Antimicrobial susceptibility testing and antibiotic consumption results from 16 hospitals in Viet Nam: The VINARES project 2012 – 2013
Vu Tien Viet Dung
a,*, Do Thi Thuy Nga
a, Ulf Rydell
c, Lennart E. Nilsson
c,d, Linus Olson
g,h, Mattias Larsson
a,g,h, Håkan Hanberger
c,d,h, Marc Choisy
a,e,i,j, Dao Tuyet Trinh
b,
H. Rogier van Doorn
a, Nguyen Van Kinh
b, Nguyen Vu Trung
b, Heiman F.L. Wertheim
a,f, on behalf of the VINARES consortium
1aOxfordUniversityClinicalResearchUnit,VietNam
bNationalHospitalforTropicalDiseases,Hanoi,VietNam
cLinköpingUniversity,Linköping,Sweden
dDepartmentofInfectiousDiseases,InstitutionofClinicalandExperimentalMedicine,FacultyofMedicine,LinköpingUniversity,Sweden
eMIVEGEC,FrenchNationalResearchInstituteforSustainableDevelopment,Montpellier,France
fDepartmentofMedicalMicrobiologyandRadboudumcCenterforInfectiousDiseases,Radboudumc,Nijmegen,Netherlands
gDepartmentofPublicHealthSciences,KarolinskaInstitutet,Sweden
hTheTrainingandResearchAcademicCollaborationSweden-VietNam,Hanoi,VietNam
iTheFrenchNationalCentreforScientificResearch,Montpellier,France
jUniversityofMontpellier,Montpellier,France
ARTICLE INFO
Articlehistory:
Received2January2019
Receivedinrevisedform27May2019 Accepted4June2019
Availableonline12June2019
Keywords:
Antimicrobialresistance Surveillance
Defineddailydose(DDD)/1000patient-days Hospital-basedsurveillance
VietNam
ABSTRACT
Objective:Toestablishahospital-basedsurveillancenetworkwithnationalcoverageforantimicrobial resistance(AMR)andantibioticconsumptioninVietNam.
Methods: A16-hospital network (VietNam Resistance:VINARES)was establishedand consistedof nationalandprovincial-levelhospitalsacrossthecountry.Antimicrobialsusceptibilitytestingresults fromroutineclinicaldiagnosticspecimensandantibioticconsumptiondatainDefinedDailyDoseper 1000beddays(DDD/1000patient-days)wereprospectivelycollectedandanalysedbetweenOctober 2012andSeptember2013.
Results:Datafromatotalof24732de-duplicatedclinicalisolateswerereported.Themostcommon bacteriawere:Escherichiacoli(4437isolates,18%),Klebsiellaspp.(3290isolates,13%)andAcinetobacter spp.(2895isolates,12%).Thehospitalaverageantibioticconsumptionwas918DDD/1000patient-days.
Third-generationcephalosporinswerethemostfrequentlyusedantibioticclass(223DDD/1000patient- days, 24%),followed byfluoroquinolones(151 DDD/1000patient-days,16%)and second-generation cephalosporins(112DDD/1000patient-days,12%).Proportionsofantibioticresistancewerehigh:1098/
1580(69%)Staphylococcusaureusisolatesweremethicillin-resistant(MRSA);115/344isolates(33%)and 90/358 (25%) Streptococcus pneumoniae had reduced susceptibility to penicillin and ceftriaxone, respectively.Atotalof180/2977(6%)E.coliand242/1526(16%)Klebsiellapneumoniaewereresistantto imipenem,respectively;602/1826(33%)Pseudomonasaeruginosawereresistanttoceftazidimeand578/
1765(33%)toimipenem.OfAcinetobacterspp.1495/2138(70%)wereresistanttocarbapenemsand2/333 (1%)tocolistin.
Conclusions:ThesedataarevaluableinprovidingabaselineforAMRamongcommonbacterialpathogens inVietnamesehospitalsandtoassesstheimpactofinterventions.
©2019TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforAntimicrobial Chemotherapy.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.
org/licenses/by-nc-nd/4.0/).
* Correspondingauthorat:OxfordUniversityClinicalResearchUnit,NationalHospitalforTropicalDiseases,78GiaiPhong,DongDa,Hanoi,VietNam.
E-mailaddress:dungvtv@oucru.org(T.V.D.Vu).
1 TheVINARESconsortium:NationalHospitalforTropicalDiseases,VietnamNationalChildren’sHospital(formerly:NationalHospitalforPaediatrics),BachMaiHospital, VietDucHospital,Saint-PaulHospital,UongBiHospital,VietTiepHospital,HueCentralGeneralHospital,DaNangGeneralHospital,BinhDinhGeneralHospital,DakLak GeneralHospital,KhanhHoaProvincialHospital,ChoRayHospital,Children’sHospital1,HospitalforTropicalDisease,andCanThoCentralHospital.
https://doi.org/10.1016/j.jgar.2019.06.002
2213-7165/©2019TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforAntimicrobialChemotherapy.ThisisanopenaccessarticleundertheCCBY- NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
ContentslistsavailableatScienceDirect
Journal of Global Antimicrobial Resistance
j o u r n a lh o m e p ag e :w w w . e l s e vi e r . c o m / l o c a t e/ j g a r
1.Introduction
Antimicrobial resistance (AMR) among common bacterial pathogens is a recognised global health threat, leading to a significant increase in healthcare costs, treatment failures and deaths[1].Theissueismorepressinginlow-incomeandmiddle- income countries (LMICs) like Viet Nam, where the burden of resistantinfectionsisdisproportionate,whiledataandevidence ontheexactburdenandepidemiologyarescarce[2].
Overuseand inappropriateuseofantibioticsisanimportant driverfortheemergenceandspreadof AMR.TheWorldHealth Organization (WHO) introduced a six-point policy package on WorldHealth Day 2011, including surveillanceof antimicrobial resistanceanduseandrationalantimicrobialuse[3].Thisservedas theframeworkfortheGlobalActionPlanandmostNationalAction Planscomposed bymember states.The WHO alsopublished a comprehensivesetofrecommendationstotrackantimicrobialuse andresistanceinbacteria,andtoensureabetteruseofantibiotics andreduceantimicrobialuseinanimalhusbandry[4].InVietNam, thereis substantial overuseof antimicrobialdrugs, both inthe animalhealthsectorandinthehumanhealthsectorinhospitals andthecommunity[5,6].Anobservationalstudyofantibioticsales innorthernVietNamshowedhighproportionsoftransactionsat pharmaciesthatincludedantibiotics:24%inurbansitesand30%in ruralsites;thelargemajoritywerewithoutprescription[7].
Since 1988, a number ofnational and international efforts have been madetoimplementAMRsurveillanceinVietNam,ondifferentscales.
Table1describestheobjectives,scaleandresultsofeachproject.The VietNamResistancenetwork(VINARES)waslaunchedin2012asa collaborationbetweentheMinistryofHealth,VietnameseInfectious DiseasesSociety,OxfordUniversityClinicalResearchUnit-VietNam (OUCRU), and Linköping University, Sweden, together with 16 hospitals across the country[8]. This study describes the antimicrobial sensitivity testing (AST) results from clinical specimen isolates from the microbiologylaboratories,andantimicrobialconsumptiondatafrom the VINARES hospitals’ pharmacies between October 2012 and September2013.Theseresultsprovideanupdateonearlierresults publishedinthesituationanalysis[6]andrecommendationsfordata collectionimprovementtouseasevidencefordesignandimplemen- tationoftargeted interventionstotackleantibacterialoveruseand resistanceinVietNam.
2.Materialsandmethods
2.1.Datacollection
The VINARES network was previouslydescribed [8].Sixteen hospitalswereincluded:seveninthenorthern,threeinthecentral andsixinthesouthernregionofVietNam.Theseincludenational- level (n=7) and provincial-level (n=9) hospitals; two tropical
Table1
ResultofantimicrobialresistantsurveillanceprogramsandstudiesinVietNamfrom1988–2011.
Project Year Scale Vietnamese sites
Description Resultofprogram
NPSAR(former ASTS)[25][26]
1988–
2006 Viet Nam
9(in1988) 31(1993)
NPSAR–implementedbyMoH–wasanationalsurveillance programforAMR.Bacteriacausinginfectiousdiseasesin inpatientsandoutpatientswereisolatedandtestedfor antibioticsusceptibility.
E.coliproducingESBL:7.7%(42/548) K.pneumoniaeproducingESBL:23.7%(115/485)
ANSORP[27] 1996 Asia 1 Children’sHospital2,HCMC,participatedinaprojectof surveillanceforpneumococcalresistanceamongclinicalS.
pneumoniaeisolatesthatwerecollectedfrom14centresin11 countriesinAsiaandtheMiddleEastbetween2000and2001.
Proportionofpenicillinnon-susceptibilityS.
pneumoniae(71.4%)inVietNam,highestamong participantcountries.
Amoxicillin/clavulanicacid(AMC)resistancerate was22.2%.
ANSORP[28] 2004–
2006
Asia 1 Aprospective,multinationalsurveillancestudywith moleculartypinganalysisthatwasperformedtounderstand thechangingepidemiologyofS.aureusinfectionsinAsian countries.
UniversityofMedicineandPharmacyinHCMC,Viet Namreportedthathospital-acquiredand community-acquiredS.aureuswere74.1%and30.1%, respectively.SecondhighestratesofMRSAamong participants,lowerthanSriLanka.
Multi-center evaluation studyonS.
aureus bacteraemia [19]
2008–
2009
Global 3 Thisstudyassessedthevariationinmanagementand adherencetotreatmentguidelinesofS.aureusbacteraemia treatedconsecutivelyover1yearineightcentresintheUnited Kingdom,threeinVietNamandoneinNepal.
80patients(19%)fromVietNamhadmethicillin- resistantS.aureusbacteraemia.
GARP[6] 2009 Global 15 GARP-VietNamandtheUniversityofOxfordcollaboratedwith theVietnameseMoHtosetupnewantibioticresistance surveillanceprogram.Across-sectionalstudywasperformed tocollectantibioticresistanceandantibioticusedatafrom15 participatinghospitals2009.
E.coliresistanttocefuroxime(30–80%),toSXTfrom 60–80%.
E.coliandK.pneumoniaeproducingESBLwere15–
57%and7–73%,respectively.
40%ofPseudomonasaeruginosaand60%of Acinetobacterspp.wereresistanttoceftazidime.
SOAR[22] 2009–
2011
Global 11 AstudyonAMRsurveillanceofrespiratorypathogens.Isolates ofS.pneumoniaeandH.influenzaewereobtainedfromclinical materialstakenfromadultsandpaediatricpatientswith community-acquiredrespiratoryinfections.
47.8%and93.1%ofS.pneumoniaewerenon- susceptibletopenicillinandresistantto azithromycin.
40.5%ofH.influenzaeproducedβ-lactamase.
ResistancestoAMCforS.pneumoniaeandH.
influenzaewerelow(3.1%and2.6%,respectively).
SMART[29] 2009–
2011
Global 4 AstudyonantimicrobialsusceptibilityratesinaerobicGram- negativebacteriacausingintra-abdominalinfectionsinViet Nam(2sitesinHCMCand2sitesinHanoi).
ESBLpositiveinE.coliandK.pneumoniaewere48.1%
and39.5%,respectively.7.7%of13P.aeruginosa isolateswereresistanttoceftazidimebutnoneto ciprofloxacin.
Abbreviations:AMR,antimicrobialresistance;NPSAR,NationalProgramforSurveillanceinAntimicrobialResistance;ASTS,antimicrobialsensitivitytestingstudy;MoH, MinistryofHealth;ESBL,extendedspectrumbeta-lactamase;AMC,amoxicillin/clavulanicacid;ANSORP,AsianNetworkforSurveillanceofResistantPathogens;HCMC,Ho ChiMinhCity;MRSA,methicillin-resistantStaphylococcusaureus(S.aureus);GARP,GlobalAntibioticResistancePartnership;SXT,trimethoprim/sulfamethoxazole;SOAR, SurveyofAntibioticResistance;SMART,StudyforMonitoringAntimicrobialResistanceTrends;E.coli,Escherichiacoli;K.pneumoniae,Klebsiellapneumoniae;S.pneumoniae, Streptococcuspneumoniae;H.influenza,Haemophilusinfluenzae;P.aeruginosa,Pseudomonasaeruginosa.
diseases, two paediatric and one surgical hospital(s) (Fig. 1).
Antibioticconsumptionwasreportedmonthlybythepharmacy department.Eachdepartmentwasprovidedwithalaptopandan Excel file to enter the details of antibioticconsumption in the
intensive care unit (ICU) and whole hospital. In a few cases, patientshadtobuyoutsidemedicinethatwasnotavailableinthe pharmacydepartment(e.g.colistin).Thissituationwasignoredin thisstudy,butitdidnotaffecttheresultsasitwasveryrare.
Fig.1.Location,specialtyandtypeofthe16participatinghospitalsintheproject.
A baseline laboratory assessment was conducted at the 16 participatinghospitals.LaboratorieswereprovidedwithAmerican Type Culture Collection (ATCC) reference strains for internal qualitycontrol and wereenrolledinto themonthly UKNEQAS identification and ASTexternal quality assessment programme.
Eachlaboratoryperformedidentificationandsusceptibilitytests onanisolatethatwassentmonthlybyUKNEQAS,anduploaded theresultontheirwebsite.UKNEQUASassessedthetestresults andreturnedthereporttoOUCRU.Eachlaboratorywasprovided withalaptop,surveillancedatabasesoftware(WHONET)[9],and up-to-dateVietnamese-translated Clinical Laboratory Standards Institute (CLSI) guidelines (M100-S22) [10]. Staff from all participating sites were trained during several workshops on microbiologymethodsandtheuseofWHONET.Ahelpdeskwasset uptoaddressanyissuesthroughouttheproject.Reporteddata, includingASTresultsfrombacterialisolatesfromclinicalspeci- menssentinforroutinediagnosticsandhospital-wideantibiotic consumptioninthe16participatinghospitals,weresentmonthly from October 2012 to September 2013. The AST results were enteredmanuallyintoWHONETorexportedfromautomatedAST as VITEK2 (bioMérieux, Marcy l’Étoile, France) or LABCONN (Labsoft, Viet Nam) using BacLink (provided with WHONET).
Therewerefourhospitalsthatusedautomatedsystems,11used manualandoneusedboth.Aconfigurationfilewasdevelopedfor eachlaboratorytoconvertdata.BothASTandantimicrobialusage dataweresubmittedregularlyoronrequestbyemail.
Allduplicateisolatesforthesamepatient(identicalspecimentype andbacterium)intheASTdatasetwereexcludedfollowingWHO recommendations[11].Resultsobtainedbydiskdiffusion(DD)and minimuminhibitoryconcentration(MIC)methodswerecombined.If bothwereperformed,theMICresultwasused.Reportedresistance ratesaretheproportionofbacteriawiththeASTshowingresistance (result=R). Intermediate susceptible isolates (result=I) were not consideredasresistant.Resultswereaccepted,analysedandreported asisand,generally,noconfirmatorytestingofunexpectedresultsor rarephenotypesinreferenceorcentrallaboratorieswasperformed accordingtocurrentpracticeinmostLMICs.
2.2.Dataanalysis
The number and proportion of nine indicator bacteria — Acinetobacterspp., Pseudomonasaeruginosa (P.aeruginosa), Hae- mophilus influenzae (H. influenzae), Escherichia coli (E. coli), Klebsiella spp., Staphylococcus aureus (S. aureus), Enterobacter spp.,Enterococcusfaecium(E.faecium),andStreptococcuspneumo- niae(S.pneumoniae)–weredescribedoverallbypatientagegroup, sexandspecimentype.
AnS.aureusisolatewascountedasmulti-resistant(MRSA)ifit testedresistanttooxacillinorcefoxitininscreening.Similarly,S.
pneumoniae isolates were counted as reduced susceptibility to penicillinifresistancetooxacillinbydiskdiffusionwasreported (alsoifunconfirmedbyMICtesting).Resistancerateswerereported forallspecimenscombinedandforasubgroupofinvasiveisolates frombloodandcerebrospinalfluid(CSF).Theproportionassessed bydiskdiffusionorMICwasreportedifapplicable.Asamplewas consideredresistanttoanantibioticclassifitwasresistanttoat leastoneantibioticagentinthatclass,asperCLSIguidelines.
AntibioticconsumptionwassummarisedinnumberofDefined DailyDoseper1000beddays(DDD/1000patient-days).TheDDDis theassumedaveragemaintenancedoseperdayforadrugusedfor itsmainindicationinadults,whichcanbeobtainedfromtheWHO Antimicrobial DDD Quick Reference List [12]. The DDD/1000 patient-dayswascalculatedbyantibioticclassanddidnotdepend onbedsizeofhospital.
Foreachbacteria-antibioticpair,theresistantproportionwas calculatedastheratiobetweenthenumberofresistantisolates
and number of testedisolates for thatdrug. Allhospitalswere anonymisedandcodedfromH1toH16.Rsoftware(version3.1.11) wasusedfortheanalysis[13].
3.Results
3.1.Distributionofbacteriaandantibiotics
The ASTresults werereported for 26 808 isolatesfrom the VINARES network between October 2012 and September 2013.
Afterde-duplicationandremovaloffungi,24732bacterialisolates wereincludedintheanalysis.
The most commonly isolated organisms were: E. coli (4437 isolates,18%);Klebsiellaspp.(3290isolates,13%)–including2206 K.pneumoniaeisolates(9%);Acinetobacterspp.(2895isolates,12%) –including1668Acinetobacterbaumannii(A.baumannii)isolates (7%);P.aeruginosa(2326isolates,9%);S.aureus(2039isolates,8%);
Enterobacterspp.(1067isolates,4%);S.pneumoniae(813isolates, 3%);H.influenzae(404isolates,2%);andE.faecium(98isolates,1%).
Gram-negativebacteriaaccountedfor69%(17065isolates)and Gram-positive for 31% (7667 isolates). Sputum was the most frequently reported specimen (3625 isolates,15%) followed by blood(3222isolates,13%).Ofthetotalisolates,11%wererecovered frombloodandCSF.
Thedistributionofisolatesbygenderandageofpatientand typeofspecimenforthenineindicatorbacteriaissummarisedin SupplementaryTableS1.Among17369commonbacteria,ahigher proportionwasisolatedfrommale(66%)thanfromfemalepatients (34%),reflectingtheusualhospitalpopulationinVietNam.E.coli, K.pneumoniae,P.aeruginosaandAcinetobacterspp.(includingA.
baumannii) were mainly reportedfrom adults aged >20 years, whileS.aureusandS.pneumoniaeweremorecommonlyreported fromchildrenaged10years.
The distribution of bacteria was separately stratified by hospitalsandregion(Fig.2).Acinetobacterspp.weremostlyfound in two major general hospitals (H2and H4). H.influenzaewas mainly isolated from children, with >50% from one paediatric hospital(H11).TheproportionofE.coliwassimilaracrossthethree regions and, overall, the pathogen distribution also appeared similaracrossregions.
Antibiotic consumptionwas separately summarised by hospital and region.One hospital (H2)did not provideconsumption data.The averageantibioticconsumptionwas918DDD/1000patient-daysper hospital. Hospitalsin thecentraland southernregion hadsimilar antibioticconsumptionrates(1079and1026DDD/1000patient-days, respectively), while a lower rate was reported, on average, from hospitalsinthenorthernregion(799DDD/1000patient-days,after excludingapaediatrichospital).Themostcommonlyusedantibiotics werethird-generationcephalosporins(223DDD/1000patient-days, 24%), fluoroquinolones (151 DDD/1000patient-days,16%),second- generationcephalosporins(112DDD/1000patient-days,12%),penicil- lin combinations (111 DDD/1000 patient-days, 12%), followed by aminoglycosides (54 DDD/1000patient-days, 6%), penicillins with extended spectrum (53 DDD/1000 patient-days, 6%), fourth-generation cephalosporins (49DDD/1000 patient-days,5%), carbapenems(35 DDD/1000patient-days,4%)andglycopeptides(10DDD/1000patient- days,1%).Overall,third-generationcephalosporins wasthe largest group in all regions, followed byfluoroquinolones (Fig. 3). Two-thirds of second-generationcephalosporinswereusedinthecentralregion.
3.2.Antimicrobialsusceptibilities 3.2.1.General
TheproportionofisolateshavingAST assessedvariedbyhospital;
moreASTswereperformedatnationalhospitalsthaninprovincial hospitals(SupplementaryTableS2).Vancomycin-susceptibilitytest
forE.faeciumandimipenem-susceptibility testforP.aeruginosa weremostfrequentlycarriedoutacrossthehospitals(98%and76%, respectively).Fourhospitals(H3,H4,H8,H12)hadtested>95%of reportedisolates.
Fig. 4 illustrates the resistantproportions of each bacteria- antibioticcombination and itsrelationship withtheamountof antibiotic used and the number of isolates. Second-generation cephalosporinsarenotincludedinFig.4becausemosthospitals didnottestforthese.Morethan50%ofEnterobacteriaceaewere resistanttothird-generationcephalosporinsandfluoroquinolones.
Carbapenem-resistantproportionswerehighestamong Acineto- bacterspp.(around75%),followedbyP.aeruginosa(around50%) andlowestinEnterobacteriaceae(around 25%).S.aureusand E.
faecium were susceptible to vancomycin, but the susceptibility resultsofE.faeciumshouldbeinterpretedwithcautionbecauseof thelownumbers.S.pneumoniaewassusceptible,whilenearly50%
ofH.influenzaewereresistanttocombinationsofpenicillins.
3.2.2.AntimicrobialsusceptibilitytestingresultsofGram-positive bacteria
Table2showstheASTresultsforS.aureus,S.pneumoniaeandE.
faecium.S.aureuswas isolatedfrom2039specimens;258(13%)
werefrombloodandCSF.Among1580tested,1098(69%)S.aureus isolates wereidentifiedas MRSA. Similarresistance rateswere observed for isolates from blood and CSF isolates from all specimens. D-test for induced clindamycin resistance was not separatelyreported,thereforeresistancemaybeunderestimated.
Theresultsfrom813S.pneumoniaeisolatesincluded87from bloodandCSF.Etestsconfirmed/wereperformedin99%(353/358) of ceftriaxone, 39% (86/221) ofpenicillin and 54% (349/641) of vancomycin susceptibility tests. A total of 115/344 (33%) S.
pneumoniaeisolatesshowedreducedsusceptibilitytopenicillin;
the corresponding percentage in blood and CSF isolates was 7/30 (23%). Streptococcus pneumoniaesusceptibilityto penicillin was screened using oxacillin disks; 86 isolates were confirmedbypenicillinMICtest.Outof194oxacillindiskdiffusion results showing resistance, 87 (45%) were not confirmed by penicillinsusceptibilitytest.RegardingbloodandCSFspecimens, thereweresixisolatesconfirmedbypenicillinMICtest,andthey were all susceptible. Ten isolates (2%) of S. pneumoniae were resistant to vancomycin (two of 10 resistant isolates were confirmed by Etest). These results were not confirmed in a referencelaboratoryormolecularly,andthisshouldbeinterpreted withcaution.
Fig.3.Distributionofantibioticconsumptionbyhospitalsandregions.
Fig.2.Distributionofbacteriabyhospitalsandregions.Sumofeachgraphis100%.
Abbreviations:E.coli,Escherichiacoli;P.aeruginosa,Pseudomonasaeruginosa;S.aureus,Staphylococcusaureus;S.pneumoniae,Streptococcuspneumoniae;H.influenzae, Haemophilusinfluenzae;E.faecium,Enterococcusfaecium.
Among98E.faeciumisolates,onlyoneamoxicillinsusceptibili- ty test was performed. Of isolates, 78% were vancomycin susceptible, while the proportion of normal-level gentamicin susceptibilitywas<50%.
3.2.3.AntimicrobialsensitivitytestingresultsofGram-negative bacteria
Enterobacteriaceae susceptibility for amikacin, cefotaxime, ciprofloxacin, gentamicin, imipenem, tobramycin and trimetho- prim/sulfamethoxazole(SXT) are shownin Table 3. Amongthe 4437E.colisubmitted,527(12%)werefrombloodandCSFand992 (23%)fromurine.Morethan80%ofASTsforE.coliwerecarriedout by disk diffusion. Resistance was >50% for third-generation cephalosporins and fluoroquinolones. Lower resistance levels wereseenfor imipenemand amikacin.Resistancerates among
allisolates weregenerally higherthan proportionsobservedin bloodandCSFisolates(P<0.0001forcefotaximeandciprofloxa- cin;P=0.03forSXT).
Of the 3290 available Klebsiella spp. isolates, 2206 were K.
pneumoniae.Resistanceratestothird-generationcephalosporins andcarbapenemswere68%and16%,respectively.SimilartoE.coli, proportions of resistance to cefotaxime, ciprofloxacin and SXT werealsoloweramongbloodandCSFthanallisolates(P<0.0001 forall).Amongthe1067Enterobacterspp.isolates(82fromblood andCSF),21%wereresistanttocarbapenems.
Table4showstheASTresultsofAcinetobacterspp.,P.aeruginosa andH.influenzae.Resultsfrom2895Acinetobacterspp.(including 1668A.baumannii)weresubmitted.Dataoncolistinwereavailable fromonlyonehospitalandtwoof333isolateswerefoundtobe resistant.Resultsshowedveryhighresistantproportionsfor all Fig.4.Resistancerate,numberofisolatesandnumberofDDD/1000patient-daysperantibioticgroup.
Foreachbacteria-antibioticcombination,theresistantproportionwasshowninbottomleftfigure;thenumberofisolatesforeachorganismwasintopleftandthenumberof DDD/1000patient-daysofantibioticgroupwasinbottomright.Theresistantproportionwascalculatedfor15hospitalsthatprovidedantibioticconsumptiondata.DDD/1000 patient-daysfigureillustratestheamountofantibioticusedforallbacterialtreatment,notforanyspecificbacteria.Untestedorclinicallyirrelevantbacteria-antibiotic combinationswerenotshown.
antibiotics,from68% (amikacin) to77%(ceftazidime).Resistant proportionsofAcinetobacterspp. forimipenemandamikacinin bloodandCSFwerelowerthaninotherspecimens(P<0.0001for both).
Ofthe2326 P.aeruginosasubmittedisolates,154 werefrom bloodandCSF.Theresistanceratetoceftazidimewas33%,whichis similar toblood and CSF specimens. Of the isolates, 33% were resistanttoimipenem,while39%wereaminoglycosides-resistant.
Blood isolates had lower aminoglycosides-resistance levels in comparisonwithallisolates(P=0.04).
H.influenzaewas isolated from404specimens, including10 fromblood andCSF.Onehundredandsixty(71%)isolateswere resistant to ampicillin. The resistance rates to amoxicillin/
clavulanic acid (AMC) and cefotaxime were 39% and 44%, respectively.
4.Discussion
This study described the AST and antibiotic usage results reportedfromtheVINARESnetworkinVietNambetween2012 and 2013. Overall, the data showed high proportions of AMR amongalltestedbacteriaacrossallhospitalsinthenetwork.These results also show large variations in the resistant proportions betweenhospitals.Thishighlightstheimportanceofcontinuous monitoringoflocalantibioticuseandbacterialresistance,whichis oneofthecorestrategiesintheNationalActionPlanoncombatting Table2
SusceptibilityresultofGram-positivebacteriaisolatedintheVINARESproject.
Resistance/testedisolates(%) S.aureus S.pneumoniae E.faecium
Allspecimens (n=2039)
BloodandCSF (n=258)
Allspecimens (n=813)
BloodandCSF (n=87)
Allspecimens (n=98)
BloodandCSF (n=24)
MRSA 1098/1580(69) 145/197(74)
Vancomycin 22/823(3) 2/135(1) 10/641(2) 1/74(1) 21/96(22) 3/24(12)
Ciprofloxacin 456/1277(36) 71/189(38) 2/12(17) 0/0(NT)
Erythromycin 985/1315(75) 103/143(72) 246/289(85) 26/29(90)
Clindamycin 639/907(70) 74/118(63)
Gentamicin 435/1155(38) 55/135(41) 26/46(57) 8/9(89)
Levofloxacin 333/852(39) 40/125(32)
SXT 261/1156(23) 41/141(29)
Penicillin 115/344(33)a 7/30(22)b
Ceftriaxone 90/358(25) 9/52(17)
Amoxicillin 1/1(100) 0/0(NT)
Abbreviations:CSF,cerebrospinalfluid;MRSA,methicillin-resistantStaphylococcusaureus;SXT,trimethoprim/sulfamethoxazole.
Untestedorclinicallyirrelevantbacteria-antibioticcombinationswerenotshown.
ascreenedwithoxacillin,86isolatedwereconfirmedbyMICmethodandoneisolatewasresistanttopenicillin.
b screenedwithoxacillin,sixisolateswereconfirmedbyMICmethodandallweresusceptibletopenicillin.
Table3
SusceptibilityresultsofEscherichiacoli,Klebsiellaspp.andEnterobacterspp.isolatedintheVINARESproject.
Resistance/testedisolates(%) Escherichiacoli Klebsiellaspp. Enterobacterspp.
Allspecimens (n=4437)
BloodandCSF (n=527)
Allspecimens (n=3290)
BloodandCSF (n=413)
Allspecimens (n=1067)
BloodandCSF (n=82)
Amikacin 321/2936(11) 36/394(9) 638/2163(29) 61/329(19) 149/768(19) 11/58(19)
Cefotaxime 2342/4192(56) 240/514(47) 1479/2227(66) 101/190(53) 483/802(60) 25/48(52)
Ciprofloxacin 1758/3052(58) 188/397(47) 1222/2305(53) 139/332(42) 277/741(37) 28/63(44)
Gentamicin 1285/2655(48) 111/282(39) 1042/1989(52) 99/233(42) 294/760(39) 21/47(45)
Imipenem 180/2977(6) 15/403(4) 393/2294(17) 64/361(18) 144/665(22) 12/70(17)
SXT 1994/2803(71) 196/298(66) 1242/2007(62) 118/236(50) 360/709(51) 24/51(47)
Tobramycin 502/1309(38) 52/247(21) 588/1377(43) 65/236(28) 142/386(37) 16/48(33)
Abbreviations:CSF,cerebrospinalfluid;SXT,trimethoprim/sulfamethoxazole.
Table4
SusceptibilityresultofAcinetobacterspp.,Pseudomonasaeruginosa,HaemophilusinfluenzaeisolatedintheVINARESproject.
Resistance/testedisolates(%) Acinetobacterspp. P.aeruginosa H.influenzae
Allspecimens (n=2895)
BloodandCSF (n=313)
Allspecimens (n=2326)
BloodandCSF (n=154)
Allspecimens (n=404)
BloodandCSF (n=10)
Amikacin 1347/1993(68) 82/188(44) 329/1556(21) 13/82(16)
Cefotaxime 118/270(44) 5/10(50)
Ciprofloxacin 1298/1733(75) 74/207(36) 496/1527(32) 33/120(28) 17/269(6) 0/10(0)
Gentamicin 1385/1837(75) 130/214(61) 566/1456(39) 33/106(31)
Imipenem 1495/2138(70) 110/244(45) 578/1765(33) 36/129(28) 33/341(10) 0/10(0)
SXT 1258/1799(70) 86/192(45) 46/60(77) 3/5(60)
Ceftazidime 1650/2146(77) 124/242(51) 602/1826(33) 43/133(32)
TCC 771/1128(68) 47/141(33)
AMC 109/276(39) 2/5(40)
Erythromycin 3/3(100) 0/0(NT)
Ampicillin 160/226(71) 3/5(60)
Abbreviations:CSF,cerebrospinalfluid;SXT,trimethoprim/sulfamethoxazole;TCC,ticarcillin/clavulanicacid;AMC,amoxicillin/clavulanicacid.
drugresistance[14].Overall,lowerproportionsofresistancewere found in samples taken among isolates from blood and CSF samples,likelyreflectingdifferentproportionsofhospitalacquired isolatesamongsampletypes.
Cephalosporins,fluoroquinolonesandpenicillincovered70%of antibioticuse.Hospitalsweremorelikelytousefluoroquinolones andlesslikelytousecephalosporinsin2012(16%)thanin2008 (11%) [5]. This may be explained by the change in antibiotic susceptibilitybetweenthetwoperiods.
The antibiotic use DDD/1000 patient-days provides a rough indicationthat 91% of patients in hospitals were on antibiotic treatment.However,DDD/1000patient-daysisnotanappropriate measure with which to study the impact of antimicrobial stewardship because DDD/1000 patient-days can be misrepre- sentedby useof combination therapyor of higherdosages for certainindicationsorguidedbytherapeuticdrugmonitoring(not yetpracticedinVietNam).ThisbiasestowardshigherDDD/1000 patient-days and, thus, an overestimate of antibiotic use. As evidenceofthesebiasesinDDD/1000patient-daysmeasurements, a monthly point prevalence study conducted in ICUs within VINARES reported that the proportion of patients receiving antibioticsatsurveytimewas85%(2787/3287)andthat60%of patientswereprescribedmorethanoneantibiotic[15].Thisresult was higher compared with the point-prevalence conducted in 2008(67%(5104/7571)ofpatientsreceivingantibiotics)[5].Two possibleexplanationswerethat16of36participatingsitesofthe previous study were district-level hospitals with an expected lower usage and that there has been an increase in use of antibioticsinhospitalsovertime.Theglobalreportofantibiotic consumptionalsostatedanincreaseof20–25DDD/1000patient- daysinVietNaminthe2010–2015period[16].
Thenumberofdaysofantimicrobialtherapy(DOT)canbeused alongwithDDD/1000patient-daystoreportantibioticconsump- tionpracticesinhospital,whichoffersmoreclinicalrelevance[17].
TheDOTreportstheadministrationofasingleagentonagivenday, regardless of the number of doses administered or dosage concentration [18], which avoids overestimation of usage. The DDD/1000patient-daysallowscomparisonofantibioticuseacross countriesandhospitals,whileDOTcouldmakeconclusionsabout therelativeuseofoneantibioticcomparedwithanother[18].The measurementofDOTmightbedifficultformosthospitals[17].The measurement of antibiotic use by DDD/1000 patient-days and DOT/1000patient-dayswasdissimilarbecausetheadministered dose is dissimilar from the DDD recommended by the WHO, accordingtoastudyon130hospitalsinUSA[18].
Thecurrentstudyreportedsimilarlevelsofresistanceamong fourGram-negativebacteria—includingE.coli,K.pneumoniae,P.
aeruginosa and Acinetobacter spp. — compared with the GARP situationanalysis[6].Furtherstudiestoevaluatethechangesin antibioticconsumptionandthelikelyeffectonresistancelevelsat hospitalsare warranted to prioritisetargets of intervention. In addition,linkingmicrobiologyandsusceptibilitydatawithclinical datashouldbeaimedfor inthis surveillancenetwork,toallow assessmentoftheoriginoftheinfection(communityvs.hospital- acquired)forbetterinformingofevidence-basedguidelines.
IntheGARPsituationanalysis,anMRSAproportionofbetween 30–64%wasreported[6].Anevaluationfrom2008–2009inthree hospitalsinVietNamrevealedanMRSAproportionof19%in80 patientswithS.aureusbacteraemia[19];whilethecurrentstudy reported61%ofMRSAinbloodandCSFspecimens.Betterandmore comprehensivesurveillancemethodologymayexplainthehigher proportionsofMRSAbacteraemia.InVINARES,allMRSAtestswere performedbyoxacillinorcefoxitindiscdiffusiontest,whichisa reliableproxyfordetectionofMRSA[20].
More than 50% of S. pneumoniae were non-susceptible to penicillin.ThiswasalsoconfirmedinarecentreviewbyWHO[21],
showingtheresultofpenicillinnon-susceptibilityfrom47–48%in twocountriesintheSouth-EastAsiaand17–64%in10countriesof Western Pacific region. However, the extentof the problemis uncertain,partlyduetovariationsinhowreducedsusceptibilityis beingreportedandlargeproportionsofintermediateresults.The SOARstudyreported48%(138/289)ofpenicillin-non-susceptibili- tyamongS.pneumoniae[22].
Carbapenems were still mostly active against the tested Enterobacteriaceae. Compared with 2009, there was a slight increaseofresistancetoimipenemamongE.coli(from2%to6%) andKlebsiellapneumoniae(from10%to17%)[6].Mostimipenem- susceptibilitytestsforEnterobacteriaceaewereperformedbydisk diffusion,butthismaynotbeasreliableasbrothmicrodilutionor othermethods[23].Klebsiellaspp.showedincreasedresistanceto third-generation cephalosporins in comparison with a 2009 situation analysis study (from 40% to 66%). Fourth-generation cephalosporinsalsohadlesseffectonthesespecies.ForE.coli,the currentstudyshowedsimilarlyhighresistanceproportionstothe conventionalagents usedfor treatment,suchasSXTandthird- generationcephalosporins,incomparisonwiththe2009situation analysis(from60–80%)[6].Thesedatashowedthepersistentand increasing problem of Enterobacteriaceae resistance to third- generation cephalosporinsin the hospital setting. Even though resistancetocarbapenems in this reportis still low, thelevels tendedtoincreasefromthetimeofthe2009situationanalysisand arelikelytocontinueincreasingunlesseffectiveinterventionsare undertaken.
P.aeruginosawasstillsusceptibletoceftazidime,ciprofloxacin and imipenem, with a resistance rate around 30%. The 2009 situationanalysisshowedsimilarratesofaround40%resistanceto ceftazidimeand ciprofloxacin[6].Data fromVINARESare more likely to be representative, given the improved and quality assessed microbiological and reporting practices. P. aeruginosa was among the three most common aetiologies of hospital- acquiredinfectionsinICUsinVietNaminthesameperiod[15].
Thispoint prevalencesurvey showedhigher resistance propor- tionsofP.aeruginosatocarbapenems(55.7%)comparedwiththe current surveillance results, reflecting the larger burden of resistanceintheICUsettings.
In the current study, Acinetobacter spp. also showed 70%
resistancetoimipenem,whiletheproportionreportedinthe2009 situation analysis was 40% [6]. High carbapenem resistance in theseorganismsraisedgreatconcernfortreatmentalternatives,as colistin is usually the last resort and an increase in colistin resistanceislikelytohappen.Colistinresistancewasonlyassessed atone hospital inthenetwork, and 0.6% of333 tested isolates showed resistanceusingEtest/VITEK(whicharenottherecom- mendedstandard).Higherlevelsofcarbapenems(85%)andcolistin (1.3[n=78]and31.6%[n=38])resistancewerereportedfromtwo hospitals (of the three provincial and university hospitals participatinginVINARES)insouthernVietNaminanotherstudy from2012–2014[24].
VINARES reportedhigher AMC resistanceH.influenzaecom- paredwithSOAR(39%and2.6%,respectively).Thiscouldbethe result of increasing rate over time, or due to a large overlap betweenparticipatinghospitals.Mostimportantly,SOARsamples were from outpatients whereas the current ones were from samples sent to the micro laboratories. As microbiology is underutilised in Viet Nam (and other LMICs) this probably representsapopulationwithmoreadvancedinfectionsoramore extensive historyof pre-treatmentand,thus, selection ofresis- tancepathogens.ThisisveryillustrativeofhowthecurrentAMR surveillanceoverestimatesresistancebecauseofthisunderuseand lackofclinicalmetadataanddenominators.
Antimicrobial sensitivity testing was not conducted for all reported isolates and this could have introduced bias in the
reportedresistantproportions.Threeexplanationscanbegiven:
first,someASTsareonlyindicatedbasedontheresultsofanother (e.g.inS.aureusvancomycinwasonlytestedforMRSA);second, ASTs may have only been indicated when the isolates were suspected to be the aetiological pathogen causing clinical manifestations;andthird,ASTsmayhavebeenorderedbecause offailureinempiricaltreatment.
Forfutureeffortstoconductantimicrobialresistancesurveil- lanceand toprovidemoreusefuldatafor guidinglocal clinical treatmentandpublichealthresearch,itisimportantforclinical microbiologylaboratoriestobestrengthenedandbetterutilised.
Currently,thenumberofsamplescomingtothelaboratoryislow incomparisonwiththenumberofadmittedpatients.Itislikely thatthemoreseverepatients,transferredpatients,patientsfailing primarytreatment,andpatientswithhospital-acquiredinfections areoverrepresentedamongpatientsfromwhomsamplesaresent tothelaboratory.Clearclinicaldiagnosticand treatmentguide- lines, with consistent microbiological testing on suspicion of infectious aetiology, couldpartiallyovercome this bias.Clinical datashouldalsobeconsideredtobepartofthesurveillancedata.
Thiscouldincludeclinicalsyndrome,dateofadmission,transfer status,andantibioticusewhilesampled.
5.Conclusion
Thisprojectdemonstratesaninitiativewithalargenetwork ofhospitalstomonitorAMRinVietNam.Resistanceproportions tocommon antibiotics in 16 hospitalswere remarkablyhigh, and most have increased since the 2009 situation analysis.
Policy developmentfor pharmacies bothin hospitalsandthe community requires a structured solution to address this problem. AMR surveillance could be improved by enhancing capacityof clinical microbiologiststhrough advancedtraining and upgrading WHONET programwith more control of data entry and a pre-defined global configuration. Clinical data shouldbeincludedinthereportsfromthehospitalsinfuture.
Externalqualityassuranceisalsorecommendedforalltesting performedinthelaboratory.
Funding
TheVINARESprojectwasfundedbytheSwedishInternational DevelopmentCooperationAgency(SIDA),theWellcomeTrustof Great Britain, Linköping University (Sweden), and the Global AntibioticResistancePartnership(GARP).Thefundershadnorole instudydesign,datacollectionandanalysis,decisiontopublish,or preparationofthemanuscript.
Competinginterests Nonedeclared.
Ethicalapproval Notrequired.
Acknowledgments
TheauthorsthankhospitalsinVINARESprojectforproviding antibioticconsumptiondataandlaboratoryroutineASTresult.
AppendixA.Supplementarydata
Supplementarymaterialrelatedtothisarticlecanbefound,inthe onlineversion,atdoi:https://doi.org/10.1016/j.jgar.2019.06.002.
References
[1]Shallcross LJ, Howard SJ, Fowler T, Davies SC. Tackling the threat of antimicrobialresistance:frompolicytosustainableaction.PhilosTransR SocBBiolSci2015;370(1670).
[2]NguyenVanKinh,DoThiThuyNga,ChandnaArjun,NguyenVuTrung,Pham VanCa, Doan PhuongMai Doan, et al. Antibioticuse andresistance in emergingeconomies:asituationanalysisforVietNam.BMCPublicHealth 2013;13:1158.
[3]WHO. World Health Day 2011: policy briefs. Geneva: World Health Organization;2011. . [Accessed25July2018]http://www.who.int/world- health-day/2011/policybriefs/en/.
[4]WHO.Theevolvingthreatofantimicrobial resistance:optionsforaction.
Geneva:WorldHealthOrganization;2012.
[5]ThuTA,RahmanM,CoffinS,Harun-Or-RashidM,SakamotoJ, HungNV.
AntibioticuseinVietnamesehospitals:amulticenterpoint-prevalencestudy.
AmJInfectControl2012;40(9):840–4.
[6]NguyenVanKinh,LuongNgocKhue,CaoHungThai,TruongQuocCuong.First reportonantibioticuseandresistanceinVietnamhospitalsin2008-2009.
[439_TD$DIFF]MoH,GARP;2012.
[7]DoThiThuyNga,NguyenThiKimChuc,NguyenPhuongHoa,NguyenQuynh Hoa,NguyenThiThuyNguyen,HoangThiLoan,etal.Antibioticsalesinrural andurbanpharmaciesinnorthernVietnam:anobservationalstudy.BMC PharmacolToxicol2014;15:6.
[8]WertheimHeimanFL,ChandnaArjun,VuDinhPhu,PhamVanCa,NguyenThi DaiPhong,LamYenMinh,etal.Providingimpetus,tools,andguidanceto strengthennationalcapacityforantimicrobialstewardshipinVietNam.PLoS Med2013;10(5)e1001429.
[9]O’Brien TF, StellingJM. WHONET:an informationsystem formonitoring antimicrobialresistance.EmergInfectDis1995;1(2):66.
[10]M100-S22:PerformanceStandardsforAntimicrobialSusceptibilityTesting;
Twenty-SecondInformationalSupplement.p.188,ClinicalandLaboratory StandardsInstitute,2012.
[11]WHO.Globalantimicrobialresistancesurveillancesystem:manualforearly implementation.Geneva:WorldHealthOrganization;2015.
[12]WHO.WHOCC-ATC/DDDIndex.Geneva:WorldHealthOrganization;2017.. [Accessed30March2018]https://www.whocc.no/atc_ddd_index/.
[13]RCoreTeam.R:alanguageandenvironmentforstatisticalcomputing.Vienna:
RFoundationforStatisticalComputing;2018.https://www.R-project.org.
[14]MinistryofHealth,Nationalactionplanoncombattingdrugresistanceinthe periodfrom 2013–2020,VietNam. https://www.flemingfund.org/wp-con- tent/uploads/4abc59f751f5a94192e38da85e962356.pdf.Accessed2017.
[15]PhuVuDinh,WertheimHeimanFL,LarssonMattias,NadjimBehzad,Dinh Quynh-Dao,NilssonLennartE,etal.Burdenofhospitalacquiredinfectionsand antimicrobialuseinVietnameseadultintensivecareunits.PLoSOne2016;11(1).
[16]KleinEiliY,VanBoeckelThomasP,MartinezElenaM,PantSuraj,Gandra Sumanth,LevinSimonA,GoossensHerman,LaxminarayanRamanan,etal.
Global increase and geographic convergence in antibiotic consumption between2000and2015.ProcNatlAcadSci2018201717295.
[17]MorrisAM.AntimicrobialStewardshipPrograms:AppropriateMeasuresand MetricstoStudytheirImpact.CurrTreatOptionsInfectDis2014;6(2):101–12.
[18]PolkRE,FoxC,MahoneyA,LetcavageJ,MacDougallC.Measurementofadult antibacterialdrugusein130UShospitals:comparisonofdefineddailydose anddaysoftherapy.ClinInfectDis2007;44(5):664–70.
[19]Thwaites GE. The management ofStaphylococcus aureusbacteremiaintheUnited KingdomandVietnam:amulti-centreevaluation.PLoSOne2010;5(12).
[20]PourmandMR,HassanzadehS,MashhadiR,AskariE.Comparisonoffour diagnostic methods for detection of methicillin resistant Staphylococcus aureus.IranJMicrobiol2014;6(5):341–4.
[21]WHO.Antimicrobialresistance:globalreportonsurveillance.Geneva:World HealthOrganization;2014.
[22]VanPHBinhPT,MinhNHL,MorrisseyI,TorumkuneyD.Resultsfromthe survey ofantibioticresistance(SOAR)2009–11inVietnam. JAntimicrob Chemother2016;71(1):i93–i102.
[23]RechenchoskiDanieleZendrini,DambrozioAngélicaMarimLopes,VivanAna CarolinaPolano, SchuroffPauloAlfonso, BurgosTatiane dasNeves,etal.
Antimicrobialactivityevaluationandcomparisonofmethodsofsusceptibility forKlebsiellapneumoniaecarbapenemase(KPC)-producingEnterobacterspp.
isolates.BrazJMicrobiol2017;48(3):509–14.
[24]AnhNguyenTuan,NgaTranVuThieu,TuanHuynhMinh,TuanNguyenSi,Y Dao Minh, Chau Nguyen Van Vinh, et al. Molecular epidemiology and antimicrobialresistancephenotypesofAcinetobacterbaumanniiisolatedfrom patientsinthreehospitalsinsouthernVietnam.JMedMicrobiol2017;66 (1):46–53.
[25]PhamVanCa.ASTS-antimicrobialsensitivitytestingstudy.Presentationin NationalInfectiousDiseaseConferenceofVietNam2005.
[26]PhuongDM.QualityissuesinresistancetestinganddatainVietnam.[450_TD
$DIFF]Bach Mai Hospital; 2009. . Accessed 2018 https://cddep.org/wp- content/uploads/2017/06/en_phuong_bmh.pdf.
[27]SongJae-Hoon,JungSook-In,KoKwanSoo,KimNaYoung,SonJunSeong, ChangHyun-Ha,etal.Highprevalenceofantimicrobialresistanceamong clinical Streptococcus pneumoniae isolates in Asia (an ANSORP Study).
AntimicrobAgentsChemother2004;48(6):2101–7.
[28]SongJae-Hoon,HsuehPo-Ren,ChungDooRyeon,KoKwanSoo,KangCheol-In, Peck Kyong Ran, Yeom Joon-Sup, et al. Spread of methicillin-resistant
