Early alert from the microbiology laboratory improves the outcome of elderly patients with Enterococcus spp. bloodstream infection: Results from a multicentre prospective study

Early

alert

from

the

microbiology

laboratory

improves

the

outcome

of

elderly

patients

with

Enterococcus

spp.

bloodstream

infection:

Results

from

a

multicentre

prospective

study

M.

Falcone

a,

*

,

G.

Tiseo

b

,

F.

Dentali

c

,

E.

Foglia

d

,

M.

Campanini

e

,

F.

Menichetti

a

,

A.

Mazzone

f

,

on

behalf

of

the

FADOI

(Federazione

delle

Associazioni

dei

Dirigenti

Ospedalieri

Internisti)

and

GISA

(Italian

Group

for

Antimicrobial

Stewardship)

a

DivisionofInfectiousDiseases,DepartmentofClinicalandExperimentalMedicine,UniversityofPisa,Pisa,Italy

b

DepartmentofInternalMedicineandMedicalSpecialties,‘Sapienza’UniversityofRome,Rome,Italy

c_{DepartmentofClinicalMedicine,UniversityofInsubria,Varese,Italy} d

CentreforResearchonHealthEconomics,SocialandHealthCareManagement(CREMS),UniversityCarloCattaneo–LIUC,Castellanza,Italy

e

InternalMedicineWard,OspedaleMaggioredellaCarità,Novara,Italy

f

InternalMedicineWard,OspedaleCivile,Legnano,Italy

ARTICLE INFO

Articlehistory:

Received11December2018

Receivedinrevisedform4February2019 Accepted19February2019

Availableonline27February2019

Keywords: Enterococcusspp. Bloodstreaminfection Elderlypatients Microbiologylaboratory ABSTRACT

Objectives:Thisstudydescribestheclinicalfeaturesandoutcomesofpatientswithbloodstreaminfection (BSI)duetoEnterococcusspp.andidentifiedfactorspredictiveofmortality.

Methods: This analysis is part of a prospective multicentre observational study of consecutive hospitalisedpatientswithBSIconductedfromMarch2012toDecember2012in31internalmedicine wardsinItaly.PatientswithenterococcalBSIwereselectedfromtheentirecohort.Patientcharacteristics, therapeutic interventions and outcome were reviewed. Cox regression analysis was performedto identifyfactorsassociatedwithin-hospitalmortality.Hazardratios(HRs)and95%intervalconfidences (CIs)werecalculated.

Results:Among533patientswithBSI,41(7.7%)hadBSIbyEnterococcusspp.(28Enterococcusfaecalis,4 Enterococcus faecium and 3 eachof Enterococcus avium,Enterococcus casseliflavus and Enterococcus gallinarum).SixBSIs(14.6%)werepolymicrobial.Median(IQR)patientagewas73(66–85.5)years. In-hospitalmortalitywas24.4%.Polymicrobialinfection(HR=9.100,95%CI1.295–63.949;P=0.026),age (HR=1.261,95%CI1.029–1.546;P=0.025)andSOFAscore(HR=1.244,95%CI1.051–1.474;P=0.011)were riskfactorsforin-hospitalmortality.Conversely,receivinganalertfromthemicrobiologylaboratory beforeobtainingfinalantimicrobialsusceptibilityresultswasassociatedwithsurvival(HR=0.073,95% CI0.007–0.805;P=0.033).

Conclusion: BSI due to Enterococcus spp. in elderly patients is associated with high mortality. Polymicrobialinfection,ageandSOFAscorearefactorsassociatedwithpooroutcome.Conversely,early alertfromthemicrobiologylaboratoryimprovespatientsurvival.

©2019InternationalSocietyforChemotherapyofInfectionandCancer.PublishedbyElsevierLtd.All rightsreserved.

1.Introduction

Enterococcus spp.are animportant causeof nosocomialand community-acquiredinfectionsbothintheUSA[1,2]andEurope

[3].Traditionally,Enterococcusfaecalishasbeenconsideredasthe leading infectingspecies. Moreover,in thelast decades several Enterococcusspp.withnovelmechanismsofacquiredresistanceto

antimicrobials have became more and more prevalent [4], and vancomycin-resistant enterococci (VRE) causing hospital out-breaksworldwiderepresentanimportantthreattopatientcare

[5–8].Alargenumberofmolecularepidemiologicalstudieshave beenperformedtoobtaininsightsintothedisseminationofVRE clonesinandbetweenhospitals,infarmanimalsandinhealthy humans.Ithasbeenreportedthat>80%ofEnterococcusfaecium isolates currently recovered from US hospitals are resistant to vancomycin, and virtuallyallof them (>90%) exhibitampicillin resistance[9].Bloodstreaminfections(BSIs)dueto vancomycin-resistantE.faeciumnotonlyincursignificantcostsforhealthcare

* Correspondingauthor.

E-mailaddress:marco.falcone@unipi.it(M. Falcone).

https://doi.org/10.1016/j.jgar.2019.02.014

2213-7165/©2019InternationalSocietyforChemotherapyofInfectionandCancer.PublishedbyElsevierLtd.Allrightsreserved.

ContentslistsavailableatScienceDirect

Journal

of

Global

Antimicrobial

Resistance

systemsbutalsoplacevulnerablepatientsatgreaterhigherriskof fatalinfection[10].

In-hospital mortalityassociatedwithenterococcalBSIvaries greatlybetweenstudiesdependingonthetypeofpatients,ward ofhospitalisationandantimicrobialsusceptibilityprofile[11,12]. Inonestudyofpatientswithenterococcalbacteraemia,thecrude 30-daymortalityratewas23%withanattributablemortalityrate of6%[7],whereasinpatientswithVREbacteraemiainanother studythecrudeandattributablemortalityrateswere50.2%and 30.3%,respectively[13].Arecentstudyofpatientswithintensive care unit (ICU)-acquired enterococcal BSI showed that these infectionsareassociatedwithincreasedcasefatalitybutwithlow attributablemortality [14].Infections suchas bacteraemia and endocarditis can cause life-threatening disease with high mortalityrates,especially inelderlypatientswithmultiple co-morbidities [2,15–17]. In these cases, prompt and accurate identification and early appropriate antibiotic therapy are essentialtoreducemortality[18].

The aim of the current study was to describe the clinical features and outcomes of elderly patients with BSI due to Enterococcusspp.andtoidentifyfactorsindependentlyassociated withmortality.

2.Materialsandmethods 2.1.Studydesign

This evaluation was performed as part of a prospective multicentreobservationalstudyofconsecutivepatientswithBSI hospitalisedin31internalmedicinewardsfrom14differentItalian regionsfrom1March2012to31December2012(SNOOPIIstudy)

[19,20]. A minimum of 15 patients were requested from each

participatingcentre.Allincludedpatientsweremonitoreddaily fromtheonsetofBSIuntiltheendofhospitalisation.Approvalof thestudy protocolwas obtained from theinstitutional review boards at each hospital, which waived the requirement for obtaininginformedconsent.Thestudywasconductedunderthe auspicesoftheFederationofAssociationofExecutivesofHospital Internists(FADOI).

All consecutive patients aged 18 years admitted to the hospital with BSI or those developing a BSI during hospital-isationwere included inthe study. Among them, all patients withBSIcausedbyEnterococcusspp.havebeenanalysedinthe currentstudy.

In each centre, trained physicians prospectively collected information onpatient characteristics,dailyclinical eventsand illnesscharacteristicsandreporteditonastandardclinicalrecord form.Morespecifically,demographicdatasuchasageanddateof hospital admission wererecorded. For each patient, the inves-tigators recorded past medical history, reason for hospital admission and co-morbidities as assessed by the Charlson comorbidity index. Information about co-morbid medical con-ditionssuchasdiabetes,cardiovasculardisease,chronic obstruc-tive pulmonary disease, liver disease and cancerwas obtained through review of the medical records. Clinical data such as presenceoffever, hypotension,tachycardia,tachypnoea,altered mentalstatusandsignsoforganfailureattheonsetofBSIwere recorded.Thepresenceofamedicaldevice(peripherallyinserted central catheter, dialysis catheter, central venous catheter or implanted line) and urinary catheter was recorded. Relevant therapies,includingcytotoxicchemotherapy,systemic corticoste-roids,immunosuppressiveagents or radiationtherapy,in the3 monthspriortoBSIweredocumented.AccordingtotheNational HealthcareSafetyNetwork(NHSN),theprobablesourceofBSIwas assessedaccording tothe available clinicaland microbiological information and was classified by the investigators using the

following categories: genitourinarytract; respiratorytract; gas-trointestinal; skin and soft tissue; surgical site; line-related; endocarditis;orunknown[21].Lengthofstaywascalculatedasthe numberofdaysfromthedateofhospitaladmissiontothedateof discharge.Inaddition,clinicalmanagementofindwelling cathe-ters, initiation of antimicrobial therapy and various clinical outcomes related to the BSI were also recorded. Information about the reporting of positive blood cultures by the clinical microbiologylaboratory (includingtimeofblood culture collec-tion,timeofmicrobiologypositivityandtimeofcommunicationof thepreliminaryresultstophysicians)wascollectedinthestudy dataform.Twowaysofreportingofthebloodculturesresultswere considered[22]:

Earlyalertfromtheclinicalmicrobiologylaboratory:definedas aphonecallfromthemicrobiologisttothephysicianincharge abouttheidentifiedspeciesgrowinginthebloodculture.The physicianinchargewasinformedbytelephoneabouttheresult of species identification before completion of antimicrobial susceptibilitytesting;and

Conventionalreporting:cliniciansreceivedinformation about theisolatedorganismonlywhenthebloodculturereportwas completed and returned fromthe laboratory to the hospital ward.

Thetimefrombloodculturecollectiontoreportofbloodculture resultswascalculatedandexpressedinhoursandwasrepresented by(i)thetimefrombloodculturecollectionand thephonecall fromthemicrobiologistiftheearlyalertwasreceivedor(ii)the timefrombloodculturecollectiontothestandardreceiptofblood cultureswhenconventionalreportingwasperformed.

2.2.Definitions

ABSIwasdefinedaccordingtothestandarddefinitionsofthe US Centers for Disease Controland Prevention (CDC) [23]. BSI episodes were classified as: (i) nosocomial, if a positive blood culturewasobtainedfrompatientswhohadbeenhospitalisedfor 48h;(ii)healthcare-associated,ifapositiveblood culturewas obtainedfromapatientatthetimeofhospitaladmissionorwithin 48hofadmissionifthepatient(a)hadreceivedintravenous(i.v.) therapyathomeorhadself-administeredi.v.medicaltherapyin the 30 days before the BSI, (b) had received wound care or specialisednursingcarethrough a healthcareagency, family or friends, (c) had attended a hospital or haemodialysis clinic or receivedi.v.chemotherapyinthe30daysbeforetheBSI,(d)was hospitalisedinanacute-carehospitalfor2daysinthe90days beforetheBSIor(e)residedinanursinghomeorlong-termcare facility;and(iii)community-acquired,ifapositivebloodculture obtained at the time of hospital admission or within 48h of hospital admission for patients who didnot _fit thecriteria for healthcare-associatedinfection[24].Neutropeniawasdefinedas anabsoluteneutrophilcountof<500cells/mm3_{within30 days}

beforetheBSI[25].Immunosuppressivetherapywas de_finedas theuseofsteroids(prednisolone>0.5mg/kg/dayorequivalentfor >1month),chemotherapyoranti-tumournecrosisfactor, cyclo-phosphamide,azathioprine,methotrexate,mycophenolatemofetil orcalcineurininhibitortherapywithinthepast3months.Chronic kidneydiseasewasdefinedaccordingtoKidneyDiseaseOutcomes QualityInitiative(K/DOQI)guidelines[26].

2.3.Microbiologicalanalysis

Microbiologicalexaminationwasperformedonsputum,urine andbloodsamplesaccordingtostandardsofpractice.Allisolates frompatientswereidentifiedinacentralmicrobiologylaboratory

from each study site using automated methods such as BacT/ ALERT1(bioMérieux;Marcy-l’Étoile,France)andVITEK12broth microdilution(bioMérieux)accordingtoEuropeanCommitteeon Antimicrobial Susceptibility Testing (EUCAST) interpretative standards[27].Insome,butnotall,centresbacterialidentification was carried out by matrix-assisted laser desorption/ionisation time-of-flightmassspectrometry(MALDI-TOF/MS)(VITEK1MS; bioMérieux).

2.4.Aimofthestudyandstatisticalanalysis

The primary goal of this study was to describe the clinical featuresandoutcomesofBSIepisodescausedbyEnterococcusspp. inpatientshospitalisedinItalianinternalmedicinewardsandto identifyriskfactorsforin-hospitalmortalityamongthesepatients. The original SNOOPII study expected the inclusion of a minimum of 15 consecutive patients with BSI from each participatingcentre. Of these cases, patients withenterococcal BSIwereselectedforthecurrentstudy.

Statisticalanalysiswasperformedusingcommerciallyavailable statisticalsoftwarepackages[IBMSPSSStatisticsv.20.0(IBMCorp., Armonk, NY)and Rv.3.0.2 (R DevelopmentCoreTeam,Vienna, Austria)].ContinuousvariableswerecomparedbyStudent’st-test if normally distributed and by Mann–Whitney U-test if non-normallydistributed.Categoricalvariableswereevaluatedusingχ2

testortwo-tailedFisher’sexacttest.Values forcontinuousand categorical variables are expressed as the meanstandard deviation or median [interquartile range (IQR)] and as the percentage of the group from which they were derived, respectively. Multivariate analysis to identify independent risk factorsfor in-hospital mortalitywas performedusinga logistic regressionmodel.

VariableswithaP-valueof<0.10andthoseconsideredclinically relevantaccordingtoavailableliteraturedatawereincludedina stepwiseCox proportionalhazardmodel.Independentvariables withP<0.05afterstepwiseeliminationwereincludedinthefinal multivariateregression.Thehazardratio(HR)and95%confidence interval (CI) were calculated to evaluate the strength of any association.Statisticalsignificancewasestablishedat 0.05.All reportedP-valuesaretwo-tailed.

3.Results

Fig.1 shows the flowchart of the study.During the study period, a total of 533 patients with BSI were included in the SNOOPIIstudy.Amongthese,257episodes(48.2%)werecaused byGram-positivebacteria.Enterococcusspp.wereresponsiblefor 41BSIs (7.7%) observedin internalmedicine wards.Overall,35 (85.4%)weremonomicrobialand6(14.6%)werepolymicrobial.Of the41enterococcalBSIpatients, 28(68.3%)hadmonomicrobial BSI caused by E. faecalis and 4 (9.8%) had monomicrobial BSI causedbyE.faecium.ThreemonomicrobialBSIswerecausedby Enterococcus avium, Enterococcus casseliflavus and Enterococcus gallinarum,respectively.OfthesixpolymicrobialBSIs,twowere associatedwithcultureofmethicillin-susceptibleStaphylococcus aureus, one with Staphylococcus epidermidis (considered a skin contaminant), one with Klebsiella pneumoniae and two with Candidaalbicans.Allcases,withtheexceptionof onecaseof E. faecium–Candidaco-infection,receivedappropriatetherapy with-inthefirst48hfrombloodculturecollection.Inallcases,onlythe firstsetofbloodcultureswaspositiveformorethanone micro-organism,whilsttheremainingbloodcultureswerepositiveonly forEnterococcusspp.

Five isolates (12.2%) were resistant to ampicillin and four isolates(9.8%)wereresistanttovancomycin(twoE.faecium,oneE. casseliflavusandoneE.gallinarum).

Table 1 showsthe clinical features and outcomesof the41

patientswithenterococcalBSI.Patientsincludedinthestudywere elderlywithamedianageof73years (IQR66–85.5years).The majority of episodes were community-onset, with 17 (41.5%) healthcare-associated and 10(24.4%) community-acquiredBSIs. The most commonsources of infectionwerethe genitourinary tract (34.1%) and gastrointestinal tract (22.0%), followed by primary BSI (19.5%).Clinicians received anearly alert fromthe microbiologylaboratoryfor20patients(48.8%).Themedian(IQR) timefrombloodculturecollectiontoknowledgeoftheresultswas shorterinthecaseofearlyalertfrommicrobiologycomparedwith conventional reporting [39h (24–48h) vs. 96h (72–120h); P<0.001]. After receiving the phone call from microbiology, clinicians decided to change the antimicrobial therapy in 10 patients: 4 patients discontinued fluoroquinolones and started ampicillin plus aminoglycosides (n=2), imipenem (n=1) or piperacillin/tazobactam (TZP) (n=1), respectively; 2 patients discontinuedTZPand started linezolid;3patientsdiscontinued cephalosporinsandstartedteicoplanin(n=2)orampicillin(n=1), respectively; and 1 patient added aminoglycosides to their ampicillin therapy. All patients who underwent a change in antimicrobialtherapyaftertheearlyalertfromlaboratorysurvived (datanotshown).

Thein-hospitalmortalityratewas24.4%(10/41).Table2shows acomparisonbetweensurvivorsandnon-survivors.Non-survivors hadahighermedianage(84.5yearsvs.70years;P=0.022),were morefrequentlyaffectedbyahealthcare-associatedBSI(70.0%vs. 32.3%;P=0.035), hadahigher medianSequentialOrganFailure Assessment(SOFA)scoreatpresentation(11vs.2;P<0.001)and weremorecommonlyaffectedbymultipleco-morbidities(median Charlsonco-morbidityindex,4vs.3;P=0.049).Conversely,among survivors,cliniciansmorefrequentlyreceivedanalert communi-cation from the microbiology laboratory regarding the blood culturepositivity(58.1%vs.20.0%;P=0.036).

Inthemultivariateanalysis(Table3),polymicrobialinfection (HR=9.100,95%CI1.295–63.949;P=0.026),age(HR=1.261,95%CI 1.029–1.546;P=0.025)andSOFAscore(HR=1.244,95%CI1.051– 1.474; P=0.011) were risk factors for in-hospital mortality. Conversely, an alert from the clinical microbiology laboratory representedtheonlyfactorindependentlyassociatedwithsurvival (HR=0.073,95%CI0.007–0.805;P=0.033).

4.Discussion

Thisstudydescribestheepidemiologyandclinicalfeaturesof elderlypatientswithenterococcalBSIhospitalisedin31different internalmedicinewards.Moreover,factorsindependently associ-ated with in-hospital mortality among these patients were

identified. Previously knownfactors, suchas presence of poly-microbialinfection,ageand SOFA score,wereidentifiedasrisk factorsformortality.Moreover,anewpeculiarvariable,anearly alertfrom themicrobiology beforethereceiptofblood culture results,improvedtheoutcomeofthesepatients.

Themostimportantfindingofthecurrentstudyisthateffective communicationbetweentheclinicalmicrobiologylaboratoryand prescribing physicians could promote early optimisation of antimicrobialregimensresultinginanimprovedoutcome.Infact, we showed that receiving an early communication from the microbiologylaboratoryabouttheisolatedpathogenallowedearly modificationofthetreatmentregimen.Sincedelayindeliveryof appropriateantibiotictherapyisassociatedwithhighermortality

[18],receiptofanearlycommunicationfromthelaboratorycould be essential in suchcases. Processes involving communication betweenlaboratoriesandclinicalunitsarethemosterror-prone

partsoflaboratorytesting[28].Bloodcultureresultsmayremain pendingforseveraldays,butreportingsigni_ficanttestresultsto clinicalstaffbyphoneorinwrittenformcouldbecrucialtoguide antimicrobialtherapy.Thetimelinessandaccuracyofpreliminary bloodculturereportscouldaffectpatientoutcomes[22,29].

Table1

Clinicalfeaturesandoutcomesofpatientswithbloodstreaminfection(BSI)dueto Enterococcusspp.(N=41).

n(%)a Demographiccharacteristics

Age(years)[median(IQR)] 73(66–85.5)

Age65years 33(80.5)

Malesex 20(48.8)

Hospitalisationinlast3months 20(48.8) Antibiotictreatmentinlast3months 19(46.3) Meansofacquisition Community-acquired 10(24.4) Healthcare-associated 17(41.5) Nosocomial 14(34.1) Co-morbidconditions Cardiovasculardisease 28(68.3) Diabetes 16(39.0) COPD 11(26.8) Solidmalignancy 10(24.4)

Chronickidneydisease 9(22.0)

Liverdisease 6(14.6)

Neutropenia 1(2.4)

Charlsoncomorbidityindex[median(IQR)] 3(2–5) SOFAscore[median(IQR)] 3(1–8) Clinicalmanifestations Fever(>38_{C) 38(92.7)} Hypotensionb 7(17.1) Tachycardia 30(73.2) Tachypnoea 17(41.5) PaO2/FiO2<300 9(22.0)

Alteredmentalstatus 16(39.0) Hepaticdysfunctionc 4(9.8) Renaldysfunctiond 12(29.3) Alteredcoagulatione 6(14.6) Lactate>2mmol/L 11(26.8) Sourceofinfection Genitourinarytract 14(34.1) Respiratorytract 2(4.9) Gastrointestinaltract 9(22.0) Skinorsofttissue 4(9.8)

Endocarditis 4(9.8)

PrimaryBSI 8(19.5)

Septicshock 6(14.6)

Lengthofhospitalstay(days)[median(IQR)] 15(8–30) Alertfromclinicalmicrobiologylaboratory 20(48.8) In-hospitalmortality 10(24.4) IQR,interquartilerange;COPD, chronic obstructive pulmonary disease;SOFA, SequentialOrganFailureAssessment;PaO2,partialpressureofarterialoxygen;FiO2,

percentageofinspiredoxygen.

a_{Dataaren(%)unlessotherwisestated.}

b _{Definedassystolicbloodpressure(SBP)<90mmHg,meanarterialpressure}

<70mmHgoradecreaseinSBP>40mmHg.

c_Defined

asbilirubin>2mg/dLortransaminases>2theupperlimitofnormal.

d _Defined

as urine output <0.5mL/kg/h for 2h despite adequate fluid resuscitationorincreaseincreatinine>0.5mg/dLor44.2mmol/L.

e_{Defined as international normalised ratio (INR) >1.5, activated partial}

thromboplastintime(aPTT)>60sorplateletcount<100103_/mL.

Table2

Comparisonbetweensurvivorsandnon-survivorsinpatientswithbloodstream infection(BSI)duetoEnterococcusspp.a

Survivors (N=31) Non-survivors (N=10) P-value Demographiccharacteristics

Age(years)[median(IQR)] 70(65–81) 84.5(77.25–88) 0.022* Malesex 14(45.2) 6(60.0) 0.414 Hospitalisationinlast3months 11(35.5) 9(90.0) 0.003* Antibiotictreatmentinlast3

months 13(41.9) 6(60.0) 0.319 Meansofacquisition Community-acquired 10(32.3) 0 0.039* Healthcare-associated 10(32.3) 7(70.0) 0.035* Nosocomial 11(35.5) 3(30.0) 0.750 Co-morbidconditions Cardiovasculardisease 20(64.5) 8(80.0) 0.360 Diabetes 11(35.5) 5(50.0) 0.413 COPD 7(22.6) 4(40.0) 0.280 Solidmalignancy 6(19.4) 4(40.0) 0.186 Chronickidneydisease 7(22.6) 2(20.0) 0.864 Liverdisease 4(12.9) 2(20.0) 0.581 Neutropenia – 1(10.0) 0.075 SOFAscore[median(IQR)] 2(1–4) 11(6–14.25) <0.00* Charlsoncomorbidityindex

[median(IQR)] 3(2–4) 4(3–8) 0.049* Sourceofinfection Genitourinarytract 9(29.0) 5(50.0) 0.077 Respiratorytract 2(6.5) 0 0.410 Gastrointestinaltract 8(25.8) 1(10.0) 0.294 Skinorsofttissue 3(9.7) 1(10.0) 0.976 Endocarditis 2(6.5) 2(20.0) 0.209 PrimaryBSI 7(22.6) 1(10.0) 0.383 Septicshock 3(9.7) 3(30.0) 0.114 Alertfromclinicalmicrobiology

laboratory 18(58.1) 2(20.0) 0.036* Speciesdistribution Enterococcusfaecalis 21(67.7) 7(70.0) 0.894 Enterococcusfaecium 4(12.9) 0 0.232 OtherEnterococcusspp. 3(9.7) 0 0.307 Polymicrobial 3(9.7) 3(30.0) 0.114 Antimicrobialsusceptibilityprofile

Ampicillin-susceptible 24(77.4) 8(80.0) 0.864 Ampicillin-resistant 3(9.7) 2(20.0) 0.386 Vancomycin-resistant 4(12.9) 0 0.232 Definitiveantimicrobialtherapy

AmpicillinAG 9(29.0) 2(20.0) 0.575 TZPAG 4(12.9) 0 0.232 CarbapenemAG 9(29.0) 3(30.0) 0.953 TeicoplaninAG 2(6.5) 3(30.0) 0.048* VancomycinAG 3(9.7) 0 0.307 Linezolid 2(6.5) 0 0.410 Others 2(6.5) 2(20.0) 0.209 IQR,interquartilerange;COPD,chronicobstructivepulmonarydisease;SOFA,Sequential OrganFailureAssessment;AG,aminoglycoside;TZP,piperacillin/tazobactam.

a _{Dataaren(%)unlessotherwisestated.} * _{Statisticallysignificant(P0.05).}

Table3

Coxregressionanalysisofriskfactorsforin-hospitalmortalityinpatientswith bloodstreaminfectionduetoEnterococcusspp.

Riskfactor HR(95%CI) P-value Polymicrobialinfection 9.100(1.295–63.949) 0.026* Age 1.261(1.029–1.546) 0.025* SOFAscore 1.244(1.051–1.474) 0.011* Alertfromclinicalmicrobiologylaboratory 0.073(0.007–0.805) 0.033* HR, hazard ratio; CI, confidence interval; SOFA, Sequential Organ Failure Assessment.

*

However, this procedure is often challenging. A recent study highlightedthekeycommunicationbarriersbetweenmicrobiology laboratories and clinical units[30]. Disruption at the interface between laboratoriesandclinicalunits,mutuallackofinsightintoeachother’s areaofexpertise,andlimitedlaboratoryservicesareallbarriersthat needtobeaddressedinordertoimprovecommunication[30].For thesereasons,the earlyalertfromthemicrobiologylaboratoryto clinicalunitsshouldrepresentoneofthekeypointsofantimicrobial stewardshipprogrammes(ASPs).Ononehand,establishingASPsin hospitalscanbeanefficientframeworktofacilitatecommunication andto increaseinsight betweenthe microbiologylaboratoryand clinicalstaff.Ontheother hand,includinganinterventionofearly alert from the microbiology laboratory to clinical wards could guarantee the effectivenessofASPs.

In thecurrentstudy,in-hospitalmortality ofpatientswithBSIdue to Enterococcus spp. was in linewithprevious reports in other hospitalsettings[2,17].BSIsduetoEnterococcusspp.areassociated with different rates of mortality depending on the source of infection,withgastrointestinalsourcesshowingthehighestcase fatality rates [2,31]. Similarly, infective endocarditis caused by Enterococcus spp. is responsible for high mortality [16]. Other factors,suchasthepresenceofmultipleco-morbidities, malignan-cy,diabetes, inflammatoryboweldiseaseandcardiovascularfailure, negativelyaffecttheoutcome ofpatientswithenterococcalBSI

[12,16].Polymicrobialinfection,ageandSOFAscorewereidentified

asfactorsindependentlyassociatedwithmortality.Ithasrecently beendemonstratedthatE.faecaliscanfacilitateco-infectionwith othermicro-organisms,inparticularwithEscherichiacoli,through the suppression of nuclear factor kappa-B (NF-

k

B)-driven responsesinmacrophages[32]andtheincreaseofmetaboliccue thatfacilitatesE.coligrowth [33].Enterococcus spp.canalsobe foundtogetherwithCandidaspp.whenthereistranslocationof intestinal pathogens from the gut into the bloodstream [31]. Besides,independentlyfrom theco-isolated pathogen, thepresence ofEnterococcusspp.asapartofapolymicrobialinfectionappearsto increasetheriskofshort-termmortality[34].

Thisstudyhassomelimitations.First,thesamplesizeisnotso large, but slightly lower than other studies [12]. Second, only patientshospitalisedinmedicalwardswereincluded.Thus,these findingsare notgeneralisabletoothersettings suchas surgical wardsorICUs.Finally,theinclusion ofpolymicrobial infections couldhaveinfluencedthestudyresults.However,thesixcasesof polymicrobialinfectionwererepresentedbyBSIinwhichonlyone bloodculturewaspositivefortwomicro-organismswhereasthe subsequentcultureswerepositiveonlyforEnterococcusspp.,and allcaseswiththeexceptionofonecaseofE.faecium–Candida co-infectionreceivedadequateantimicrobialtherapy.Thestrengthsof thisstudyarethatit isa multicentrestudyand thatdatawere prospectivelycollectedusingstandardprocedures.

Inconclusion,BSIduetoEnterococcusspp.inelderlypatientsis associatedwithahighriskofin-hospitalmortality.Presenceofa polymicrobialinfection,ageandSOFAscoreonthedayofBSIwere factorsassociatedwithincreasedmortality.Conversely,anearly alert from the microbiology laboratory before obtaining blood culturesresultscanhelpclinicianstoselectappropriate antimicro-bialtherapy,thusimprovingpatientsurvival.Rapidandefficacious communicationfromthemicrobiologylaboratorytoclinicalunitsis crucialandshouldbeimplementedinhospitalsettings.

Funding None. Competinginterests Nonedeclared. Ethicalapproval Notrequired. References

[1]Low DE, Keller N, Barth A, Jones RN.Clinical prevalence, antimicrobial susceptibility,andgeographicresistancepatternsofenterococci:resultsfrom theSENTRYAntimicrobialSurveillanceProgram,1997–1999.ClinInfectDis 2001;32(Suppl.2):S133–45.

[2]BillingtonEO,PhangSH,GregsonDB,PitoutJD,RossT,ChurchDL, etal. Incidence, riskfactors, andoutcomesforEnterococcus spp.bloodstream infections:apopulation-basedstudy.IntJInfectDis2014;26:76–82. [3]deKrakerME,JarlierV,MonenJC,HeuerOE,vandeSandeN,GrundmannH.

Thechanging epidemiologyof bacteraemiasin Europe:trends from the EuropeanAntimicrobialResistanceSurveillanceSystem.ClinMicrobiolInfect 2013;19:860–8.

[4]EliopoulosGM.Increasingproblemsinthetherapyofenterococcalinfections. EurJClinMicrobiolInfectDis1993;12:409–12.

[5]WillemsRJ,TopJ,vanSantenM,RobinsonDA,CoqueTM,BaqueroF,etal. Globalspreadofvancomycin-resistantEnterococcusfaeciumfromdistinct nosocomialgeneticcomplex.EmergInfectDis2005;11:821–8.

[6]Wisplinghoff H,Bischoff T,TallantS,SeifertH,Wenzel RP, EdmondMB. NosocomialbloodstreaminfectionsinUShospitals:analysisof24,179cases from a prospective nationwide surveillance study. Clin Infect Dis 2004;39:309–17.

[7]Caballero-GranadoFJ,BecerrilB,CuberosL,BernabeuM,CisnerosJM,PachónJ. Attributablemortalityrateandduration ofhospitalstayassociated with enterococcalbacteremia.ClinInfectDis2001;32:587–94.

[8]OrsiGB,FalconeM,VendittiM.Surveillanceandmanagementof multidrug-resistantmicroorganisms.ExpertRevAntiInfectTher2011;9:653–79. [9]HidronAI,EdwardsJR,PatelJ,HoranTC,SievertDM,PollockDA,etal.NHSN

annualupdate:antimicrobial-resistantpathogensassociatedwith healthcare-associated infections:annual summaryof datareported tothe National HealthcareSafetyNetworkattheCentersforDiseaseControlandPrevention, 2006–2007.InfectControlHospEpidemiol2008;29:996–1011.

[10]Lee T, Pang S, Abraham S, Coombs GW. Antimicrobial-resistant CC17 Enterococcusfaecium:thepast,thepresentandthefuture.JGlobAntimicrob Resist2018;16:36–47.

[11]McBrideSJ,UptonA,RobertsSA.Clinicalcharacteristicsandoutcomesof patientswithvancomycin-susceptibleEnterococcusfaecalisandEnterococcus faeciumbacteraemia—afive-yearretrospectivereview.EurJClinMicrobiol InfectDis2010;29:107–14.

[12]Bar K,Wisplinghoff H, WenzelRP, Bearman GM,Edmond MB.Systemic inflammatory response syndrome in adult patients with nosocomial bloodstreaminfectionsduetoenterococci.BMCInfectDis2006;6:145. [13]SongX,SrinivasanA,PlautD,PerlTM.Effectofnosocomial

vancomycin-resistantenterococcalbacteremiaonmortality,lengthofstay,andcosts.Infect ControlHospEpidemiol2003;24:251–6.

[14]OngDS,BontenMJ,SafdariK,SpitoniC, FrenckenJF,WitteveenE,etal. Epidemiology, management, and risk-adjusted mortality of ICU-acquired enterococcalbacteremia.ClinInfectDis2015;61:1413–20.

[15]FalconeM,RussoA,VendittiM.Optimizingantibiotictherapyofbacteremia andendocarditisduetostaphylococciandenterococci:newinsightsand evidencefromtheliterature.JInfectChemother2015;21:330–9.

[16]PericàsJM,CerveraC,MorenoA,Garcia-de-la-MàriaC,AlmelaM,FalcesC, etal.OutcomeofEnterococcusfaecalisinfectiveendocarditisaccordingtothe lengthofantibiotictherapy:preliminarydatafromacohortof78patients. PLoSOne2018;13:e0192387Erratumin:PLoSOne2018;13:e0196317. [17]PohCH,OhHM,TanAL.Epidemiologyandclinicaloutcomeofenterococcal

bacteraemiainanacutecarehospital.JInfect2006;52:383–6.

[18]Fernández-HidalgoN,AlmiranteB,GavaldàJ,GurguiM,PeñaC,deAlarcónA,etal. Ampicillinplusceftriaxoneisaseffectiveasampicillinplusgentamicinfortreating Enterococcusfaecalisinfectiveendocarditis.ClinInfectDis2013;56:1261–8. [19]MazzoneA,DentaliF,LaReginaM,FogliaE,GambacortaM,GaragiolaE,etal.

Clinicalfeatures,short-termmortality,andprognosticriskfactorsofseptic patientsadmittedtointernalmedicineunits:resultsofanItalianmulticenter prospectivestudy.Medicine(Baltimore)2016;95:e2124.

[20]FalconeM,TiseoG,DentaliF,LaReginaM,FogliaE,GambacortaM,etal. Predictingresistantetiologyin hospitalizedpatients withblood cultures positiveforGram-negativebacilli.EurJInternMed2018;53:21–8. [21]USCentersforDiseaseControlandPrevention(CDC).CDC/NHSNsurveillance

definitionsforspecifictypesofinfections.2019..[Accessed5July2019]

https://www.cdc.gov/nhsn/PDFs/pscManual/17pscNosInfDef_current.pdf. [22]Bouza E,Sousa D, Muñoz P, Rodríguez-Créixems M,FronC, LechuzJG.

Bloodstreaminfections:atrialoftheimpactofdifferentmethodsofreporting positivebloodcultureresults.ClinInfectDis2004;39:1161–9.

[23]HoranTC,AndrusM,DudeckMA.CDC/NHSNsurveillancedefinitionofhealth care-associatedinfectionandcriteriaforspecifictypesofinfectionsinthe acutecaresetting.AmJInfectControl2008;36:309–32.

[24]FriedmanND,KayeKS,StoutJE,McGarrySA,TrivetteSL,BriggsJP,etal.Health care-associatedbloodstreaminfectionsinadults:areasontochangethe accepted definition of community-acquired infections. Ann Intern Med 2002;137:791–7.

[25]FreifeldAG,BowEJ,SepkowitzKA,BoeckhMJ,ItoJI,MullenCA,etal.Clinical practiceguidelinefortheuseofantimicrobialagentsinneutropenicpatients withcancer:2010updatebytheInfectiousDiseasesSocietyofAmerica.Clin InfectDis2011;52:e56–93.

[26]NationalKidneyFoundation.K/DOQIclinicalpracticeguidelinesforchronic kidneydisease:evaluation,classification,andstratification.AmJKidneyDis 2002;39(2Suppl.1):S1–266.

[27]MagiorakosAP,SrinivasanA,CareyRB,CarmeliY,FalagasME,GiskeCG,etal. Multidrug-resistant,extensivelydrug-resistantandpandrug-resistant bacte-ria:an internationalexpertproposalfor interimstandard definitionsfor acquiredresistance.ClinMicrobiolInfect2012;18:268–81.

[28]HawkinsR.Managingthepre-andpost-analyticalphasesofthetotaltesting process.AnnLabMed2012;32:5–16.

[29]Schifman RB, MeierFA,SouersRJ.Timeliness andaccuracy ofreporting preliminarybloodcultureresults:aCollegeofAmericanPathologistsQ-probes studyof65institutions.ArchPatholLabMed2015;139:621–6.

[30]SkodvinB,AaseK,BrekkenAL,CharaniE,LindemannPC,SmithI.Addressing the key communication barriers betweenmicrobiology laboratoriesand clinicalunits:aqualitativestudy.JAntimicrobChemother2017;72:2666–72. [31]FalconeM,RussoA,IraciF,CarfagnaP,GoldoniP,VulloV,etal.Riskfactorsand outcomes for bloodstream infections secondary to Clostridium difficile infection.AntimicrobAgentsChemother2015;60:252–7.

[32]TienBYQ,GohHMS,ChongKKL,Bhaduri-TagoreS,HolecS,DressR,etal. Enterococcusfaecalispromotesinnateimmunesuppressionandpolymicrobial catheter-associated urinary tract infection. Infect Immun 2017;85: pii: e00378-17.

[33]KeoghD,TayWH,HoYY,DaleJL,ChenS,UmashankarS,etal.Enterococcal metabolitecuesfacilitateinterspeciesnichemodulationandpolymicrobial infection.CellHostMicrobe2016;20:493–503.

[34]CeciM,DelpechG,SparoM,MezzinaV,SánchezBruniS,BaldacciniB.Clinical andmicrobiologicalfeaturesofbacteremiacausedbyEnterococcusfaecalis.J InfectDevCtries2015;9:1195–203.