www .e l s e v i e r . c o m / l o c a t e / b j i d
The
Brazilian
Journal
of
INFECTIOUS
DISEASES
Original
article
Effects
of
a
bundled
Antimicrobial
Stewardship
Program
on
mortality:
a
cohort
study
Lucas
Miyake
Okumura
a,∗,
Monica
Maria
Gomes
da
Silva
b,
Izelandia
Veroneze
b aHospitalPharmacyDepartment,ClinicalHospital,UniversidadeFederaldoParaná(UFPR),Curitiba,PR,BrazilbInfectiousDiseaseControlService,ClinicalHospital,UniversidadeFederaldoParaná(UFPR),Curitiba,PR,Brazil
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received28November2014 Accepted11February2015 Availableonline17April2015
Keywords:
AntimicrobialStewardship Antimicrobial
DefinedDailyDose Cohort
a
b
s
t
r
a
c
t
Objectives:ToassessabundledAntimicrobialStewardshipProgramanditseffecton mor-tality.
Data:EightmonthsofclinicalelectronicmedicalrecordsandAntimicrobialStewardship Programregistrieswereusedassourceofdata.
Method:ThisisahistoricalcohortstudyconductedinaBrazilianUniversityHospital.Eligible patientswereadmittedtogeneralwardsorintensivecareunitsandhadanantimicrobial therapyprescribedandassessedbydifferentstrategies:BundledAntimicrobialStewardship Program(bundledinterventionconsisted ofclinicalpharmacistchartreview,discussion withmicrobiologist and infectiousdiseasephysicians, local educationand continuous follow-up)orConventionalAntimicrobialStewardshipProgram(clinicalpharmacistchart reviewanddiscussionwithinfectiousdiseasephysician).Primaryoutcomefromthisstudy was30-daymortality,whichwascomparedbetweengroups,byusingKaplan–Meiersurvival curveandlog-ranktest.OtheroutcomesincludedDefinedDailyDosesper1000patient-days andoccurrenceofresistantbacteria.
Results:From533patients, 491wereeligible forthe study,ofwhich 191patientswere includedtoAntimicrobialStewardshipProgramand300toConventionalstrategy.In gen-eral,theywerelikelytobemaleandagewassimilaringroups(58.9vs55.5years,p=0.38). Likewise,CharlsonComorbidityIndexwasnotstatisticallydifferentbetweengroups(2.6vs 2.7,p=0.2).Bloodstreamsiteinfectionswerefrequentlydiagnosedinbothgroups(30.89% vs26%,p=0.24).Otherlesscommonsitesofinfectionswerecentralnervoussystemand lungs.TheASPgrouphadhighersurvivalrates(p<0.01)andtheriskdifferencewas10.8% (95%CI:2.41–19.14).TherewerelessDefinedDailyDosesper1000patient-days(417vs557.2, p<0.05)andhigherratesofresistantbacteriaidentifiedintheASPgroup(83%vs17%). Conclusion:BundledASPwasthemosteffectivestrategy,withreducedmortalityandDefined DailyDosesper1000patient-days.
©2015ElsevierEditoraLtda.Allrightsreserved.
∗ Correspondingauthor.
E-mailaddress:[email protected](L.M.Okumura). http://dx.doi.org/10.1016/j.bjid.2015.02.005
Background
AntimicrobialStewardshipPrograms(ASP)seek tooptimize clinicaloutcomesandtoreduceunwantedeventsrelatedto inappropriateuseofantimicrobialdrugtherapy(ADT).1,2
According to the Infectious Disease Society of Amer-ica and Society for Healthcare Epidemiology of America,1
the collaboration between an Infectious Diseases Medical Doctor(IDMD),aclinicalpharmacist,andother profession-als improves patient outcomes by conducting prospective audits/feedback,formularyrestriction,localeducation, imple-mentingevidence-basedguidelines,de-escalationand esca-lation of ADT, dose optimization, and intravenous-to-oral therapyswitch.
Common outcomes used to assess the effectiveness of StewardshipProgramsincludespecificindicatorssuchasdrug waste,daysoftherapy,Defined Daily Doses(DDDs), reduc-tionofantimicrobialresistance,andrateofClostridiumdifficile infections.3,4ASP-relatedresearchhasalsoreportedpositive
effectsonhospitallengthofstayandmortality,5,6butstillthey
arescarceininternationalliteraturebecausesuchoutcomes canbeinfluencedbymultipleconfounders.
Recently, one systematic review has evaluated a single intervention–ADTde-escalating–onmortality.7Infact,such
evidencewaspoorlyplannedanditdoesnotillustratewhat StewardshipProgramsareallabout.Despitetheimportance ofADTde-escalating andits largeapplicability topreserve largespectrumantibiotics,otherinterventionssuchasswitch therapy(intravenous-to-oral)and ADTinitiationare largely performedbyStewardships.8Currentresearchhaveevidenced
thattheyplayaroleinreducingresistantbacteriaandother undesiredevents.8–11
ThereisademandtoconductASPresearchesthatillustrate what is performed during daily clinical practice. Notwith-standing,IDsocieties1,2endorsetheneedtoassesswhether
a“bundled”ASPstrategy,withasmanyaspossible interven-tions,improvespatients’outcomes.2
Objectives
Thisresearchaimedtoassesstheeffectsofabundled ASP strategyon30-daymortality.Secondaryobjectivesfromthis studyassessedbundledASPeffectsonDDDandoccurrence ofresistantbacteria.
Methods
Studydesignandsetting
ThisisaretrospectivecohortstudyconductedinaBrazilian publicuniversityhospitalwith550bedsandanaverageof60% occupation.
Thishospitalhasafiveday/weekASPandthecore mem-bersfromthisteamincludetwoIDphysicians(preceptorand IDresident)andonepharmacist(resident).
Patients
From FebruarytoSeptember2013,patientswhoweremore than18yearsoldwereincludedinthisstudywhenadmitted toadultgeneralwardorintensivecareunit(ICU).Tomeet eligi-bilitycriteria,theyalsoneededtohaveadrug-relatedproblem intheirADTprescriptionevaluatedbyaclinicalpharmacistat thefirstorseconddayofdrugtherapy.Exclusioncriteriafor thisstudywere:admissiontootherwards(cardiac,oncology andhematologywardsandotherspecialtyunits);patientsnot assessedbyclinicalpharmacist;non-acceptanceof interven-tionssuggestedbyASP.
Duringthestudyperiod,patientswereassignedtodifferent StewardshipProgramsaccording tohumanresources avail-ability.Thereby,twostrategieswereconcomitantlyperformed andpatientscouldreceiveeitheraconventionalASPora bun-dledASP.
Bundled strategy consisted in daily clinical pharmacist ADTproblemsscreeningbyusingchartreviews,labresults and electronic system review; discussion with ID MD and microbiologist(i.e.dailyvisitstolaboratorytodiscuss possi-bilitiestonarrowingorincreasingantibioticsspectrum);local education to prescribersto improve drug therapyuse; and providecontinuousfollow-up(untilclinicalresolutionor dis-charge,whenapplicable).
ConventionalstrategyconsistedofapassiveASP,whereby aclinicalpharmacistperformedthesamedrugtherapy prob-lemsscreeninganddiscussedeachcasewithIDphysicians. Wheneveraninterventionwasnecessary,aphonewasused tocommunicatewithprescribers.
Datacollection,baselinecharacteristicsandoutcomes
Onepharmacist(LMO)collecteddatafromASPinterventions registries,hospital’smedicalrecordandpharmacydispensing registries.
TocomparebaselineASPstrategies,weconsidered demo-graphicvariables(ageandsex),clinicalconditions(primary siteofinfection,Charlson’sComorbidityIndex–CCI, admis-sion to ICU or general ward)12 and days of follow-up/per
patient.
The primaryoutcome from this study was 30-day mor-tality,whichwascomparedbetweengroups.Otherassessed outcomesincluded:DDDreduction,interventionsperformed toimproveADTandoccurrenceofresistantbacteriainblood cultures.Inordertoassesssuchoutcomesthefollowing defi-nitionswereused:
• 30-daymortality:timeperiodsincepatientassessmentby conventional orbundled ASPstrategy,till discharge (sur-vive)oreventofdeath.
• DDD was expressed as DDD/thousand patient-days and was calculated according to World Health Organization criteria.13
• Resistantbacteriawerealltypesofisolatedorganismsthat had adocumented drugresistanceina sterilebiological sample(onlybloodculturewasconsidered).Wedidnot con-siderinthisoutcomedrugsensitivebacteria.
Finally,weconductedasubgroupanalysisbyincludingall patientswhohadnotacceptedinterventionbyASPtoanalyze theimpactonmortality.
Studysample
Thecalculated sample sizefor the present study was 186 patientsineachstudyarmbasedonapreviousstudy14and
considering a relative risk of 0.57 on mortality, two-sided hypothesis, 5%alphaand 80% power. Wechosethis study becauseitwasconductedinacriticalcaresetting,whichis alsopartofourinclusioncriteria.Moreover,suchastudywas alsoconductedinaBrazilianhospitalandshowedimpacton mortality.
Dataanalysisandstatisticalmethods
Descriptivestatistics were used toanalyzecontinuous and dichotomousvariables.Inferentialanalyseswereperformed tocomparepatientcharacteristics,wherebycontinuous vari-ableswereassessedwitht-test(normalcurveassumedwith Komolgorov–Smirnovtest)andproportionvariableswere ana-lyzedwithchi-squareorFisher’sexacttest,asadequate.
To assess the primary outcome, a survival analysis (non-parametric Kaplan–Meier method and log-rank test) compared the occurrence of fatal events between groups. Wealsodeterminedpointestimatevalues,suchasabsolute risk reduction (ARR), relative risk (RR), number needed to treat(NNT)and95%confidenceintervals(CI).Incidencerates (IR)of30-daymortalityperthousandpatient-dayswerealso reportedbymeans ofcomparisonto other epidemiological studies.
Atlast,as wedidnotbalance patientsbetweengroups, different baseline characteristics could influence the pri-mary outcome. Therefore, all critical variables at baseline (p<0.2)wereincludedinaCoxproportional-hazards regres-sion. Selected variables were included one-by-one in the multivariateanalysismodelandcovariateswerecontrolledby variablesthathadap-valuevariationgreaterthan0.1.
AllstatisticalanalyseswereconductedusingSPSSv.20.0 andconsideredatwo-sidedhypothesistestand5%probability ofatypeIerror.Allp-valueslowerthan0.05wereconsidered statisticallysignificant.
Ethicalissues
The local Bioethics Committee, which complies with Helsinki’sDeclaration,approvedthisstudyandthecertificate numberofanalysisisCAAE26619414.2.0000.0096.Lastly,we used the STROBE Statement to write this manuscript and provideadequatereporting.15
Results
Patientcharacteristics
During the study period, pharmacists assessed more than 3000patientsand533hadundergoneclinicalpharmacistsDRP screeningandhadADTproblems.Ofthose,191receivedthe
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Probability to sur viv e Conventional group Optimized group P<.01 Days
Fig.1–SurvivalcurvesofdifferentASPstrategies.
bundledASPStrategyand300patientsreceivedthe conven-tionalASP.Other42patientsdidnothavetheirdrugtherapy modifiedassuggestedbytheASPteam,andthereforewere analyzedseparately(Table2).
Insummary,therewerepredominantlymalepatients,who hadasimilarage(58.9vs55.5years,p=0.17)andCCIscore(2.6 vs2.7,p=0.2).Asexpected,ICUpatientshadgreaterCCIscores thanpatientsingeneralwards(Table1).
ThemostcommonsitesofinfectioninbothASPgroups were,respectively,bloodstream(30.9%vs26%,p=0.19), respi-ratory (28.3% vs38%, p<0.03), and urinary tract infections (14.1% vs6.7%, p=0.01). Other less commonsites of infec-tion were centralnervoussystemand ventilator-associated pneumonia(VAP).
Primaryoutcome:crudemortalityandgroupcomparison
Ingeneral,166(34%)patientsexperiencedafataleventand the overall 30-day mortality incidence was 1.6 deaths per 100patient-days.Therewasasignificantdifferencebetween grouprates(1.1vs1.9deaths/100patient-days,p=0.002).To calculatetheincidenceratedenominator,follow-upbetween groupsoughttobestatisticallythesame,whichwasconfirmed withanon-significantp-valueof0.23(Table1).
Survivalanalysisshowedthat30-daymortalitywaslower with bundled ASP (p<0.01) (Fig. 1). TheRR was28% lower in the bundled ASPgroup (RR=0.72, 95% CI 0.54–0.94) and we observed an absolute risk reduction of 10.7% (95% CI 2.41–19.14), which means that forevery nine patients who receiveASP,onewillbenefitfromthisintervention(NNT=9.28, 95%CI5.22–41.54)(Fig.2).
Secondaryoutcomes
TheaverageDDDper1000patient-dayswassignificantlylower inthebundledASPgroup:417(±56.1)vs557.2(±10.25),p<0.01. Therewasahigherrateofresistantbacteriainthe Bun-dledASPgroup(83%vs17%),andcommonlyisolatedbacteria wereK.pneumonia(eitheronedrugresistant,carbapenemase
Table1–Patientcharacteristics.
Characteristics BundledASP (n=191) ConventionalASP (n=300) p-value
No.ofpatients % No.ofpatients %
Age,inyears Mean(sd)
58.9(18.3) 55.5(17.3) 0.38
Sex,female 79 41.3 143 47.7 0.17*
Generalward 54 28.3 68 22.7 0.33
Intensivecareunits 137 71.7 232 77.3 0.2*
Primaryinfectionsite
Bloodstreamsite 60 31.4 78 26 0.19*
Centralnervoussystem 6 3.1 6 2 0.43
Gastrointestinal 25 13.1 44 14.7 0.62
Skinandsofttissue 14 7.3 23 7.7 0.89
Respiratorytract 54 28.3 114 38 0.03* Ventilationassociated 5 2.6 15 5 0.19* Urinarytract 27 14.2 20 6.6 0.01* CharlsonComorbidity Index Mean(sd) 2.6(2.3) 2.7(2.4) 0.2* Generalward Mean(sd) 2.6(2.4) 2.6(2.4) – Intensivecareunits
Mean(sd) 2.8(1.8) 3.0(2.3) – Daysoffollow-up Mean(sd) 25.9(32.7) 19.5(21.1) 0.23 30-Daymortality 52(27.2) 114(38) <0.01 Intensivecareunits 42(21.9) 99(33) <0.01 Generalward 10(5.2) 15(5) <0.01
Abbreviations:sd,standarddeviation;95%CI,95%confidenceinterval.
Dataaren(%)unlessotherwisestated.
∗ ConsideredtobeincludedintheCoxregression.
Table2–Effectsofnon-acceptingASPintervention.
Bundled vs conventionalASP (total=491patients) Bundled+non-accepted interventions vs observation (total=533patients) ARR 10.8%(95%CI2.4–19.1) 8.82%(95%CI0.8–16.8) RR 0.72(95%CI0.54–0.94) 0.77(95%CI0.6–0.98)
orextendedspectrumbeta-lactamasepositive),followedby methicillin-resistant negative coagulase Staphylococcus (MR NCS)(Table4).
Therewere 227accepted interventions,which consisted inchangingADT(11%),improvingdosage(39%),interruption oftreatment(28%),initiationofADT(8%), escalationor de-escalation(6%and8%,respectively)(Table2).
Lastly,inasubgroupanalysisconsideringthe42excluded patients, whohad notaccepted anintervention, wewould haveobserveda2%decreaseintheefficacyofASP(10.8%vs 8.8%)anda6%reductioninRR(RR,0.72–0.77)(Table2).
Baselinecharacteristicsandimpacton30-daymortality
Sixcovariates(sex,CCI,UTI,VAP,RT,andadmissiontoICU) wereincludedinthemultivariatemodel(seebivariateanalysis
0% 11% (26) 39% (89) 28% (63) 8% (18) 8% (18) 6% (13) 5% 10% 15% 20% 25% 30% 40% 50% 45% 35% ADT change Dosage Suspension Initiation Escalation De-escalation
Fig.2–InterventionsperformedtoimproveAntimicrobial DrugTherapy(ADT).Notes:Therewere14interventions performedbyASPafterdiscussinglaboratorypreliminary results(Table4),suchasmorphology,cultureor
biochemistryfindings,whereby8wereantimicrobial initiation,4escalations,and2de-escalations.“ADT Change”accountedforIV/POswitchandsamespectrum modification(vancomycin→daptomycin).
inTable1)toassess whethertheycould influenceprimary outcome.
Coxregressiondeterminedthattwovariablescould inde-pendentlypredicttheriskofdeath:CCIandadmissiontoICU
Table3–Independentpredictorsof30-daymortality afterCoxproportional-hazardsregression.
Variable p-value aHR 95%CI Sex 0.917 0.89 0.75–1.39 BSI 0.040 0.64 0.41–0.98 RT 0.974 1.36 0.68–1.49 UTI 0.939 1.05 0.52–1.83 VAP 0.220 0.56 0.78–2.96 CharlsonComorbidityIndex* 0.003 1.09 1.03–1.16
AdmissiontoICU* 0.005 0.93 0.89–0.98
Abbreviations:aHR,adjustedHazardRatio;BSI,BloodSiteInfections;
RT,RespiratoryTractInfection;UTI,UrinaryTractInfection;VAP, Ventilator-AssociatedPneumonia;SE,standarderror;ICU; Inten-siveCareUnit;95%CI,95%confidenceinterval.
IntensiveCareUnitswaschangedto“AdmissiontoICU”,tofacilitate reading.
∗ Covariateswithstatisticalsignificance.
(Table3)mayhaveinfluenced30-daymortalityandtheywere slightlyhigherinconventionalASP.
Discussion
Mortalityreduction
Thisstudy assessedthe effectivenessofa bundledASP by using mortalityand other relevantoutcomes such asDDD reduction,occurrenceofresistantbacteria,andinterventions performedtoimproveADT.The10%absoluteriskreduction betweengroupscanbeconsideredasanimportantclinical effect.Previousstudies14,16usedregressiontechniquesor
dif-ferentperiodstoassess theeffectsofASPimplementation, whichisdifferentfromourstudydesign;wecompared differ-entASPstrategies.
Nonetheless,we conductedaretrospectivecohort study andfutureprospectiveinvestigations–randomizedcontrolled trials–wouldbeimportanttoanswerwhetherdifferenttypes ofASPstrategiesresultindifferentoutcomes.This informa-tionshouldbeendorsedbytwootherfindingsinourstudy: • TheRR(0.72,95%CI0.54–0.94)washigherifcomparedtoa
previousresearch(RR=0.57)14andweattributethis
differ-encetosmallbutimportanteffectivenessofnon-bundled strategy.
• In multivariate analysis, CCI and admission to ICU were considered risk factors and could have influ-enced30-daymortalitybetweengroups.Propensityscores could not improve the balance between the two groups because all patients that received both strategies were included.
When we input patients who did not accept an inter-vention by bundled ASP (previously excluded from study), we observed increased mortality rates. Outcomes may be directly impacted by accepting or not ASP suggestions. Therefore, non-acceptance of an ASP intervention may be discouraged, and institutions with low rates of ASP acceptanceshoulddelineate strategiestochangethis risky behavior.
Loweruseofantimicrobialdrugtherapy
Regarding DDDreduction,A-IIlevel evidencefrom interna-tional guidelines1 indicates that ASP interventions reduce
unnecessary ADT prescription. Previously, a multicenter (9 hospitals) cohort study16 showed that 206(38%)out of542
patients received inappropriate use of ADT. Such rate of inadequate ADTreassureswhy bundledstrategymay have performedsomanyinterventionstooptimizeantibioticsuse. In our study, renal function dose adjustments (dose reduction) and drug interruption were the most prevalent interventionsandcomprisedalmost70%fromallADT opti-mizations performed by the bundled strategy. For these reasons, we believed that bundled ASP interventions were directlyresponsibleforDDDreduction.
Althoughthis scenariosuggestsless antibioticsuse and costreduction,itisimportanttorememberthatDDDdecrease neitherimpliesglobalcost-savingsnorcost-effectiveness,asit onlyaccountsforADTconsumption.Othervariablesshouldbe addressedtoassesstheimpactofASPoneconomicoutcomes –suchaslengthofstayandhumanresourcecosts.
Morebacteriaidentification:causeorconsequence?
Interestingly,bundledASPhadmorepositivebloodcultures than conventional ASP,bothingeneralwardsand ICU.We believethatthesefindingsarelikelytobecausedbybundled ASP.Inotherwords,localauditsandcontinuouseducationto physiciansmayhavepromotedbacteriaidentification.
Becausethisstudywasnotpoweredanddesignedforsuch assessment,wedidnotmakefurtherassessmentsthrough regressionanalysis,asthiscouldleadtountruestatements. However,otherhypothesesthathavebeenbroughtupfrom theseobservationsare:istherateofmicrobialidentification influencedbylocallyeducatingphysiciansandbybetter infec-tious disease management? DoesASP intervention leadto morecultures?DoesASPstimulatemoreaccurateinfection diagnosisordoesmorebacterialisolationultimatelyleadsto reducedmortality?
Resistant bacteria were more often identified in the bundledASPgroup,likeresistantKlebsiellaspp,including car-bapenemase producers, methicillin-resistant staphylococci, andresistantAcinetobacterbaumanii.Alloftheaforementioned microorganisms are associatedwithhigher deathrates17,18
and,evenso, bundledASPwas associatedwithbetter out-comes.
Limitations
Thismanuscripthasseverallimitations.Firstly,thiswasa ret-rospectivestudy;thus,datacollectionandsourceofclinical registriesarenaturaldrawbacksforthesestudiesincluding: incomplete data, censoring, and non-blinded data analy-sis. Wepreferred acohort designinstead ofaclinical trial because,inthispreliminaryassessment,wehadtoinvestigate whether different ASPstrategies couldimpact patient out-comes.Moreover,designingaproperrandomizedclinicaltrial forantimicrobialstewardshipisrathercomplicated,especially foraprospectiveauditandfeedbackstrategy;properpatient allocationconcealmentandeducation-feedbackprocessmay
Table4–Resistantbacteriaisolatedfrombloodculturesaccordingtowardsandtogroups.
Bacteria BundledASP ConventionalASP ICU(n) Non-ICU(n) ICU(n) Non-ICU(n)
K.pneumoniae(KPC+orESBL) 6 – – – NCS 5 1 – 2 A.baumanii(resistant) 5 – – – NCS(MR) 3 – 1 – S.aureus(MRSA) 3 – 1 1 P.aeruginosa 1 2 – – E.cloacae – 1 – 1 E.faecalis – 1 – 1 B.cepacia 1 – – – C.freundii – 1 – – E.faecium(VRE) 1 – – – E.aerogenes – 1 – – K.pneumoniae – 1 – – P.aeruginosa (resistant) 1 – – – S.viridans 1 – – – Total(42isolated) 27(64%) 8(19%) 2(5%) 5(12%)
Abbreviations:ICU,intensivecareunit;KPC+,Carbapenemaseresistance;ESBL,extendedspectrumbeta-lactamase;NCRMR,non-coagulase methicillin-resistantStaphylococcus;MRSA;methicillin-resistant Staphylococcusaureus;VRE, vancomycin-resistantenterococci.ResistantA. baumaniiincludesresistancetoatleastthreedrugs(cephalosporins,penicillins,aminoglycosides,orquinolones).
leadtobetterantibioticprescribingpracticesovertime.19,20
Suchmethod-relatedopportunitiesshouldbeconsideredin futurestudies.
Secondly,“carebundles”ofdifferent antimicrobial stew-ardship interventions are often implemented concurrently (i.e.,educationofprescribers,formularyrestriction, prospec-tiveaudit,and feedback).21 Therefore, consideringalsothe
existenceofinfectionpreventionbundles,researchers aim-ingtoteaseoutthecontributionofeachintervention.7may
find difficulties to isolate true clinical effects on patients’ outcomes. Nevertheless, the IDSA recommend researching pragmaticandpractice-orientedASP,1whichwasourchoice
whendesigningthepresentstudy.
Furthermore,ourstudydesigndidnotallowfor quantifi-cationofthe antimicrobialstewardshipeffectsonbacterial resistance, as it takes time for the benefits to be evi-dent. That is why the majority of published stewardship papersadoptedquasi-experimentaldesigns:typicallybefore and after implementation studies,where “treatment” allo-cation and other potential confounding factors are not controlled.1,22–24
We observed a significant difference between groups regarding the rate of respiratory tract and urinary tract infections. Since multivariate analysis indicated that both sitesofinfectionwerenotindependentlyassociatedwithan increase on 30-day mortality, we believe these differences betweengroupswereduetoimbalancedpatientsallocation. Ifononehandsuchstatisticalprocedurewouldleadto bal-anced groups, on the other hand in multivariate analysis toassess whether groupdifferences could predict primary outcome only CCI and admission to ICU played a role on 30-daymortality.Thus, futurestudies shouldstrongly con-sidercontrollingforthesetwovariables,althoughtheywere onlysignificant afteracceptinga p-value<0.2inunivariate analysis.
Finally,ourstudysettingwasauniversityhospital,where there are resident rotations, so more DRP may be found. Moreover,professionalsmaybemorewillingtoaccept inter-ventions.
One should consider the external validity of our study before implementing our results in distinct services or comparingthemwithotherinvestigations;thisstudywas con-ductedinaLatinAmericancountry.
Finalconsiderations
ThebundledASPprovedtobeaneffectivewaytoimprove antimicrobialdrugtherapy,byreducing30-daymortalityand DDD/1000patient-days.Moreover,CCIandadmissiontoICU werelikelytoincreasemortality,sopatientswiththeserisk factorsshouldreceivemoreattentioninfutureASPstudies.
Worldwide, ASP implementation has risen in response to the growing threat of antimicrobial resistance amidst the diminishing pipeline of new antibiotics. We believe that antimicrobial stewardship will continue to evolve in the upcoming decade, and among various interventions, prospective audit and feedback will probably be the most implementedstrategy,inviewofitsclearadvantages: particu-larly,lackofprescribers’oppositionandprescribingbehavior modification.25
Futureresearchesshouldfocusonevaluatingtheroleand waysto improve clinical effectivenessofdifferent bundled strategies,especiallythose interventionsthatmay increase therateofbacteriaisolation,whichtranslatesintocorrectly selectingADTtotreatspecificmicroorganisms.Thereisalso anurgentneedtostandardizeoutcomesaswellasdevelop novel study designsthat can objectively assess antimicro-bial stewardshipinterventions, despite the limitations and opportunitiesinherenttoASPsheterogeneousstructuresand process.
Funding
TheBrazilianMinistryofEducationprovidedamonthly schol-arshiptothefirstauthor(LMO).Theotherauthorshavenot receivedanyspecificgrantforthisresearch.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
Acknowledgements
We acknowledgethe contribution of all professionals that improvedpatientoutcomesbydedicatingtheirtimeto hos-pital ASP. In addition, we thank the Brazilian Journal of InfectiousDiseasedoubleblindreviewprocess,whichmade timelycontributionstoourmanuscript.
r
e
f
e
r
e
n
c
e
s
1. DellitTH,OwensRC,McGowanJEJr,etal.InfectiousDiseases SocietyofAmericaandtheSocietyforHealthcare
EpidemiologyofAmericaGuidelinesfordevelopingan institutionalprogramtoenhanceantimicrobialstewardship. ClinInfectDis.2007;44:159–77.
2. CarlingP,FungT,KillionA,TerrinN,BarzaM.Favorable impactofamultidisciplinaryantibioticmanagement programconductedduring7years.InfectControlHosp Epidemiol.2003;24:699–706.
3. LawtonRM,FridkinSK,GaynesRP,McGowanJE.Practicesto improveantimicrobialuseat47UShospitals:thestatusof the1997SHEA/IDSApositionpaperrecommendations.Infect ControlHospEpidemiol.2000;21:256–9.
4. PearSM,WilliamsonTH,BettinKM,GerdingDN,GalgianiJN. DecreaseinnosocomialClostridiumdifficile-associated
diarrheabyrestrictingclindamycinuse.AnnInternMed. 1994;120:272–7.
5. YeoCL,WuJE,ChungGWT,ChanDSG,ChenHH,HsuLY. Antimicrobialstewardshipauditingofpatientsreviewedby infectiousdiseasesphysiciansinatertiaryuniversity hospital.AntimicrobialResistInfectControl.2013;2:29. 6. RosaRG,GoldaniLZ,SantosRP.Associationbetween
adherencetoanantimicrobialstewardshipprogramand mortalityamonghospitalisedcancerpatientswithfebrile neutropaenia:aprospectivecohortstudy.BMCInfectDis. 2014;14:286.
7. SilvaBN,AndrioloRB,AtallahAN,SalomãoR.De-escalation ofantimicrobialtreatmentforadultswithsepsis,severe sepsisorsepticshock.CochraneDatabaseSystRev. 2013;3:CD007934.
8. SiegelRE,HalpernNA,AlmenoffPL,LeeA,CashinR,Greene JG.Aprospectiverandomizedstudyofinpatientantibiotics forcommunity-acquiredpneumonia.Theoptimaldurationof therapy.Chest.1996;110:965–71.
9. ShojaniaKG,YokoeD,PlattR,FiskioJ,MalufN,BatesDW. Reducingvancomycinuseutilizingacomputerguideline: resultsofarandomizedcontrolledtrial.JAmMedInform Assoc.1998;5:554–62.
10.StiefelU,PatersonDL,PultzNJ,GordonSM,AronDC,Donskey CJ.Effectoftheincreasinguseofpiperacillin/tazobactamon theincidenceofvancomycin-resistantenterococciinfour academicmedicalcenters.InfectControlHospEpidemiol. 2004;25:380–3.
11.LautenbachE,LaRosaLA,MarrAM,NachamkinI,BilkerWB, FishmanNO.Changesintheprevalenceof
vancomycin-resistantenterococciinresponseto
antimicrobialformularyinterventions:impactofprogressive restrictionsonuseofvancomycinandthird-generation cephalosporins.ClinInfectDis.2003;36:440–6.
12.CharlsonME,PompeiP,AlesKL,MacKenzieCR.Anewmethod ofclassifyingprognosticcomorbidityinlongitudinalstudies: developmentandvalidation.JChronicDis.1987;40:373–83. 13.WHOCollaboratingCentreforDrugStatisticsand
Methodology[internet].GuidelinesforATCclassificationand DDDassignment.Oslo,Norway:WHO;2013.Availablefrom: http://www.whocc.no/atcddd
14.MarraAR,deAlmeidaSM,CorreaL,etal.Theeffectof limitingantimicrobialtherapydurationonantimicrobial resistanceinthecriticalcaresetting.AmJInfectControl. 2009;37:204–9.
15.vonElmE,AltmanDG,EggerM,PocockSJ,GotzschePC, VandenbrouckeJP.TheStrengtheningtheReportingof ObservationalStudiesinEpidemiology(STROBE)statement: guidelinesforreportingobservationalstudies.AnnIntern Med.2007;147:573–7.
16.AndersonDJ,MoehringRW,SloaneR,etal.Bloodstream infectionsincommunityhospitalsinthe21stcentury:a multicentercohortstudy.PLOSONE.2014;9:e91713. 17.DattaR,HuangSS.Riskofinfectionanddeathdueto
methicillin-resistantStaphylococcusaureusinlong-term carriers.ClinInfectDis.2008;47:176–81.
18.TumbarelloM,VialeP,ViscoliC,etal.Predictorsofmortality inbloodstreaminfectionscausedbyKlebsiellapneumoniae
carbapenemase–producingK.pneumoniae:importanceof combinationtherapy.ClinInfectDis.2012;55:943–50.
19.FraserGL,StogsdillP,DickensJDJr,WennbergDE,SmithRPJr, PratoBS.Antibioticoptimization.Anevaluationofpatient safetyandeconomicoutcomes.ArchInternMed. 1997;157:1689–94.
20.SolomonDH,VanHoutenL,GlynnRJ,etal.Academic detailingtoimproveuseofbroad-spectrumantibioticsatan academicmedicalcenter.ArchInternMed.
2001;161:1897–902,
http://dx.doi.org/10.1001/archinte.161.15.1897.
21.MarwickC,DaveyP.Carebundles:theholygrailofinfectious riskmanagementinhospital.CurrOpinInfectDis.
2009;22:364–9.
22.SocietyforHealthcareEpidemiologyofAmerica,Infectious DiseasesSocietyofAmerica,PediatricInfectiousDiseases Society.Policystatementonantimicrobialstewardshipbythe SocietyforHealthcareEpidemiologyofAmerica(SHEA),the InfectiousDiseasesSocietyofAmerica(IDSA),andthe PediatricInfectiousDiseasesSociety(PIDS).InfectControl HospEpidemiol.2012;33:322–7.
23.KakiR,ElligsenM,WalkerS,SimorA,PalmayL,DanemanN. Impactofantimicrobialstewardshipincriticalcare:a systematicreview.JAntimicrobChemother.2011;66:1223–30. 24.McGowanJEJr.Antimicrobialstewardship–thestateofthe
artin2011:focusonoutcomeandmethods.InfectControl HospEpidemiol.2012;33:331–7.
25.CharaniE,EdwardsR,SevdalisN,etal.Behaviorchange strategiestoinfluenceantimicrobialprescribinginacutecare: asystematicreview.ClinInfectDis.2011;53:651–62.