Triage tool for suspected COVID-19 patients in the emergency room: AIFELL score

brazjinfectdis2020;24(5):458–461

w w w . e l s e v i e r . c o m / l o c a t e / b j i d

The

Brazilian

Journal

of

INFECTIOUS

DISEASES

Brief

communication

Triage

tool

for

suspected

COVID-19

patients

in

the

emergency

room:

AIFELL

score

Ian

Levenfus

_,

_Enrico

_Ullmann

_,

_Edouard

_Battegay

_,

Macé

M.

Schuurmans

a_{UniversityHospitalZurich,DivisionofInternalMedicine,Zurich,Switzerland}

b_{UniversityofZurich,FacultyofMedicine,Zurich,Switzerland}

c_{UniversityofLeipzigMedicalCenter,PsychotherapyandPsychosomatics,DepartmentofPediatricPsychiatry,Leipzig,Germany}

d_{TechnischeUniveristatDresden,DepartmentofMedicine,Dresden,Germany}

e_{SouthUralStateUniversity,DepartmentofMedicalBiology,Chelyabinsk,Russia}

f_{UniversityHospitalZurich,DivisionofPulmonology,Zurich,Switzerland}

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received31May2020 Accepted19July2020

Availableonline20August2020

Keywords: COVID-19 SARS-CoV-2 Score Emergency

a

b

s

t

r

a

c

t

Clinicalpredictionscoressupporttheassessmentofpatientsintheemergencysettingto determinetheneedforfurtherdiagnosticandtherapeuticsteps.Duringthecurrent COVID-19pandemic,physiciansinemergencyrooms(ER)ofmanyhospitalshaveaconsiderably higherpatientloadandneedtodecidewithinashorttimeframewhomtohospitalize.Based onourclinicalexperiencesindealingwithCOVID-19patientsattheUniversityHospitalin Zurich,wecreatedatriagescorewiththeacronym ¨AIFELL ¨consistingofclinical,radiological andlaboratoryfindings.

Thescorewasthenevaluatedinaretrospectiveanalysisof122consecutivepatientswith suspectedCOVID-19fromMarch untilmid-April2020.Descriptivestatistics,Student’s

t-test,ANOVAandScheffe’spost-hocanalysisconfirmedthediagnosticpowerofthescore. TheresultssuggestthattheAIFELLscorehaspotentialasatriagetoolintheERsetting intendedtoselectprobableCOVID-19casesforhospitalizationinspontaneouslypresenting orreferredpatientswithacuterespiratorysymptoms.

©2020SociedadeBrasileiradeInfectologia.PublishedbyElsevierEspa ˜na,S.L.U.Thisisan openaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).

Due to the worldwide spread of SARS-CoV-2 and rapidly increasing numbers of infections, the novel coronavirus becameaconsiderablestrainforemergencyrooms(ER), espe-ciallywhenseveralsuspectedcaseswithunspecificgeneral orrespiratorysymptomsarriveatthesametime. Identifica-tionofmorecriticalpatientsintheERforhospitalizationisa challengesincethedetectionofSARS-CoV-2in

nasopharyn-∗ _{Correspondingauthor.}

E-mailaddresses:[email protected],[email protected](I.Levenfus).

gealswabsbyquantitativepolymerasechainreaction(qPCR) stillrequiresmanyhours(>6hinoursetting).Therefore,the qPCRresultcurrentlycannotbeusedinthefrontlinesetting todecidewhomtohospitalizeandwhocanbemanagedas anoutpatient.RapidpointofcaretestsforSARS-CoV-2were beingdevelopedatthetimeofthestudybutwerenotvalidated forroutineuse.1

https://doi.org/10.1016/j.bjid.2020.07.003

1413-8670/©2020SociedadeBrasileiradeInfectologia.PublishedbyElsevierEspa ˜na,S.L.U.ThisisanopenaccessarticleundertheCCBY license(http://creativecommons.org/licenses/by/4.0/).

brazj infect dis.2020;24(5):458–461

459

Table1–DistributionofincludedpatientsandtheirclinicallyassignedAIFELLscores.

ConsecutivepatientswithsuspectedCOVID-19admittedviatheERfromMarchthroughmid-April2020 (n=122;72males,50females)

NasopharyngealqPCRresult SARS-CoV-2positive(n=70,age 60.6years±14.2)

SARS-CoV-2negative(n=52,57.9 years±17.9)

ComponentsoftheAIFELLscore:Alteredsmellortaste(yes/no), Inflammation(CRP≥30mg/L),Infiltrates(yes/no),Fever(≥38◦C), ElevatedLDH(>400U/L),Lymphocytopenia(absolutecount<1.45G/L)

Diagnosis COVID-19StageI

(n=10)

COVID-19Stage IIa(n=31)

COVID-19Stage IIb(n=29)

Otherrespiratoryproblemslike exacerbatedCOPD,bronchialasthma, bacterialpneumonia,aspiration pneumonitis,otherviralinfections (influenza,metapneumovirus),cardiac failure

MeanAIFELLscore 1.8±0.8 4.6±0.8 2.2±1.1

4.19±1.28

TotalnumbernAIFELLpositive(4–6points) 0 29(93.5%) 29(100%) 5(9.6%)

TotalnumbernAIFELLnegative(0–3points) 10(100%) 2(6.5%) 0 47(90.4%)

Legend:Resultsgivenasmean±SDunlessindicatedotherwise.StageIIIpatientswithprogressivesystemicinflammationwereusuallyadmitted directlytoICUfromnormalwardorotherhospitalsinoursetting,notthroughtheER.Therefore,theyarenotmentionedinthistable.SD, standarddeviation.

Asafrontlinephysician,whosetaskwastoevaluateand triagepatientsarrivingwithsymptomssuggestingCOVID-19 intheERcoronavirusunitoftheUniversityHospitalinZurich,

thefirstauthor wasconfrontedwiththeproblemofwhom

tochooseforhospitalizationduetoprobableCOVID-19and whomtodischargewhilsttheqPCRresultsoftheswabwere pending.

Thehospitalizationcriterionwastoselectpatientsatrisk fordevelopingmoreseveresymptomsleadingtorespiratory failure(COVIDStagesIIorIII2_{).Duringclinicalroutineworkin}

thefrontline unit,the questionofascorearosetosupport thetriageprocess andtoassist other physiciansinsimilar situations.

We therefore followed up consecutively hospitalized patientswithprovenCOVID-19inordertodetermineinitial features which may help distinguishing probable COVID-19 casesfrom other respiratory problems. In30 personally encounteredconsecutivelyhospitalizedpatientswith qPCR-provenCOVID-19studiedinitiallyasapilotcohort,wefound that elevated C-reactive protein (CRP) and lactate dehy-drogenase (LDH) levels as well as lymphocytopenia were characteristic laboratory patterns. Evidence obtained from literaturesearches usingthe keywords “COVID-19”, “SARS-CoV-2”, “laboratory”and “patients” inPubMed proved that theselaboratoryabnormalitieswereassociatedwith COVID-19.3–5_{Additionally,mostpatientsinthepilotcohortpresented}

withanelevatedbodytemperatureandshowedunilateralor bilateral pulmonaryinfiltrates inconventional chest radio-graphy. Several patients mentioned spontaneously having notedan attenuation ofsmell or taste and this symptom, althoughnotwidelyrecognizedasatypicalfeatureatthetime, wasconsideredtobeCOVID-19-associated.6 _Whenthe

rele-vantparaclinicalcomponentswereanalyzed,cut-off values becamemoreevidentandasimplescorewascreatedbased ontypicalclinicalinformationroutinelyavailableinourER.

TheAIFELLscoreincludesanAlteredsenseofsmell/taste,

Inflammation (C-reactive protein ≥30mg/L), radiological

Infiltrates,Fever(≥38.0◦_{C),ElevatedLactatedehydrogenase}

(LDH)levels(>400U/L)andLymphocytopenia(absolutecount <1.45G/L).Thescoreiscalculatedbyadding thenumberof criteria metatinitialpresentationinthe ER,whereaseach criterionequalsonepoint(scorerangefrom0to6points).

Toassessthescore,weapplieditretrospectivelyto consec-utivepatientswithsuspectedCOVID-19admittedviatheER fromMarchuntilmid-April2020.Onlythosecasesevaluated withchestimagingandabloodtestincludingatleasttwoof thethreeconsideredbloodparametersatpresentationinthe ER,whodidnotdeclinethegeneralresearchconsent,were included.Of122patientswithsuspectedCOVID-19,52cases turnedouttohaveotherrespiratoryproblems.

SARS-CoV-2 positive patients (n=70) were classified accordingtothestagessuggestedbySiddiqiandMehra (Sup-plementaryFigureS.1).2

ThestudywasapprovedbytheInstitutionalReviewBoard oftheCantonofZurich(CantonalEthicsCommittee,Nr. 2020-00854).Aftertestingfornormaldistributionandstandardized outliers,wecreatedanewvariablenamed“paraclinical mea-surements”byincludingthez-standardizedmeanvaluesof LDH,CRP,inverselypoledserumlymphocytesandauricular body temperature.Weafterwardssummarizedthisvariable with lung infiltrates seen by imaging and alterations of smell/taste indicated by patient history to get our predic-tor scorenamed“AIFELL”. Student’st-testsand an ANOVA withScheffe’spost-hoctestswereperformedforgroup com-parisons.Forallanalyses,weusedMSExcel2016andSPSS 24.

ThemeanageofourSARS-CoV-2positivesubjects(n=70) was60.6years±14.2(standarddeviation)vs.57.9years±17.9 ofourSARS-CoV2negativesubjects(n=52;t=.908;p=0.37). There were significantlydifferentAIFELLscoresbefore and afterz-standardization(t=5.77,p<0.001)betweenSARS-CoV-2 positivepatients(mean=2.43±0.15standarderror)and SARS-CoV-2 negativepatients(mean=1.30±0.12)(Supplementary FigureS.2).

A score of ≥4 points/criteria met at presentation was highly associated with qPCR-based SARS-CoV-2 detection

460

Table2–ANOVAofintergroupdifferencesoftheAIFELLscoregroups(1–6)usingobjectiveparaclinicalmeasurementsor

thewholearrayofAIFELLcomponents.

Paraclinicalmeasurements WholearrayofAIFELLcomponents AIFELLscorebyfrontlineexperiences ANOVA/Scheffe‘s

post-hoctest

ANOVA/Scheffe‘s posthoctest

ofpositivecomponents=points Mean±SD n df=116,F=10.55, p<0.0001 Mean±SD n df=116,F=31.525, p<0.0001 1 −0.54±0.28 12 1<2;1<3;1 <***4;1<*5;1 <**6 0.79±0.93 12 1<*4;1<*5;1 <*6;1<2;1<3 2 −0.21±0.43 26 2<3;2<4;2<*5; 2<6 1.18±0.55 26 2<3;2<*4;2<* 5;2<*6 3 −0.04±0.32 16 3<5;3<6 1.66±0.67 16 3<**5;3<*6; 4 0.01±0.65 31 4<**5;4<6 2.21±0.93 31 4<5;4<*6 5 0.52±0.42 24 2.77±0.83 24 6 0.30±0.38 8 4.30±0.38 8

Legend:Paraclinicalmeasurements=z-standardizedmeanvaluesofserumLDH,CRP,inverseabsolutelymphocytecountandtemperature measuredauricularly.WholearrayofAIFELLcomponents=sumofparaclinicalmeasurements,lunginfiltratesandalteredsmellortaste. *p<.001;**p<.01;***p=.04.Inonly86ofthe122includedcases,LDHvaluesweredetermined.Smellandtastealterationswereactivelymentioned bythepatientsandnotroutinelyaskedbythephysicians.Therefore,thenumberofpositivecasesisonly19.Patientswithoutanypositive componentsrelatingtotheAIFELLscore(0,n=5)werenotincludedinstatisticalgroupcomparisons.LDH,lactatedehydrogenase;CRP, C-reactiveprotein;SD,standarddeviation.

innasopharyngealswabsanddevelopmentofsymptomatic

COVID-19 (Stages II or III), thus justifying hospitalization. Scoresbetween0and3wereassociatedwithotherrespiratory conditions(Table1).

StageIII patients (severedisease) with extra-pulmonary systemic hyperinflammation, ARDSor symptoms ofshock wereusuallytransferredtointensivecareunit(ICU)from nor-malwardsorfromotherhospitals.DocumentedcasesofICU transfersofpatientswhodeterioratedinthecourseofdisease (progressionfromStageIItoStageIII)duringthehospitalstay (n=14)showedmeanAIFELLscoresatthedayofadmissionto ICUof5±0.68.

The ANOVA and post-hoc calculations were performed tosubstantiate groupdifferences moreclearly and verified significantdifferencesofevaluatedscorecomponentvalues betweendifferentscorevalues(Table2).Wefound,for exam-ple,thatthegroupofpatientswithanAIFELLscoreof5points hadhigherparaclinicalcomponentvaluesthanthegroupwith 2points(p<0.001).UsingthewholearrayofAIFELL compo-nents,itcould beshownthat thegroup with6pointshad significantlyhigherAIFELLcomponentvaluesthanthegroup with3points(p<0.0001).

Basedontheevaluationoftheinitialdataof30patients, wegeneratedtheAIFELLscoreasasimpletriageinstrument fortheERsettingconsistingoffrequentlyavailableelements likepatientsymptoms(fever,alteredsmellortaste),laboratory tests(differentialbloodcount,CRP,LDH)andimaging. After-wards,weevaluated itsdiagnostic performanceinalarger number of consecutive patients hospitalized for suspected COVID-19.

A hostrisk score dealing with comorbidities of COVID-19 patients7 _{as well as scores predicting critical illness}8

orhyperinflammation9_{inhospitalizedpatientswithproven}

COVID-19have been previously published.However, no ER triagescoretoidentifyprobableCOVID-19casesinmore criti-calstages(IIandIII)hasbeenproposedyet.TheAIFELLscore usesonlyfrequentlyobtaineddatausuallyavailableboth,in

theERandthegeneralpracticesetting.Otheradditional lab-oratory parameterslikeferritin,troponin10 _andD-dimers,11

whichareelevatedinseverecasesofCOVID-19,12_aswellas

more sophisticated parameters such asinterleukin (IL)-611

and solubleIL-2receptor13 may alsobeofinterest,but are not routinely evaluated orare notwidelyavailable. There-fore,theseadditionalparametersarelessapplicableforERor generalpracticetriagepurposes.

DuringtheCOVID-19pandemicandpartlyscarcemedical resources,theAIFELLscoremaybeusefulforselecting symp-tomaticCOVID-19casesfrompatientswithratherunspecific generalorrespiratorysymptomsintheERorgeneralpractice settingwhoshouldimmediatelygetaSARS-CoV-2swabdue tohigherprobabilityofthedisease.Thescoreisnotintended to identifyasymptomatic or oligosymptomatic SARS-CoV-2 infections (COVID-19 StageI),which generallycan bedealt withintheoutpatientsetting.

The major limitation of this work is the single-center evaluationofonlyalimitednumber ofpatients. Due toits retrospectivenature,somevalueswerenotobtainedfromall patients. For example,in36 ofall includedcasesLDH val-uesweremissingasitwasnotmeasuredineveryadmitted patient.Smellandtastealterationswereactivelymentioned bythepatientsandnotroutinelyaskedbythephysicians dur-ingthestudyperiod.

ThestrengthsoftheAIFELLscoreareitssimplicity, imme-diateavailabilityaswellaswideapplicabilityduetosimple components.TheAIFELLscoreobviouslyneedstobe prospec-tivelyappliedinlargercohortsofpatientstogainmorereliable dataregardingitsdiagnosticyield.

Author

contributions

IL:idea,datacollection,assemblyandinterpretation, statis-tical analysis, manuscriptwriting; EU: statistical expertise, adviceandanalysis,manuscriptwriting;EB:idea,manuscript

brazj infect dis.2020;24(5):458–461

461

correctionandapproval;MMS:idea,analysisand interpreta-tionofdata,manuscriptwriting,correctionandapproval.

ILandMMStaketheresponsibilityforthepaperasawhole.

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

Appendix

A.

Supplementary

data

Supplementarymaterialrelatedtothisarticlecanbefound, intheonlineversion,atdoi:10.1016/j.bjid.2020.07.003.

r

e

f

e

r

e

n

c

e

s

1. SheridanC.Fast,portabletestscomeonlinetocurb coronaviruspandemic.NatBiotechnol.2020, http://dx.doi.org/10.1038/d41587-020-00010-2. 2. SiddiqiHK,MehraMR.COVID-19illnessinnativeand

immunosuppressedstates:aclinical-therapeuticstaging proposal.JHearLungTransplant.2020,

http://dx.doi.org/10.1016/j.healun.2020.03.012.

3. ZhouF,YuT,DuR,etal.Clinicalcourseandriskfactorsfor mortalityofadultinpatientswithCOVID-19inWuhan,China: aretrospectivecohortstudy.Lancet.2020;6736:1–9.

4. Rodriguez-MoralesAJ,Cardona-OspinaJA,Gutiérrez-Ocampo E,etal.Clinical,laboratoryandimagingfeaturesofCOVID-19: asystematicreviewandmeta-analysis.TravelMedInfectDis. 2020:101623,http://dx.doi.org/10.1016/j.tmaid.2020.101623.

5.HuangC,WangY,LiX,etal.Clinicalfeaturesofpatients infectedwith2019novelcoronavirusinWuhan,China. Lancet.2020;395:497–506.

6.LechienJR,Chiesa-EstombaCM,DeSiatiDR,etal.Olfactory andgustatorydysfunctionsasaclinicalpresentationof mild-to-moderateformsofthecoronavirusdisease (COVID-19):amulticenterEuropeanstudy.EurArch Otorhinolaryngol.2020;2,

http://dx.doi.org/10.1007/s00405-020-05965-1. 7.ShiY,YuX,ZhaoH,WangH,ZhaoR,ShengJ.Host

susceptibilitytosevereCOVID-19andestablishmentofahost riskscore:findingsof487casesoutsideWuhan.CritCare. 2020;24:108.

8.LiangW,LiangH,OuL,etal.Developmentandvalidationofa clinicalriskscoretopredicttheoccurrenceofcriticalillness inhospitalizedpatientswithCOVID-19.JAMAInternMed. 2020,http://dx.doi.org/10.1001/jamainternmed.2020.2033. 9.MehtaP,McauleyDF,BrownM,etal.Correspondence

COVID-19:considercytokinestormsyndromesand.Lancet. 2020;6736:19–20.

10.LippiG,LavieCJ,Sanchis-GomarF.CardiactroponinIin patientswithcoronavirusdisease2019(COVID-19):evidence fromameta-analysis.ProgCardiovascDis2020.2019, http://dx.doi.org/10.1016/j.pcad.2020.03.001.

11.GaoY,LiT,HanM,etal.Diagnosticutilityofclinical laboratorydatadeterminationsforpatientswiththesevere COVID-19.JMedVirol.2020,

http://dx.doi.org/10.1002/jmv.25770.

12.VelavanTP,MeyerCG.MildversussevereCOVID-19: laboratorymarkers.IntJInfectDis.2020;95:304–7. 13.ChenG,WuD,GuoW,etal.Clinicalandimmunologic

featuresinsevereandmoderateformsofCoronavirus Disease2019.JClinInvest2020.2020,