Are antibiotic-resistant pathogens more common in subsequent episodes of diabetic foot infection?

Article

Reference

Are antibiotic-resistant pathogens more common in subsequent episodes of diabetic foot infection?

LEBOWITZ, Dan, et al.

Abstract

After antibiotic therapy of an initial diabetic foot infection (DFI), pathogens isolated from subsequent episodes might become more resistant to commonly prescribed antibiotics. If so, this might require a modification of the current recommendations for the selection of empiric antibiotic therapy. This study investigated whether the Infectious Diseases Society of America (IDSA) DFI guideline recommendations should be modified based on the number of past DFI episodes.

LEBOWITZ, Dan, et al . Are antibiotic-resistant pathogens more common in subsequent

episodes of diabetic foot infection? International Journal of Infectious Diseases , 2017, vol.

59, p. 61-64

DOI : 10.1016/j.ijid.2017.04.012 PMID : 28450198

Available at:

http://archive-ouverte.unige.ch/unige:95539

Disclaimer: layout of this document may differ from the published version.

1 / 1

Are antibiotic-resistant pathogens more common in subsequent episodes of diabetic foot infection?

Dan Lebowitz

, Karim Gariani

, Benjamin Kressmann

, Elodie von Dach

, Benedikt Huttner

, Placido Bartolone

, Nam Lê

, Morad Mohamad

,

Benjamin A. Lipsky

, Ilker Uçkay

*

aServiceofGeneralInternalMedicine,GenevaUniversityHospitalsandFacultyofMedicine,UniversityofGeneva,Geneva,Switzerland

bServiceofInfectiousDiseases,GenevaUniversityHospitalsandFacultyofMedicine,UniversityofGeneva,4,rueGabriellePerret-Gentil,1211Geneva14, Switzerland

cServiceofDiabetologyandEndocrinology,GenevaUniversityHospitalsandFacultyofMedicine,UniversityofGeneva,Geneva,Switzerland

dOrthopaedicSurgeryService,GenevaUniversityHospitalsandFacultyofMedicine,UniversityofGeneva,Geneva,Switzerland

eInfectionControlProgram,GenevaUniversityHospitalsandFacultyofMedicine,UniversityofGeneva,Geneva,Switzerland

fUniversityofOxford,Oxford,UK

ARTICLE INFO

Articlehistory:

Received9January2017

Receivedinrevisedform8April2017 Accepted13April2017

Corresponding Editor: Eskild Petersen, Aarhus,Denmark

Keywords:

Antibioticresistance Diabeticfootinfections Pathogens

Newepisodes Recurrence

SUMMARY

Background:Afterantibiotictherapyofaninitialdiabeticfootinfection(DFI),pathogensisolatedfrom subsequentepisodesmightbecomemoreresistanttocommonlyprescribedantibiotics.Ifso,thismight requireamodificationofthecurrentrecommendationsfortheselectionofempiricantibiotictherapy.

This study investigated whether the Infectious Diseases Society of America (IDSA) DFI guideline recommendationsshouldbemodifiedbasedonthenumberofpastDFIepisodes.

Methods:Thiswasasingle-centreretrospectivecohortsurveyofDFIpatientsseenduringtheyears 2010to2016.

Results:Atotal1018episodesofDFIin482adultpatientswereidentified.Thesepatientswerefollowed- upforamedianof3.3yearsafterthefirstDFIepisode.Thetotalnumberofepisodeswas2257andthe medianintervalbetweenrecurrentepisodeswas7.6months.AmongtherecurrentDFIs,thecausative pathogenswerethesameasinthepreviousepisodeinonly43%ofcases(158/365).Staphylococcusaureus wasthepredominantpathogeninallepisodes(range1to13episodes)andwasnotmoreprevalentwith theincreasingnumberofepisodes.DFIsweretreatedwithsystemicantibioticsforamediandurationof 20days(interquartilerange11–35days).Overall,therewasnosignificantincreaseintheincidenceof antibioticresistancetomethicillin,rifampicin,clindamycin,orciprofloxacinovertheepisodes(Pearson’s Chi-squaretestp-valuesof0.76,1.00,0.06,and0.46,respectively;correspondingp-valuesfortrendof 0.21,0.27,0.38,and0.08,respectively).

Conclusions:AfterthesuccessfultreatmentofaDFI,recurrentepisodesarefrequent.Ahistoryofa previousDFIepisodedidnotpredictagreaterlikelihoodofanyantibiotic-resistantisolateinsubsequent episodes.Thus,broadeningthespectrumofempiricantibiotictherapyforrecurrentepisodesofDFIdoes notappearnecessary.

©2017TheAuthor(s).PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.

ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by- nc-nd/4.0/).

Introduction

Patientswhohavehadonediabeticfootinfection(DFI)areat highriskoffutureepisodes.Inaddition,theyareusuallytreated withprolongeddurationsoftherapy,oftenwitharelativelybroad antibiotic spectrum, for recurrentepisodesofDFI (Uçkayet al., 2015;Uçkayetal.,2016;Uçkayetal.,2014).Antibioticuseisthe major clinical risk factor for promoting antibiotic resistance (Harbarth et al., 2000).The healthcare-associated transmission

* Correspondingauthorat:GenevaUniversityHospitalsandFacultyofMedicine, 4,rueGabriellePerret-Gentil,1211Geneva14,Switzerland.Tel:+41223729828;

fax:+41223723987.

E-mailaddress:[email protected](I.Uçkay).

1 DanLebowitzandKarimGarianicontributedequallyasfirstauthors.

http://dx.doi.org/10.1016/j.ijid.2017.04.012

1201-9712/©2017TheAuthor(s).PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

ContentslistsavailableatScienceDirect

International Journal of Infectious Diseases

j o u r n a l h o m ep a g e : w w w . e l s e v i e r . c o m / l o c a te / i j i d

ofresistantpathogensislikelywhenDFIpatientsarehospitalized orrequirefrequentpodiatriccareinspecializedcentres(Agostinho etal.,2013).HavingsubsequentDFIepisodestheoreticallyraises theriskofantibiotic-resistantinfectionsdeveloping(Zenelajetal., 2014).

To help prevent resistance and to reduce antibiotic-related costsandadverseeffects,the2012InfectiousDiseasesSocietyof America(IDSA)clinicalpracticeguidelinesforthediagnosisand treatment of diabetic foot infections (Lipsky et al., 2012) recommendprescribingantibioticsthat(1)haveprovenefficacy intreatingDFIs;(2)covercommonGram-positivecocci;and(3) have limited coverageof Gram-negativepathogens. Commonly used empiric oral options are clindamycin, co-trimoxazole, levofloxacin,andamoxicillin–clavulanate,administeredforabout 1–3 weeks for soft tissue infections and 4–6 weeks for non- amputatedosteomyelitiscases. These guidelinesalso statethat chronic,previouslyantibiotic-treated,orsevereinfectionsusually require broader spectrum regimens (Lipsky et al., 2012), but specificrecommendationscannot begivenbecauseof a lackof publishedcomparativedata.OtherguidelinesforDFIsalsoavoid offeringspecificempiricantibioticsuggestionsforthesamereason (Garianietal.,2014).

Thisstudywasundertakentoinvestigatewhetherornotthere is an effect of having a past DFI episode on the likelihood of antibioticresistanceinpathogensisolatedfromsubsequentDFIs.

Thisinformationcouldinformwhetherornotphysiciansshould considera pasthistoryofDFIwhen choosingempiricantibiotic therapy. Of note, this study was not designed to address the surgicalapproach toDFI or factors regarding thelikelihood of achievingremission,whichareaddressedelsewhere(Uçkayetal., 2015;Uçkayetal., 2016;Uçkayetal., 2014;Lipskyet al.,2012;

Garianietal.,2014).

Methodsandsetting

This was a single-centre retrospective cohort survey of DFI patientsseenduringtheperiodJanuary2010toDecember2016at GenevaUniversityHospital.Thishospital,theonlypublichospital inGenevaandalsocoveringsomeareasofneighbouringFrance, has an estimated average antibiotic consumption of 59 daily defineddoses(DDD)per100patient-days.Theinstitutionemploys a clinical pathway for managing DFI that includes submitting informationtoadatabaseonDFI.Therewerenomajorchangesin infectionpreventionorantibioticstewardshippoliciesduringthe studyperiod.Aspartofahospital-widequalityprogramme,the medical directoratehaswaived the need for individualpatient informed consent for the use of this clinical pathway, which includesallDFIs,independentoftheirseverity.DL,KG,BK,andEvD, allofwhomareexperiencedwithinfectiousdiseasesdatabases, completedthedataandbuiltthedatabase.Aresearchnurse(BK) andaninfectiousdiseasesphysician(IU)(Uçkayetal.,2009),both ofwhomspecializeincaringforDFIs,supervisedtheaccuracyof thedataand distinguishedbetweenwoundsthatwereinfected versuscolonizedandbetweencultureisolatesthatwerecausative pathogens versuslikely contaminantor colonizing microorgan- isms.

Definitionsandstatisticalanalysis

TheclinicalpathwayandDFIdefinitionarebasedontheIDSA DFIguidelines(Lipskyetal.,2012)andonspecialistconsultation (BAL). Episodes of infection after a first DFI (within the study period)weredefinedasaneworrecurrentDFIiftheyoccurredin thesame anatomical foot localizationand presentedat least 2 monthsafterthepriorepisode.Theaimwastoexcludepersistent DFIsfromthefinalanalysis.Itwasdecidedagainstconsideringall

prescribingof outpatient antibiotictherapy for non-DFI-related infectionsorperioperativeprophylaxisintheincludedpatients.

The three pathogens predominantlyisolated frommicrobio- logicalcultureofeachindividualDFIwererecorded,astheywere also the organisms against which treating clinicians usually targetedtheirantibiotictherapy.Thepathogencountwascensored atthreemicroorganisms.Iftherewasanassociatedsurgery,the intraoperative tissue or bone specimens were taken. In non- operatedcases,tissueorbonespecimenswereselected,iffeasible.

Otherwisepuswassampled.Superficialswabswithoutdirectpus contact, as well as enrichment broth cultures, were excluded.

Clinicalcurewasdefinedastheanamnestic,laboratory,andclinical resolutionofthesignsandsymptomsoftheformerDFI.

Thelaboratoryinitiallyprocessedallspecimensforculturein accordancewiththeClinical andLaboratoryStandardsInstitute recommendations (Performance Standards for Antimicrobial SusceptibilityTesting, 2007), before switching to theEuropean Committeecriteriain2014(EuropeanCommitteeonAntimicrobial SusceptibilityTesting,2014).Thestandardmicrobiologylaborato- ryincubationtimewas5days.Clonaltypingofmicroorganisms was not done routinely. Focus was placed on the antibiotic resistancetofourofthemostfrequentlyprescribedagentsforDFIs:

methicillin, rifampicin, ciprofloxacin, and clindamycin. Because antibiotictherapy would have affectedbacteria that were only colonizersduringthefirstepisode,orwerenewlyacquiredinthe intervalbetweeninfections(Agostinhoetal.,2013),itwasdecided toanalyzeantibiotic resistanceepidemiologicallyoverthetime periodofthestudyratherthanforeachpathogenforeveryepisode ofDFI.Forexample,insteadofanalyzingwhetheranEscherichia coliisolateculturedduringafirstinfectionwasstillsusceptibleto ciprofloxacin in the subsequent episode, it was determined whetherciprofloxacin resistance occurredamong allpathogens isolatedinanysubsequentepisode(s).

Forgroupcomparisons,Pearson’sChi-squaretestwasused.The p-valuefortrendassessedchangesovertimeandepisodes.Stata software(version9.0;StataCorp.,CollegeStation,TX,USA)was usedforthedataanalysis.

Results

Patients,episodes,andtherapy

Thestudyhadaccesstodataonatotalof1018DFIepisodes(279 infemales),withamedianfollow-upof3.3yearsafterthefirstDFI episode(interquartile(IQR)range0.8–9.0years).Amongthe482 diabeticpatientsincludedinthiscohortstudy(whohadamedian duration of diabetes of 15 years), 244 suffered a second DFI episode,132athird,71afourth,39afifth,18asixth,10aseventh, 10aneighth,sixaninth,andthreeatenth.Thesixeleventhto fourteenth episodesinvolvedonly threepatients. Overall,there were 2257 episodes, of which a total of 540 were follow-up episodes. The median interval between DFI episodes was 7.6 months(IQR2.2–30.2months).

Themedianpatientageonadmissionwas69years,bodymass indexwas28kg/m²,ankle-brachialindexwas1.0,andC-reactive proteinlevelwas62mg/l.MostDFIsinvolvedthefore-foot,but65 (15%)involvedthehind-footandankleand38(9%)involvedthe mid-foot. Overall, 392 (39%) episodes were complicated by underlyingosteomyelitis.

TheDFIsweretreatedwithsystemicantibioticsforamedian durationof20days(IQR11–35days),includingamedianof5days intravenously (IQR 0–12 days). The six most frequently used antibioticdrugclasseswerebeta-lactams(n=1017),glycopeptides (n=116),quinolones(n=91),co-trimoxazole(n=49),clindamycin (n=46),andrifampicin (n=25).Themediannumber ofsurgical debridementswas1(range0–7);596oftheseinvolved(partial) 62 D.Lebowitzetal./InternationalJournalofInfectiousDiseases59(2017)61–64

lower extremity amputations. Overall, 610 (60%) cases were clinicallyconsideredasischemic,98underwentangioplasty,and 98receivedhyperbaricoxygentherapy.

Pathogens

Amongthe1018DFIepisodes,thenumberofepisodeswithtwo pathogenswas381(37%)andthenumberwiththreepathogens was163(16%).Thefivemostfrequentlyisolatedmicroorganisms were Staphylococcus aureus (325 episodes), coagulase-negative staphylococci(n=35),Enterococcusfaecalis(n=40),Streptococcus agalactiae(n=26),andPseudomonasaeruginosa(n=61).S.aureus was the predominant pathogen in all episodes of recurrent infection (from 1 to 13). S. agalactiae and coagulase-negative staphylococciwererarelyencounteredbeyondepisode7,whereas enterococciandP.aeruginosawerefoundequallyinhigherepisode numbers.Amongthe365DFIsoccurringatleastthreetimes,the threedominantpathogensshowedpartialconcordancebetween episodesonly43%ofthetime(158/365).In57%ofepisodes,the pathogens isolated were unrelated tothose found in the prior episode.

Antibioticresistanceacrossepisodesinthestudyandthemedical centre

The incidence of antibiotic resistance of DFI isolates to methicillin, rifampicin, clindamycin, and ciprofloxacin did not increase significantlyover subsequent DFI episodes(Chi-square testp-valuesof0.76,1.00,0.06,and0.46,respectively).Incontrast, therewasatendencytowardslowerratesofantibioticresistance fromepisode1to3,withp-valuesfortrendof0.21,0.27,0.38,and 0.08, respectively. Table 1 shows the comparison of rates of antibioticresistanceforallclinicalisolatesin DFIepisodes1–3.

Table2revealstheratesofresistancetoantibioticsofsomeofthe key pathogens at the study institution during the period investigated.

TherateofresistancetomethicillinofS.aureusisolatesinthese DFIepisodeswasnothigherthantheaveragefortheclinicalstrains at the medical centre in 2011 (75/325 vs. 605/2630, p=0.98).

Moreover,theS.aureusstrainsoftheDFIpatientshadalowerlevel ofresistancetoclindamycin(74/325vs.736/2630, p=0.05)and rifampicin(11/325vs.26/2630,p=0.01),andtendedtohavelower ratesofresistancetociprofloxacin(61/325vs.605/2630,p=0.08).

P. aeruginosa also tended to have lower rates of resistance to ciprofloxacininDFIpatientscomparedtoallclinicalisolatesinthe medicalcentre(9/61vs.98/1230,p=0.08).Whenanalyzingthe61 PseudomonasDFIisolatesseparately,thedifferenceinciprofloxa- cin resistance across episodes 1 to 3 and overall was not statistically significant (Chi-square test, p=0.87), which was confirmed in the trend analysis (p=0.42). Overall, while the proportionofantibioticresistancedecreasedovertheepisodesfor DFI patients (Table 1), it remained stable for all other clinical isolatesattheinstitution(Table2).

Discussion

Inthissingle-centrecohortstudyinvolving1018episodesof DFIin482adultpatients,theconcordanceofthethreedominant woundpathogenswasonly43%forsubsequentepisodes,while the pathogens isolated were unrelated to those in the prior episodeinnearlytwo-thirdsofcases.Basedonthemicrobiolog- icalfindings,many so-calledDFIrecurrencesare probablynew episodes.

FromthefirstepisodeofDFI,theratesofresistanceofwound isolatesofS.aureusandP.aeruginosaweresignificantlyhigherthan those of all clinicalisolates in thestudy institution. Moreover, while theproportionof antibioticresistance for DFIsdecreased overtime,itremainedstableforallotherclinicalisolatesatthe institutionduringthe6-yearstudyperiod,forwhichthereisno apparentexplanation.Noincreasedoccurrenceofneworspecific bacterialspecieswiththeincreasingnumberofDFIepisodeswas detected.Thedurationofantibiotic therapyadministeredtothe patientswasconsistentwiththerecommendationsinwidelyused DFIguidelines(Lipskyetal.,2012;Garianietal.,2014).Thechoice of antimicrobial agents, largely consisting of oral amoxicillin– clavulanate,clindamycin,orfluoroquinolones,wassimilartothe experiencereportedbylargeVeteranshealthnetworksintheUSA (Finckeetal.,2010).RegardingthemicrobiologicalprofileofDFI pathogens,thepresentdataareconsistentwiththoseofreports from CentralEuropean and North Americaninstitutions,which haveshownapredominanceofS.aureusandotheraerobicGram- positive pathogens (Uçkay et al., 2014; Harbarth et al., 2000;

Charlesetal.,2015).Thisfindingisincontrasttothosereportedin publicationsfrom(sub)tropicalcountriesinAsiaandAfrica,which have shown a predominance of Gram-negative pathogens, especiallyP.aeruginosa(Uçkayetal.,2014).

Thisstudyhasseverallimitations,themostimportantbeing that the assessment of resistance patterns was restricted to methicillin, ciprofloxacin, rifampicin, and clindamycin (Czekaj etal.,2011).TheseveritygradingoftheDFI(e.g.,accordingtothe IDSAguidelines(Lipskyetal.,2012))wasnotuniformlyassessed, nor was resistance to otherantibiotics, suchas co-trimoxazole (Harbarth et al., 2015), linezolid, tetracyclines, daptomycin, glycopeptides, piperacillin–tazobactam, colistin (Valour et al., 2013),orcarbapenems,whichareoccasionallyusedformethicil- lin-resistantS. aureus, sepsis,or non-fermentingGram-negative

Table2

Ratesofantibioticresistanceinallclinicalisolatesoftheselectedkeypathogensduringthestudyperiod—entirehospital.

Antibiotic MSSA Streptococcusagalactiae Pseudomonasaeruginosa

Year 2010 2013 2016 2010 2013 2016 2010 2013 2016

Methicillin 0% 0% 0% 0% 0% 0% -^a -^a -^a

Clindamycin 11% 12% 17% 22% 22% 20% -^a -^a -^a

Rifampicin 1% 1% 1% 0% 0% 5% -^a -^a -^a

Ciprofloxacin 3% 4% 3% -^a -^a -^a 8% 10% 9%

MSSA,methicillin-susceptibleStaphylococcusaureus.

aDrugnotusedforthisorganism.

Table1

Ratesofantibioticresistanceaccordingtotheincreasingnumberofepisodesof diabeticfootinfection.

AllpathogenscausingDFI,byepisode p-Value^a

Episode1 Episode2 Episode3

49% 23% 14% 0.21

53% 25% 11% 0.08

54% 23% 8% 0.38

46% 23% 17% 0.27

DFI,diabeticfootinfection.

ap-Valuefortrend.

bacteria. This choice was made based on the relatively small number of DFI episodes that were treated with these other antibioticagentsandtheexpectationthatalloftheGram-positive pathogenswouldbesusceptibletolinezolid(Valouretal.,2013), daptomycin(Jugunetal.,2013),andglycopeptides.

Second,theresultsshouldnotbeinterpretedassuggestingthat newpathogens wereselectedbyongoing antibiotic therapy,as commonlyoccursinclinicalpractice(Al-Mayahietal.,2015).In this study, onlypatients who had an antibiotic-freeinterval of several weeks before experiencing a new episode of DFI were included.Thisantibiotic-freetimewindowisagoodcompromise betweenselectingrecurrentepisodes(uptoaquarteroffurther DFI episodes recurred at 2 months) and new DFIs. No other publicationappearstohavementionedthemediandurationuntil onsetofantibioticresistanceinDFIs.

Third,thisstudywassolelyepidemiologicalandthusdidnot prospectivelyinvestigatethedevelopmentofaspecificantibiotic resistanceforagivenpathogeninanindividualpatient.Takinginto account the considerable variation in DFI microbiology among episodes,suchadetailedindividualizedanalysiswasnotfeasible whenanalyzingover1000events.

Fourth, infection prevention programmes, such as the promotion of hand hygiene or antibiotic stewardship, might have reduced the prescription of selected antibiotics and indirectlydecreased theoverall resistance tothese antibiotics.

However,itis believedthattheinfluenceof thesehypothetical programmesis small, becausemost suchprogrammestarget a reduction in nosocomial infections and not the antimicrobial resistance, and many DFIs are treated in the community and involve many different antibiotics. There are alsono scientific datatoindicatethatantibioticstewardshipprogrammesdirectly reduce theresistance patterns of DFIs. Additionally,there was nomajorchangeinthesepoliciesatthestudyinstitutionduring thestudyperiod.

Fifth,DFIepisodestreatedelsewhereorbeforethestudyperiod mighthavebeenmissed,especiallyforpatientsresidinginGeneva foronlyafewyears.However,becausethestudycentrehasbeen thelargestandtheonlypublicmedicalcentreforseveraldecades, itisbelievedthatthisisunlikely.

Finally,theresultsmightreflectthesituationinmanyWestern countries,butcannotbegeneralizedtootherregionsinresource- poor countries or those in which there is a high burden of community-andhealthcare-associatedantibiotic-resistantpatho- gens, where over-the-counter antibiotic agents are widely available, or that have a high prevalence of non-fermenting Gram-negativerodsinDFIs(Uçkayetal.,2014).

Inconclusion,basedontheresultsofthisstudy,itissuggested that for medical centres similar to this one, the current IDSA recommendations for the empiric oral antibiotic treatment of recurrentmildtomoderateDFIs(Lipskyetal.,2012)donotneedto bemodifiedtoencouragebroaderspectrumcoveragebasedona historyofpastDFIepisodes.Itwouldbebeneficialfortheseresults tobeconfirmedatothersites,especiallyinprospective,patient- levelstudies.

Ethicalapproval

Notrequiredforthisreview.

Funding

Therewasnofundingforthiswork.

Conflictofinterest

Therewasnofundingforthepreparationofthismanuscript.

BAL has served as a consultant to KCI/Acelity, Innocoll, and Dipexium. IU has received research funding from Innocoll.

However, the content of this paper has no relation with the consultancyofanyoftheauthors.Allauthorsdeclarenofinancial support,grants,financialinterests,orconsultancythatcouldlead toconflictsofinterest.

Authorcontributions

Allauthors contributed tothewriting and reviewing of the manuscript.

Acknowledgements

WethanktheMedicalDirectorsofGenevaUniversityHospitals and the teams of the Laboratory of Bacteriology and the Orthopaedic Service for their support. We are indebted to Mr AbdessalamCherkaouiforprovidingresistancedata.

References

AgostinhoA,RenziG,HausteinT,JourdanG,BonfillonC,RougemontM,etal.

Epidemiologyandacquisitionofextended-spectrumbeta-lactamase-producing Enterobacteriaceaeinasepticorthopedicward.SpingerPlus2013;2:91.

Al-MayahiM,CianA,LipskyBA,SuvàD,MüllerC,LandelleC,etal.Administrationof antibioticagentsbeforeintraoperativesamplinginorthopedicinfectionsalters cultureresults.JInfect2015;71:518–25.

CharlesPG,UçkayI,KressmannB,EmonetS,LipskyBA.Theroleofanaerobesin diabeticfootinfections.Anaerobe2015;34:8–13.

CzekajJ,DinhA,MoldovanA,VaudauxP,GrasG,HoffmeyerP,etal.Efficacyofa combinedoralclindamycin-rifampicinregimenfortherapyofstaphylococcal osteoarticularinfections.ScandJInfectDis2011;43:962–7.

EuropeanCommitteeonAntimicrobialSusceptibilityTesting.BreakpointTablesfor InterpretationofMICsandZoneDiameters. Version4.

FinckeFG,MillerDR,ChristiansenCL,TurpinRS.Variationinantibiotictreatment fordiabeticpatientswithseriousfootinfections:aretrospectiveobservational study.BMHHealthServiceRes2010;10:193–204.

GarianiK,UçkayI,LipskyBA.ManagingDiabeticFootInfections:AReviewofthe NewGuidelines.ActaChirurgicaBelg2014;114:7–16.

HarbarthS,SamoreMH,LichtenbergD,CarmeliY.Prolongedantibioticprophylaxis aftercardiovascular surgeryand its effectonsurgical site infectionsand antimicrobialresistance.Circulation2000;101:2916–21.

HarbarthS,vonDachE,PaganiL,Macedo-VinasM,HuttnerB,OlearoF,etal.

Randomizednon-inferioritytrialtocomparetrimethoprim/sulfamethoxazole plus rifampicin versus linezolid for the treatment of MRSA infection. J AntimicrobChemother2015;70:264–72.

JugunK,VaudauxP,GarbinoJ,PaganiL,HoffmeyerP,LewD,etal.Thesafetyand efficacyofhigh-dosedaptomycincombinedwithrifampicinforthetreatment ofGram-positiveosteoarticularinfections.IntOrthop2013;37:1375–80.

Lipsky BA, Berendt AR, Cornia PB, Pile JC, Peters EJ, Armstrong DG, et al.

2012InfectiousDiseasesSocietyofAmericaclinicalpracticeguidelineforthe diagnosisandtreatmentofdiabeticfootinfections.ClinInfectDis2012;54:132–

73.

PerformanceStandardsforAntimicrobialSusceptibilityTesting.17^thInformational Supplement.DocumentM100-S17.USA: Clinicaland LaboratoryStandards Institute;2007.

UçkayI,Vernaz-HegiN,HarbarthS,SternR,LegoutL,VautheyL,etal.Activityand impactonantibioticuseandcostsofadedicatedinfectiousdiseasesconsultant onasepticorthopaedicunit.JInfect2009;58:205–12.

UçkayI,GarianiK,PatakyZ,LipskyBA.Diabeticfootinfections:state-of-the-art.

DiabetesObesMetab2014;16:305–16.

UçkayI,Aragón-SánchezJ,LewD,LipskyBA.Diabeticfootinfections:whathavewe learnedinthelast30years.IntJInfectDis2015;40:81–91.

UçkayI,GarianiK,Dubois-FerrièreV,SuvàD,LipskyBA.DiabeticFootInfections:

Recent literature and cornerstones of management. Curr Opin Infect Dis 2016;29:145–52.

ValourF,DutroncH,DinhA,CazorlaC,PavèseP,LesensO,etal.Difficult-to-treat Gram-negativeboneandjoint infections:efficacyand safetyofprolonged intravenouscolistin.IntJAntimicrobAgents2013;41:197–9.

ZenelajB,BouvetC,LipskyBA,UçkayI.Dodiabeticfootinfectionswithmethicillin- resistantStaphylococcusaureusdifferfromthosewithotherpathogens?.IntJ LowExtremWounds2014;13:263–72.

64 D.Lebowitzetal./InternationalJournalofInfectiousDiseases59(2017)61–64