandcompletely,thendecorticationshouldbeperformedwithoutdelay.Itshouldbenoted,however,assurgeonsbecomemoreadeptatthoracos-copy,asmallerfractionofpatientssubjectedtothoracoscopyrequiredecortication.InarecentstudyfromtheUnitedKingdom,WallerandRengarajanreportedthattheysuccessfullyper-formeddecorticationviathoracoscopyin21of36patients(58%)TABLE2.FACTORSASSOCIATEDWITHPOORPROGNOSISINPATIENTSWITHPARAPNEUMONICEFFUSION PleuralfluidispusPleuralfluidbacterialsmearsarepositivePleuralfluidglucoseislessthan60ml/dlPleuralfluidbacterialculturesarepositivePleuralfluidpHislessthan7.20PleuralfluidLDHismorethanthreetimestheupperlimitofnormalPleuralfluidisloculated Definitionofabbreviation:LDHlacticaciddehydrogenase.Factorsarelistedindecreasingorderofimportance.(51).Open-drainageproceduresarereservedforthosepatientswhoaretooilltoundergothoracoscopyorthoracotomy.ConflictofInterestStatementR.W.L.doesnothaveafinancialrelationshipwithacommercialentitythathasaninterestinthesubjectofthismanuscript.1.LightRW.Pleuraldiseases,4thed.Baltimore:Lippincott,WilliamsandWilkins;2001.2.AndrewsNC,ParkerEF,ShawRR,WilsonNJ,WebbWR.Managementofnontuberculousempyema.AmRevRespirDis3.HasleyPB,AlbaumMN,LiY-H,FuhrmanCR,BrittonCA,MarrieTJ,SingerDE,ColeyCM,KapoorWN,FineMJI.Dopulmonaryradio-graphic“ndingsatpresentationpredictmortalityinpatientswithcom-munity-acquiredpneumonia?ArchInternMed4.FineMJ,AubleTE,YealyDM,HanusaBH,WeissfoldLA,SingerDE,ColeyCM,MarrieTJ,KapoorWN.Apredictionruletoidentifylow-riskpatientswithcommunity-acquiredpneumonia.5.BartlettJG,FinegoldSM.AnaerobicinfectionsofthelungandpleuralAmRevRespirDis6.LightRW,GirardWM,JenkinsonSG,GeorgeRB.ParapneumonicAmJMed7.LightRW,MacGregorMI,BallWCJr,LuchsingerPC.Diagnosticsig-ni“canceofpleural”uidpHandPco8.ChengD-S,RodriguezRM,RogersJ,WagsterM,StarnesDL,LightRW.Comparisonofpleural”uidpHvaluesobtainedusingbloodgasma-chine,pHmeter,andpHindicatorstrip.9.ColiceGL,CurtisA,DeslauriersJ,HeffnerJ,LightRW,LittenbergB,SahnS,WeinsteinRA,YusenRD.Medicalandsurgicaltreatmentofparapneumoniceffusions:anevidence-basedguideline.10.ChamCW,HaqSM,RahamimJ.Empyemathoracis:aproblemwithlatereferral?11.BowditchHI.Paracentesisthoracic:ananalysisof25casesofpleuriticAmMedMonthly12.SniderGL,SalehSS.Empyemaofthethoraxinadults:reviewof10513.SasseS,NguyenT,TeixeiraLR,LightRW.Theutilityofdailytherapeuticthoracentesisforthetreatmentofearlyempyema.14.StormHKR,KrasnikM,BangK,Frimodt-MollerN.Treatmentofpleuralempyemasecondarytopneumonia:thoracocentesisregimenversustubedrainage.15.SimmersTA,JieC,SieB.MinimallyinvasivetreatmentofthoracicThoracCardiovascSurg16.ShankarS,GulatiM,KangM,GuptaS,SuriS.Image-guidedpercutane-ousdrainageofthoracicempyema:cansonographypredicttheout-EurRadiol17.AliI,UnruhH.Managementofempyemathoracis.AnnThoracSurg18.AshbaughDG.Empyemathoracis:factorsin”uencingmorbidityand19.TillettWS,SherryS,ReadCT.Theuseofstreptokinase-streptodornaseinthetreatmentofpostpneumonicempyema.JThoracSurg20.LightRW,NguyenT,MulliganME,SasseSA.Theinvitroef“cacyofvaridaseversusstreptokinaseorurokinaseforliquefyingthickpurulentexudativematerialfromloculatedempyema.Lung2000;178:13…18.21.BerghNP,EkrothR,LarssonS,NagyP.Intrapleuralstreptokinaseinthetreatmentofhaemothoraxandempyema.ScandJThoracCardio-vascSurg22.BourosD,SchizaS,PatsourakisG,ChalkiadakisG,PanagouP,SiafakasNM.Intrapleuralstreptokinaseversusurokinaseinthetreatmentofcomplicatedparapneumoniceffusions:aprospective,double-blindAmJRespirCritCareMed23.Jerjes-SanchezC,Ramirez-RiveraA,ElizaldeJJ,DelgadoR,CiceroR,Ibarra-PerezC,ArroligaAC,PaduaA,PortalesA,VillarrealA,etal.Intrapleural“brinolysiswithstreptokinaseasanadjunctivetreat-mentinhemothoraxandempyema:amulticentertrial.24.LaisaarT,PuttseppE,LaisaarV.Earlyadministrationofintrapleuralstreptokinaseinthetreatmentofmultiloculatedpleuraleffusionsandpleuralempyemas.ThoracCardiovascSurg 80PROCEEDINGSOFTHEAMERICANTHORACICSOCIETYVOL3200625.TemesRT,FollisF,KesslerRM,PettSBJr,WernlyJA.Intrapleural“brinolyticsinmanagementofempyemathoracis.26.MoultonJS,MoorePT,MenciniRA.Treatmentofloculatedpleuraleffusionswithtranscatheterintracavitaryurokinase.AJRAmJRoent-27.BishopNB,PonS,UshayHM,GreenwaldBM.Alteplaseinthetreat-mentofcomplicatedparapneumoniceffusion:acasereport.28.SkeeteDA,RutherfordEJ,SchlidtSA,AbramsJE,ParkerLA,RichPB.Intrapleuraltissueplasminogenactivatorforcomplicatedpleuraleffu-JTrauma29.WeinsteinM,RestrepoR,ChaitPG,ConnollyB,TempleM,MacarthurC.Effectivenessandsafetyoftissueplasminogenactivatorinthemanagementofcomplicatedparapneumoniceffusions.30.RayTL,BerkenboschJW,RussoP,TobiasJD.TissueplasminogenactivatorasanadjuvanttherapyforpleuralempyemainpediatricJIntensiveCareMed31.ChinNK,LimTK.Controlledtrialofintrapleuralstreptokinaseinthetreatmentofpleuralempyemaandcomplicatedparapneumoniceffu-32.DaviesRJO,TraillZC,GleesonFV.Randomisedcontrolledtrialofintrapleuralstreptokinaseincommunityacquiredpleuralinfection.33.BourosD,SchizaS,TzanakisN,ChalkiadakisG,DrositisJ,SiafakasN.Intrapleuralurokinaseversusnormalsalineinthetreatmentofcompli-catedparapneumoniceffusionsandempyema:arandomized,double-blindstudy.AmJRespirCritCareMed34.TuncozgurB,UstunsoyH,SivrikozMC,DikensoyO,TopalM,SanliM,ElbeyliL.Intrapleuralurokinaseinthemanagementofparapneu-monicempyema:arandomisedcontrolledtrial.IntJClinPract35.DiaconAH,TheronJ,SchuurmansMM,VanDeWalBW,BolligerCT.Intrapleuralstreptokinaseforempyemaandcomplicatedparapneu-moniceffusions.AmJRespirCritCareMed36.MaskellNA,DaviesCW,NunnAJ,HedleyEL,GleesonFV,MillerR,GabeR,ReesGL,PetoTE,WoodheadMA,etal.;FirstMulticenterIntrapleuralSepsisTrial(MIST1)Group.U.K.controlledtrialofintrapleuralstreptokinaseforpleuralinfection.NEnglJMed37.SimpsonG,RoomesD,HeronM.Effectsofstreptokinaseanddeoxyribo-nucleaseonviscosityofhumansurgicalandempyemapus.38.SilenML,NaunheimKS.Thoracoscopicapproachtothemanagementofempyemathoracis:indicationsandresults.ChestSurgClinNAm39.LandreneauRJ,KeenanRJ,HazelriggSR,MackMJ,NaunheinKS.Thoracoscopyforempyemaandhemothorax.40.CassinaPC,HauserM,HillejanL,GreschuchnaD,StamatisG.Video-assistedthoracoscopyinthetreatmentofpleuralempyema:stage-basedmanagementandoutcome.JThoracCardiovascSurg41.LawrenceDR,OhriSK,MoxonRE,TownsendER,FountainSW.Thora-coscopicdebridementofempyemathoracis.AnnThoracSurg42.StriffelerH,GuggerM,ImHofV,CernyA,FurrerM,RisHB.Video-assistedthoracoscopicsurgeryfor“brinopurulentpleuralempyemain67patients.AnnThoracSurg43.WaitMA,SharmaS,HohnJ,DalNogareA.Arandomizedtrialofempyematherapy.44.ThurerRJ.Decorticationinthoracicempyema:indicationsandsurgicalChestSurgClinNAm45.PothulaV,KrellensteinDJ.Earlyaggressivesurgicalmanagementofparapneumonicempyemas.46.AngelilloMackinlayTA,LyonsGA,ChimondeguyDJ,PiedrasMA,AngaramoG,EmeryJ.VATSdebridementversusthoracotomyinthetreatmentofloculatedpostpneumoniaempyema.AnnThoracSurg47.NeffCC,vanSonnenbergE,LawsonDW,PattonAS.CTfollow-upofempyemas:pleuralpeelsresolveafterpercutaneouscatheterdrainage.48.RzymanW,SkokowskiJ,RomanowiczG,LassP,DziadziuszkoR.De-corticationinchronicpleuralempyema:effectonlungfunction.JCardiothoracSurg49.DeslauriersJ,JacquesLF,GregoireJ.RoleofEloesser”apandthoraco-plastyinthethirdmillennium.ChestSurgClinNAm50.EmpyemaCommission.CasesofempyemaatCampLee,Virginia.51.WallerDA,RengarajanA.Thoracoscopicdecortication:aroleforvideo-assistedsurgeryinchronicpostpneumonicpleuralempyema.ThoracSurg