rev bras hematol hemoter. 2015;37(1):67–68
Revista
Brasileira
de
Hematologia
e
Hemoterapia
Brazilian
Journal
of
Hematology
and
Hemotherapy
w w w . r b h h . o r g
Letter
to
the
Editor
Why
and
how
validate
criteria
by
manual
smear
review
to
improve
laboratory
productivity?
DearEditor,
Ireadthearticle“Arethereviewcriteriaforautomated
com-pletebloodcountsoftheInternationalSocietyofLaboratory
Hematologysuitable forall hematologylaboratories?”,
pub-lishedinRBHH2014;36(2);219–225,1andIwouldliketomake
somecommentsaboutit.
Thepurposeofthestudyisremarkableanditisknownthat
eachLaboratoryhastodefinetheirowncriteriaforreviewing
bloodsmears.Severalstudieshavereporteddifferent
expe-rienceswhenmanualmethodswerereplacedbyautomated
hematology systems, all ofwhich described improvements
in quality and precision, in addition to faster reportable
results.2–4Comaretal.appliedthereviewcriteriaofthe
Inter-nationalSocietyforLaboratoryHematologyintheirstudywith
someadaptations according to local requirements.A total
of1977wholeblood sampleswere analyzedontwo pieces
ofequipment and 100leukocyteswere countedbyasingle
observer.Amongotherresults,theauthorsreportedhighrates
ofmicroscopyreviewsandanunacceptablepercentageoffalse
negativeresults(>5%).
Someconsiderationsshouldbementionedaboutthis
eval-uation. First,and mostimportantinmyview,isrelated to
microscopicanalysis.AccordingtotheNational Committee
forClinical LaboratoryStandards(NCCLS)document
‘Refer-enceLeukocyte(WBC)DifferentialCount,H20-A2’5somerules
mustbefollowedinordertoobtainareliableanalysis,such
as:(i)threebloodfilmsfromeachspecimenshouldbe
pre-pared(twoareusedfortheprocedureandthethirdiskeptas
aspare);(ii)alargernumberofbloodfilmsmustbeprepared
forleukopenicsamples;(iii)200WBCshouldbecountedon
eachslide;and(iv)atleasttwoexaminersmustbeableto
clas-sifyallnormalandabnormalcells.Theserecommendations
werenotfollowedinthestudy,whichmayexplain,according
totheauthors,“theinadequateperformanceofbothpiecesof
equipment”.Otherstudies,usingtheNCCLSDocumentasan
evaluationprotocol,showeddifferentresults.Rusickaetal.6
testedtheefficiencyofflaggingusingtheXE-2100deviceand
showedthattherateoffalsenegativeresultsoftheimmature
granulocytecategorywasonly1%andtheefficiencyrateof
myeloidprecursorcelldetectionwas84%.Regardingthefalse
negativeresultsforblasts,theratewas1%(5samples),allof
themshowingleukocytecounts<2.5×109/L,andfourofthem
wereflaggedbytheatypicallymphflagorIGflag.Thismeans
that,althoughtheblastcellwasnotdetected,thebloodcell
smearshouldbeexaminedduetoadditionalflags. Another
study waspublishedbyStammingeretal.7 where800cells
were analyzedpersample.Theauthorsconsideredthe Left
Shifttobepositiveifbandcellswere>0.9×109/Lor6%and/or
aneutrophilproportion>80%.TheefficiencyoftheLeftShift
flagwas0.86(sensitivity0.53andspecificity0.92).Whenthe
combination LeftShiftplusneutrophiliawasapplied,there
wasasignificantincreaseintheefficiency(0.92),sensitivity
(0.83)andspecificity(0.92).Consideringthevariationsin
mor-phologicaldefinitionofthebandcell,thecountofthiscellin
theroutinelaboratoryisapparentlyunreliable.
Insummary,theprogressofhematologyautomationand
theachievedgoodlevelsofprecisionandtheaccuracyincell
countingareincontestable.Theexaminationand
identifica-tionofthousandsofcellsineachsample,theimprovementin
technologiesandtheincorporationofnewparametersprovide
reliable and applicableinformationfordiagnosis inseveral
clinicalconditions.Automationwillmostlikelynevertotally
replacethemicroscopicreviewofbloodcells.The
examina-tion ofred cell morphology is crucial for the diagnosis of
anemia,andautomationdoesnotprovidealltheinformation
thatispotentiallyimportanttothephysician.8Thepurpose
ofautomationistoprovidefasterreportableresults,toreduce
thetechnologisthands-ontime,inadditiontoprovidinghigh
qualityand precision.Eachlaboratorymustdefinethe best
criteriatoachievetheirperformancegoals.Severaltoolsand
guidelinesareavailabletoanalyzetheperformanceof
equip-ment,andtodefinethebestrulesforspecificneeds.
Conflicts
of
interest
Dr.HelenaGrottoisMedicalAdvisorofSysmexLatinAmerica
68
revbrashematolhemoter.2015;37(1):67–68r
e
f
e
r
e
n
c
e
s
1.ComarSR,MalvezziM,PasquiniR.Arethereviewcriteriafor automatedcompletebloodcountsoftheInternationalSociety oflaboratoryHematologysuitableforallhematology
laboratories?RevBrasHematolHemoter.2014;36(3):219–25.
2.LantisKL,JayneHarrisR,DavisG,RennerN,FinnWG. Eliminationofinstrument-drivenreflexmanualdifferential leukocytecounts.AmJClinPathol.2003;119(5):656–62.
3.GuarnerJ,AtuanMA,NixB,MishakC,VejjajivaC,CurtisC, etal.Processtoevaluatehematologicalparametersthatreflex tomanualdifferentialcellcountsinapediatricinstitution. ClinLab.2010;56(1–2):21–7.
4.BriggsC,LinssenJ,LongairI,MachinSJ.Improvedflagging ratesontheSysmexXE-5000comparedwiththeXE-2100 reducethenumberofmanualfilmreviewsandincrease laboratoryproductivity.AmJClinPathol.2011;136(2):309–16.
5.NCCLSDocumentH-20-A2.Referenceleukocyte(WBC) differentialcount(proportional)andevaluationof
instrumentalmethods;approvedstandard-secondedition. CLSI.2007;27(4).
6.RuzickaK1,VeitlM,Thalhammer-ScherrerR,SchwarzingerI. ThenewhematologyanalyzerSysmexXE-2100:performance
evaluationofanovelwhitebloodcelldifferentialtechnology. ArchPatholLabMed.2001;125(3):391–6.
7.StammingerG,AuchD,DiemH,SinhaP.Performanceofthe XE-2100leucocytedifferential.ClinLabHaem.
2002;24(5):271–80.
8.FordJ.Redbloodcellmorphology.IntJLabHematol. 2013;35(3):351–7.
HelenaZerlottiWolfGrotto∗
SysmexLatinAmérica&TheCaribbean,SãoPaulo,SP,Brazil
∗Correspondenceto:RuadoParaíso,148,Cj.31,04103-000São
Paulo,SP,Brazil.
E-mailaddress:[email protected]
Received28July2014
Accepted29September2014
Availableonline20November2014
http://dx.doi.org/10.1016/j.bjhh.2014.11.007
1516-8484/©2014Associac¸ãoBrasileiradeHematologia,
HemoterapiaeTerapiaCelular.PublishedbyElsevierEditora