Table Of Content

Viewpoint Quality standards and samples hypercholesterolaemia has long been known to refine or resolve genetic risks identifiedbyfamilyhistory.Theadventof in genetic testing genetic testing for thisdisorder infamilies for whom the causal mutation has been identified has led to improved diagnostic David Ravine,1,2,3 Graeme Suthers4,5 accuracy andsimplifiedthe process of risk assessmentamongrelatives.7 Genetictestingisincreasinglyrequested for healthy people identified by family history as being at high risk of having The most critical performance indicator for medical laboratories is the delivery of accurate test inherited a mutation predisposing to results. In any laboratory, there is always the possibility that random or systematic errors may aseriousadult-onsetautosomaldominant occur and place human health and welfare at risk. Laboratory quality assurance programmes monogenic disorder. Many of these continue to drive improvements in analytical accuracy. The most rigorously scrutinised data on conditions have no associated subclinical laboratory errors, which come from transfusion medicine, reveal that the incidence of analytical manifestations. Referred to as predictive, errorshasfallentolevelswheremostoftheresidualriskisnowfoundinpreanalyticallinksinthe presymptomatic or predisposition testing, chain from patient to result, particularly activities associated with ordering of tests and sample thiscategoryoftestingisdistinguishedby collection. This insight is important for genetic testing because, like pretransfusion testing of absence of corroborating clinical or other patients with unknown blood groups, a substantial proportion of genotyping results cannot be confirmatory data. This scenario is not immediately verified. An increasing number of clinical decisions, associated personal and social unique to late-onset dominant disorders choices, and legal outcomes are now influenced by genetic test results in the absence of other andisoften seeninotherareasofgenetic confirmatorydata.Anincorrecttestresultmayleadtounnecessaryandirreversibleinterventions, testing including prenatal, preimplanta- which may in themselves have associated risks for the patient, inaccurate risk assessment tion genetic diagnosis, and carrier testing regardingthedisease,missedopportunitiesfordiseasepreventionorevenwrongfulconvictionin for autosomal and X linked recessive a court of law. Unfortunately, there is limited information available about the risk of preanalytical disorders.Itisalsoaprominentfeatureof errors associated with, and few published guidelines regarding, sample collection for genetic ‘personalised medicine’,8 where geno- testing. The growing number and range of important decisions made on the basis of genetic typing is used to predict responses to, or findings warrant a reappraisal of current standards to minimise risks in genetic testing. sideeffects from,drugs. Corroboration of evidence is an impor- Errors are an established, unwelcome genetic test processing chain from test tant aspect of clinical care. For example, featureinallareasofhealthcare.1Erroneous selection by the clinician to clinical deci- a normal chest radiograph from a patient testresultsarewidespread2andarearecog- sions based on the result.3 Although the with severe congestive cardiac failure nised cause of iatrogenic harm. Incorrect iterative improvements in proficiency wouldnormallypromptachecktoensure genetic test results take the issue further. testing by molecular genetic testing labo- the correct image is being viewed. Many Besidesintroducingrisksofunnecessaryand ratorieshavebeenencouraging,45theyare clinicalmeasurementsarecorroboratedby irreversible interventions for those being alsoareminderthatgenetictestingisnot repeat measurements. The measurement tested, they may also lead to missed immune fromthe usual risks oferror. of blood pressure, for example, can iden- opportunitiesfordiseasepreventionamong tify patients at increased risk of stroke. relatives. With genetics emerging as a RISKS ASSOCIATED WITH GENETIC Although no other clinical features may substantialdisciplinewithinpathologyand TESTING be available to corroborate the result, laboratorymedicine,thereisaresponsibility Testing for medically significant genetic repeated measurements over timeprovide now to reassess current standards for variants is growing rapidly, although it is a background in which outlying or minimisingrisksoferrorinthefield. stillasmallcomponentofmedicaltesting incongruentfindings can beevaluated. As with all pathology investigations, overall. In the UK in 2005e2006, genetic Now, consider the case of testing a errors may occur at any point in the tests accounted for <0.1% of all labora- DNA sample from a young woman for toryinvestigations.6Comparedwiththeir aBRCA1mutation,whichwaspreviously working familiarity with most pathology identifiedinhermother,whohasastrong investigations, many clinicians still have familyhistoryofearlyonsetbreastcancer. littleexperienceoftestingforinheritedor Theinvestigationwillyieldaresultthatis ThispaperisfreelyavailableonlineundertheBMJ Journalsunlockedscheme,seehttp://jcp.bmj.com/site/ denovogenomic variants. qualitativelydifferentfromtheoutputsof about/unlocked.xhtml As well as the usual challenges associ- mostclinicalinvestigations.Thewoman’s ated with bringing new technologies into family history alone will have identified 1SchoolofPathologyandLaboratoryMedicine;2Western healthcare, there are further issues associ- herashavinga1in2riskofinheritingthe AustralianInstituteofMedicalResearch,Universityof ated with genetic investigations to mutation and, in contrast to many WesternAustralia,Perth,WesternAustralia,Australia; 3PathWest,QEIIMedicalCentre,Nedlands,Western consider.Geneticfindingsmayaidthetask medicaltestresults,theoutcomeisbinary. Australia,Australia;4DepartmentofPaediatrics, of assessing clinical symptoms. They may The mutation is either present, in which UniversityofAdelaide,Adelaide,SouthAustralia, also confirm a diagnosis long before the case she has a high lifetime risk of breast Australia;5RoyalCollegeofPathologistsofAustralasia, onset of symptoms in an individual and ovarian cancer, or absent, with her Sydney,NewSouthWales,Australia patient. This is especially the case for cancer risk returning to the background CorrespondencetoDrGraemeSuthers,SAClinical inheriteddisorderswithreadilyidentifiable population level for women of her age. GeneticsService,SAPathology,Women’s&Children’s Hospital,NorthAdelaide,SA5006,Australia; biomarkers.Forexample,serumcholesterol Unless verified by replicate testing, there [email protected] testing of relatives at risk of familial are usually no further opportunities to JClinPatholMay2012Vol65No5 389 Viewpoint blood in the tube’ (WBIT).16 17 They also Case study reveal other risks, including incomplete labelling in approximately 11 samples per A 40-year-old male patient was diagnosed with an extra-adrenal paraganglioma. Family 1000.1618Suchlabellinghasbeenshownto history revealed a fourth degree relative who presented with medullary thyroid cancer at result ina 40-foldincreased riskofsample 25yearsofage.Early-onsetmedullarythyroidcanceristhehallmarkofmultipleendocrine errors compared with the risk associated neoplasia type 2, which is an autosomal dominant disorder due to mutations in the RET withproperlylabelledsamples.19 gene.Somemutationcarriersremainhealthy,buttheirriskofdevelopingthyroidcanceris Therearenocomparabledataforgenetic high.RETmutationscanalsobeassociatedwithparagangliomas.Thepatientwastested testing. It is important to recognise, foraheritableRETmutationbyanaccreditedlaboratory,andawell-recognisedpathogenic however, that current reported pretrans- mutationwasidentified.Heproceededtohaveaprophylacticthyroidectomy,whichisthe fusion error rates are likely to be lower recommended management for mutation carriers. The patient’s unaffected children than those prevailing in the era before subsequentlyhadpredictivegenetictestingandnonewasidentifiedwiththemutation.The there was heightened emphasis on risk relative with thyroid cancer was eventually reviewed and testing demonstrated that minimisation strategies.20 It is possible she did not have our patient’s RET mutation. Instead, a different mutation was identified that the situation with genetic testing in the same gene. Repeat sample collection and analysis demonstrated that the RET may be more hazardous as there are mutation originally identified in the index patient was a laboratory artefact. The revised additional known risks associated with resulteliminatedtheearlierjustificationforhisprophylacticthyroidectomyandalsoforthe sample collection. These include labelling predictivegenetictestingofhischildren.Hewassubsequentlyshowntohaveaheritable of samples collected from interventional mutationintheSDHDgene.Althoughtheamendeddiagnosisalsohasfamilialimplications, procedures,21 such as amniotic fluid, his children are now disenchanted and have declined further involvement with the clinic. chorionic villus and fetal blood sampling, from patients in delivery rooms22 and synchronously fromrelatives.23 corroborate the finding. Furthermore, the orderingoftestsandsamplecollection,are Analytical error rates in pretransfusion test result and any associated clinical or beyondtheirscope. testing have now fallen to levels where psychosocial sequelae are very likely to There are very few published reports mostresidualmistakesoccurinpreanalyt- remain unchallenged over a lifetime, and about the incidence of errors in genetic ical phases, particularly those associated possibly beyondinto the nextgeneration. testing.Onereportdocumented1.7errors withbloodsamplecollectionandlabelling. In this setting, the riskof errorin genetic per 1000 investigations (excluding errors Thesametrendiswelldocumentedinother testing is not merely academic. Many in test selection and interpretation).14 generallaboratoryarenas.2425Itisnotable genetic testing services have their own Errors were equally divided between that a range of error-prone preanalytical anecdotes and, with permission from the preanalytical (sample collection and tasksareoftencompletedbypeoplebeyond patient involved, we include an example handling) and analytical phases. A useful the jurisdiction of laboratories, including fromourownexperience(seecasestudy). insightfromthisauditisthatthereported referring clinicians, clients and carers as The case demonstrates that genetic error rate is in the same range as those wellasworkerscontributingtodataentry, testing errors can have wide ranging and reported in other types of medical specimencollectionandtransport. lastingsequelae.Forexample,thefactour testing.15 In fact, the occult nature of an patient did not develop thyroid cancer erroneouspredictivegenetictestresult,as RISKS OF PREANALYTICAL ERRORS IN could well have been erroneously attrib- illustratedinthecasestudy,demonstrates GENETIC TESTING uted in the future to the ‘success’ of his thatestimatesoftheprevalenceofadverse Acknowledgementofrisksofpreanalytical thyroidectomy, when the real reason was consequences associated with predictive errors, particularly sample mislabelling, his risk was negligible in the first place. testing are best derived from independent raises the question whether the safety of Although the consequences of a genetic corroborating data. However, these data patients undergoing genetic testing is testingerrorcanhavenodiscernibleeffect, can only come from resampling and adequatelyprotectedbythecurrentscope particularly if the error remains unde- retestingand,asfarasweareaware,they ofgenetictestingQAprogrammes. tected, a range of outcomes is possible, arescanty andanecdotal. The prospect of a mislabelled sample including occasional severe harm or destined for genetic testing highlights an avoidable death. INSIGHTS FROM OTHER AREAS OF additional important risk associated with LABORATORY MEDICINE an adverse outcome. Medical testing is SOURCES AND FREQUENCY OF GENETIC Incontrasttothelikelihoodthatapropor- generally prompted by a patient’s history TESTING ERRORS tion of genetic testing errors will remain orsymptomsindicatinganincreasedprior The error we reported was shown to have undetected,errorsassociatedwithallogenic probability of an underlying disease occurred during the analytical phase of bloodtransfusionortissuetransplantation process. In the case of predictive genetic testing.However,itcouldjustaseasilyhave are rapidly unmasked. Life-threatening testing, the investigation is prompted by occurred before the sample arrived in the immune reactions to mismatched red cell a family history indicating a high risk of laboratoryorpostanalysis.Analyticalerrors and tissue immunotypes have focused predisposition to future illness from an have been the prime focus of molecular attention on the incidence and sources of inheritedgenemutation.Thepriorriskof geneticqualityassurance(QA)programmes errors, and also on strengthening risk abnormalityinamislabelledsampledefines developed by professional bodies in the minimisationstrategies. thelikelihoodthataresultwilldifferfrom USA,EuropeandAustralasia.9e13Whilethe Studies of pretransfusion testing report thepatient’struestatus.Forexample,if1in success of these programmes is well that0.5e0.8sampletubesper1000contain 20 samples from a patient cohort have documented, most preanalytical risks, blood from someone other than the iden- abnormal test results, then analyses of particularly those associated with the tified patient, often referred to as ‘wrong swappedsampleswillyieldcorrectresults 390 JClinPatholMay2012Vol65No5 Viewpoint for both patients more than 90% of the identification systems. A multicentre VERIFYING GENETIC TEST RESULTS time.Alternatively,aninadvertentswitch international survey of 650000 samples Genetic testing is increasingly used to of samples from individuals in whom the reportedin2003that6.1samplesper1000 resolve clinical uncertainty in a diverse priorriskofabnormalityis1in2,whichis weremislabelled16andaWBITrateof0.5 arrayofmedicalscenarios.Manyscenarios the case among siblings at risk of having samples per 1000, despite almost all offer opportunities to corroborate genetic inherited a mutation from an affected participating laboratories having written test findings. For example, a test report parent, will yield correct results for both policies on sample collection and that a child has mutations in both CFTR individualsonly50%ofthetime. labelling.27 The Division of Transfusion genes may be verified by a pre-existing In the context of a family history of Medicine at the University of California, clinical diagnosis of cystic fibrosis (an disease, the elevated prior probability of LosAngeles,asaresultofa17-yearpolicy autosomal recessive disorder caused by having inherited the disease-causing of requiring two independent samples for CFTR mutations), by finding an elevated mutation also raises the risk of harm blood typing before issuing type-specific sweat chloride concentration (a specific arising from sampling errors. This risk is blood for non-group-O patients, reported biochemicalfeatureofcysticfibrosis)orby compounded by the earlier mentioned a very similar mislabelling rate of 0.44 demonstratingthattheparentsarecarriers evidence of increased risk of sample samples per 1000.22 In keeping with the of the child’s mutations. As another mislabelling associated with synchronous policyoftheUniversityofCalifornia,Los example,anarraygenomicscanningresult sample collection fromrelatives.23 Angeles, the World Marrow Donor Asso- in a sample from a 3-year-old child with Although some indication of the likely ciation also established guidelines recom- autism might unexpectedly demonstrate background error rate associated with mending confirmatory tissue typing on a genomic deletion involving the TP53 predictivegenetictestingwouldbeuseful, duplicate samples from potential donors gene.Thefindingmaypromptadiscovery anyestimatewillinevitablybetheaverage and recipients before transplantation.28 of a previously unrecognised family ofnumerousrandomandsystematicerror The policy of confirmatory testing of a history indicative of Li Fraumini rates, all of which will also be inevitably second independent blood sample for syndrome.However, asthecasestudywe subject to regional variation and changes patients without a historical record presentillustrates,thefamilyhistorymay over time. Notwithstanding these limita- corroborating an ABO typing result from be misleading, inconclusive or absent, in tions, it is apparent that the current ad asinglesamplehassincebeenadoptedby whichcaseduplicatesamplingwouldavert hoc approach to minimising preanalytical more transfusion services and now the risk of arranging inappropriate predic- risks associated with genetic testing, extends to approximately 20% of US tivegenetictestsforrelatives.Thenumber particularly the risk of sampling errors, is laboratoryand hospital practices.18 of confirmatory options already available a patient safety issue that requires demonstratetheusefulnessofaninformed research attention. Meanwhile, it is CURRENT SAMPLE VERIFICATION and practical approach to risk mini- importantthatthoseinvolvedinthisarea PROTOCOLS IN GENETIC TESTING misationingenetic testing, particularlyin of medicine consider all potential weak- Although there has been a long-standing theabsenceofpeerrevieweddataandQA nesses in the total testing process, and focus on reducing the risk of sampling programmes that focus on the quality of then avail themselves of the full range of errors associated with transfusion and alllinksinthechainfrompatienttoresult. available riskminimisation strategies. transplantation, most contemporary reports29 30 and documents on genetic RISK MINIMISATION STRATEGIES FOR GRADIENT OF RISK MANAGEMENT testing developed by regulatory and GENETIC TESTING STRATEGIES professional bodies31e38 have primarily There is at the very least an ethical Experience from transfusion and trans- addressed strengthening laboratory QA. imperative that any genetic test finding plantation medicine provides guidance on With some notable exceptions,39 40 few withthepotentialtoinfluencesignificant risk minimisation strategies for genetic have argued in favour of adopting a more clinical,personalorsocialdecisionsshould testing. These services have implemented patient-centric approach to risk mini- prompt consideration of all associated measuresovermanyyearstodecreasethe misation, including addressing risks asso- risks, including the risk of an erroneous incidenceofsamplingerrors.Importantly, ciatedwithsamplecollection.Asurveyof result. A recent report to the Victorian both these services reject the baseline directorsof17medicalgeneticlaboratories Department of Justice43 offers an illumi- standard of analysing single specimens in Australia and New Zealand in 2005,41 nating insight into contemporary societal thatfulfilminimumlabellingcriteria,such before sampling risk minimisation guide- views on the strength of DNA evidence. as concordance of two person-specific lines were implemented, revealed a range An inquiry into the wrongful conviction identifiers on the sample tube and of approaches. In lieu of duplicate of a man for rape in July 2008 identified accompanying requestform.26 sampling, some laboratories split samples that the conviction had hinged on DNA Toensureagainsttransfusingthewrong for predictive testing into two tubes evidence derived from a single sample. At blood into a patient, transfusion services immediatelyonarrival,andthenarranged the time of the man’s conviction, it had usearangeofstrategieswhich,ingeneral independent replicate analyses. However, notbeenappreciated thatthesample had terms, have focused on strengthening the mostlaboratoriesacceptedsingleunsigned beencollectedinthesameunitasforensic chain of custody procedures in lieu of samplesforpredictivegenetictestingafter samples secured by the same doctor 30h analysisofasecondindependentlydrawn ensuring concordance of patient identi- earlierfromanotherwoman,withwhom, sample. Measures include requiring fying details on request forms and itwasundisputed,themanhadhadprior asecondpersontoconfirmconcordanceof samples. Subsequently, an Australian sexual contact. A former judge who patientidentity,requestformandsample; laboratory accreditation guideline, which conducted the reviewwarned: avoiding prelabelled tubes; appointing is not obligatory, was issued in 2006 safety officers to oversee training and recommending that predictive genetic .theDNAevidencehadbeenperceivedas processes; and using electronic patient testsbeconductedonduplicatesamples.42 beingsopowerfulbyallinvolvedinthe JClinPatholMay2012Vol65No5 391 Viewpoint casethatnoneofthefiltersuponwhich serious late-onset dominant disorders, nosisandtherapeuticoptions,includingin oursystemofcriminaljusticedependsto including familial cancers, late-onset relationtodiseasesforwhichgenetictests minimisetheriskofamiscarriageof neurodegenerative disorders and heredi- are available. Similarly, the goal of justice,operatedeffectivelyatanystage tary arrhythmia syndromes, some profes- amedicallaboratoryistoprovidetheright untilamatterofweeksbefore[theman’s] sional organisations47 48 have issued result for the right patient in a timely appealwasexpectedtobeheard. guidelinesrecommendingreplicatetesting. fashion every time. AlexanderPope wrote Thefactaninnocentmanspenttimein When considering the option of replicate in An Essay on Criticism that ‘To err is jailasaconvictedrapistonthebasisofan testing, some practitioners may contem- human.’. Three hundred years later, his uncorroboratedgeneticidentitytestresult plate arranging the test in a second labo- message is still potent. All arenas of is a sobering reminder of the extent of ratory to provide additional protection human endeavour are at risk of human potential ‘fall out’ when things go wrong against unrecognised laboratory-specific error, and the emerging discipline of in genetic testing. Clearly, such errors are risksofsystematicerror.Thisopportunity, genetictesting isnotimmune.Errorswill notsolelyaconcernformedicalproviders. however, would not be available for occur here, as they do in other areas of Sensibly, the importance of corroborating genetic tests that are currently monopo- laboratory testing, and medicine in a genetic test result will vary as will the lisedthroughpatentsorlicensing. general.Itisoflittle comfortthatsample bestwaysofachievingverification.Ideally, With ongoing genetic advances, the errors, such as WBIT, are likely to be considerationshouldbemadeofthelikely issue of risk minimisation in relation to more common than reports of adverse riskoferror;theseriousnessandfrequency genetic testing is now extending into incidents. ofanypossibleadverseconsequences;and mainstream medical practice. Pharmaco- Like the proverbial elephant in the the range, effectiveness and total cost of genomic testing has a growing role in room,weknowtheerrorsarepresentbut available risk minimisation strategies. guiding therapeutic decisions, including wehesitatetotalkaboutthem.Theissue Examinationoflegalandfinancialrisksfor managementofhepatitisCvirusinfection, must be addressed, however, because those who are professionally accountable which is the most common bloodborne errorsingenetictestinghavethepotential may also need to be included in the cost infection in the USA.49 Until now, there to prompt clinical decisions with a high analysis. hasbeenlittlecommentaryaboutwhether risk of attendant harm. They may also There is a strong argument now that thereisaneedtostrengthenthequalityof direct important life choices for those healthcare practitioners involved in preanalytical and postanalytical phases beingtested, with ramifications thatmay requesting or performing genetic tests associatedwithpharmacogenomictesting, influencehumanhealthandwelfareatall which have potentially fatal outcomes particularly in the boundary areas linking developmental stages. Some errors will associated with sampling errors should clinicalandlaboratorymedicine. invariably lead to outcomes over which recognise that the baseline standard of While risk minimisation strategies the person being tested will have no analysing single specimens fulfilling should be tailored to individual clinical control, such as wrongful conviction in minimum labelling criteria represents scenarios, cost-effectiveness is an impor- a court of law. Errors in genetic testing inadequate risk minimisation for the tant consideration. For example, if we may also waste the increasingly scarce patient.44 45 The full scope of potentially accept that a single sample is appropriate health dollars, and place individual fatal outcomes must also be considered, forprenataltestingoffetaltissue,thereare healthcare practitioners at professional, including any risk of future fatalities fewpublishedrecommendationsregarding legal and financial risk. Itis nowtime for linked to missed opportunities for disease the minimum labelling and chain of theprofessiontoconsiderthefullrangeof preventionamongrelatives.Further,given custodyrequirementsotherthaninrelation errors that are possible along the genetic the range of deleterious non-fatal to paternity testing. What impact would test processing chain from patient to outcomes associated with genetic testing, strengthened risk minimisation measures there is a case for extending this recom- have on the cost of delivering prenatal mendation to incorporate significant clin- testing? And would the extra cost be Interactive multiple choice questions ical,personal,socialorlegaldecisionsthat acceptabletothosepayingfortheservice? may bedirected by agenotype finding. In the absence of widely accepted This JCP review article has an accom- There are, however, several additional professionalguidelines,QAprogrammesor panying set of multiple choice questions measures to consider in relation to peer reviewed data for genetic test (MCQs).Toaccessthequestions,clickon genetictesting.Forexample,theoptionof sampling, the decision to implement BMJ Learning: Take this module on BMJ routine replicate testing on independent a more stringent sample protocol is ulti- Learning from the content box at the top samples has long been regarded as inap- mately directed locally based on what is right and bottom leftof the onlinearticle. propriate for prenatal genetic diagnosis, known about the clinical risks involved. Formoreinformationpleasegoto:http:// particularly if the invasive procedure Venesectionists, laboratory scientists and jcp.bmj.com/education. Please note: The required to obtain a second sample has pathologists are not necessarily aware of MCQs are hosted on BMJ Learningdthe attendant potentially serious risks, either these details. The ultimate responsibility best available learning website for forthemotherorforthefetus.Fortunately, forindividualpatientsafetyisvestedinthe medical professionals from the BMJ inthisarena,simultaneoustestingofnon- clinician requesting the investigation. On Group.Ifprompted,subscribersmustsign disease associated genetic polymorphisms the other hand, laboratory-based practi- into JCP with their journal’s username in fetal and parental DNA samples is an tioners are well placed toremind referring and password. All users must also increasingly available option to detect cliniciansofsamplingrecommendations. complete a one-time registration on BMJ sample switching, and also to assess for Learning and subsequently log in (with possiblematernalcellcontamination.46 CONCLUSION aBMJLearningusernameandpassword) It is worth noting that in other arenas, The goal of a clinician is to provide the on every visit. such as predictive genetic testing for patient with an accurate diagnosis, prog- 392 JClinPatholMay2012Vol65No5 Viewpoint result, and devise appropriate risk mini- 11. HudsonKL,MurphyJA,KaufmanDJ,etal. ClinicalandlaboratorystandardsInstitute(CLSI), misation strategies. Until such data are OversightofUSgenetictestinglaboratories.Nat 2006.CLSIdocumentMM1eA2. Biotechnol2006;24:1083e90. 33. AmericanCollegeofMedicalGenetics.Standards available, individual healthcare practi- 12. SenecaS,MorrisMA,PattonS,etal.Experience andGuidelinesforClinicalGeneticsLaboratories.2006. tioners involved in genetic testing should andoutcomeof3yearsofaEuropeanEQAscheme http://www.acmg.net/Pages/ACMG_Activities/stds- consider the associated possible risks to forgenetictestingofthespinocerebellarataxias.Eur 2002/g.htm(accessed2Aug2011). patient health and welfare, and look JHumGenet2008;16:913e20. 34. OrganizationforEconomicCooperationand 13. HertzbergM,NevilleS,McDonaldD.Externalquality Development(OECD).OECDGuidelinesforQuality beyond the baseline standard of testing assuranceofmolecularanalysisofhaemochromatosis AssuranceinMolecularGeneticTesting.ParisFrance, a single sample. genemutations.JClinPathol2006;59:744e7. 2007.http://www.oecd.org/dataoecd/43/6/ 14. HofgartnerWT,TaitJF.Frequencyofproblems 38839788.pdf(accessed2Aug2011). AcknowledgementsWearegratefultoourmany duringclinicalmolecular-genetictesting.AmJClin 35. JavaherP,KaariainenH,KristofferssonU,etal. colleagueswhocontributedtoasuccessionofvigorous Pathol1999;112:14e21. EuroGentest:DNA-basedtestingforheritable discussionsonthistopic.Weespeciallywishto 15. BoniniP,PlebaniM,CeriottiF,etal.Errorsin disordersinEurope.CommunityGenet acknowledgethosewho‘wenttheextramile’and laboratorymedicine.ClinChem2002;48:691e8. 2008;11:75e120. providedrelevantdata,referencesandopinions,aswell 16. DzikWH,MurphyMF,AndreuG,etal.An 36. ChenB,GagnonM,ShahangianS,etal.Good asrevieweddraftsofthedocument:JacquelineCarroll, internationalstudyoftheperformanceofsample laboratorypracticesformoleculargenetictestingfor KenDavis,RobElles,AndrewFellowes,JeniHood,the collectionfrompatients.VoxSanguis2003;85:40e7. heritablediseasesandconditions.MMWRRecomm lateKarenSnow-Bailey,NicoleStaplesandMichael 17. MurphyMF,StearnBE,DzikWH.Current Rep2009;58:1e37. Watson.Anyshortcomingsinthepresentationofour performanceofpatientsamplecollectionintheUK. 37. DequekerE,StuhrmannM,MorrisMA,etal.Best argumentsareourresponsibilityalone. TransfusMed2004;14:113e21. practiceguidelinesformoleculargeneticdiagnosisof FundingTheRCPApositionstatementcitedinthe 18. GrimmE,FriedbergRC,WilkinsonDS,etal.Blood cysticfibrosisandCFTR-relateddisordersdupdated article,“SampleRequirementsforMedicalGenetic banksafetypractices:mislabeledsamplesandwrong Europeanrecommendations.EurJHumGenet Testing”,waspreparedwithsupportofagrantfromthe bloodintubedaQ-Probesanalysisof122clinical 2009;17:51e65. AustralianGovernment’sDepartmentofHealth&Ageing. laboratories.ArchPatholLabMed2010;134:1108e15. 38. MattocksCJ,MorrisMA,MatthijsG,etal.A TheDepartmenthadnoroleindeterminingthecontent, 19. LumadueJA,BoydJS,NessPM.Adherenceto standardizedframeworkforthevalidationand conclusions,ordistributionofthedocument.The strictspecimen-labelingpolicydecreasesthe verificationofclinicalmoleculargenetictests.EurJ Departmenthadnoroleinthepreparationofthisarticle. incidenceoferroneousbloodgroupingofbloodbank HumGenet2010;18:1276e88. specimens.Transfusion1997;37:1169e72. 39. HarperJC,SenguptaS,VeselaK,etal. CompetinginterestsNone. 20. HowanitzPJ,WalkerK,BachnerP.Quantificationof AccreditationofthePGDlaboratory.HumReprod errorsinlaboratoryreports.Aqualityimprovement 2010;25:1051e65. ContributorsBothauthorsconceivedandwrotethe studyoftheCollegeofAmericanPathologists’Q-Probes 40. PlebaniM.Exploringtheicebergoferrorsin article. program.ArchPatholLabMed1992;116:694e700. laboratorymedicine.ClinChimActa ProvenanceandpeerreviewNotcommissioned; 21. NakhlehRE,ZarboRJ.Surgicalpathologyspecimen 2009;404:16e23. externallypeerreviewed. identificationandaccessioning:aCollegeofAmerican 41. RoyalCollegeofPathologistsofAustralasia. PathologistsQ-ProbesStudyof1,004,115casesfrom SuthersG,YuSinReportoftheAustralianGenetic Accepted16November2011 417institutions.ArchPatholLabMed TestingSurvey.2006.http://www.rcpa.edu.au/static/ PublishedOnlineFirst18January2012 1996;120:227e33. File/Assetlibrary/publicdocuments/MediaReleases/ JClinPathol2012;65:389e393. 22. FigueroaPI,ZimanA,WheelerC,etal.Nearlytwo AustralianGeneSurvey2006.pdf(accessed2Aug doi:10.1136/jclinpath-2011-200519 decadesusingthecheck-typetopreventABO 2011). incompatibletransfusions:oneinstitution’s 42. NationalPathologyAccreditationAdvisory experience.AmJClinPathol2006;126:422e6. Council,AustralianGovernmentDepartmentof 23. HoutzT,ChiafariFA,WenkRE.Controllingspecimen healthandAging.LaboratoryAccreditation REFERENCES misidentificationinparentageanalysis.Transfusion StandardsandGuidelinesforNucleicAcidDetection 1. BrennanTA,LeapeLL,LairdNM,etal.Incidenceof 2004;44:1258e9. andAnalysis.2006.http://www.health.gov.au/ adverseeventsandnegligenceinhospitalized 24. AstionML,ShojaniaKG,HamillTR,etal.Classifying internet/main/publishing.nsf/Content/ patients:resultsoftheHarvardMedicalPractice laboratoryincidentreportstoidentifyproblemsthat 11E45788D881EE37CA25732B0048D60C/$File/ StudyI.NEnglJMed1991;324:370e6. jeopardizepatientsafety.AmJClinPathol NucAcidDetect.pdf(accessed2Aug2011). 2. HowanitzPJ.Errorsinlaboratorymedicine.Arch 2003;120:18e26. 43. VincentFHR.Report:InquiryintotheCircumstances PatholLabMed2005;129:1252e61. 25. CarraroP,PlebaniM.Errorsinastatlaboratory: ThatLedtotheConvictionofMrFarahAbdulkadir 3. GiardielloFM,BrensingerJD,PetersenGM,etal. typesandfrequencies10yearslater.ClinChem JAMA.Melbourne,Australia:VictorianGovernment TheuseandinterpretationofcommercialAPCgene 2007;53:1338e42. Printer,2010.PPNo301:Session2006e10. testingforfamilialadenomatouspolyposis.NEnglJ 26. TheJointCommission.AccreditationProgram: 44. MurphyMF,KayJD.Patientidentification:problems Med1997;336:823e7. LaboratoryNationalPatientSafetyGoals.2011. andpotentialsolutions.VoxSang2004;87(Suppl 4. McGovernMM,BenachMO,WallensteinS,etal. http://www.jointcommission.org/assets/1/6/ 2):197e202. Qualityassuranceinmoleculargenetictesting 2011_NPSGs_LAB.pdf(accessed2Aug2011). 45. StainsbyD,RussellJ,CohenH,etal.Reducing laboratories.JAMA1999;281:835e40. 27. DzikWH.Newtechnologyfortransfusionsafety. adverseeventsinbloodtransfusion.BrJ 5. DequekerE,CassimanJJ.Genetictestingand BrJHaematol2007;136:181e90. Haematology2005;131:8e12. qualitycontrolindiagnosticlaboratories.NatGenet 28. HurleyCK,WadeJA,OudshoornM,etal.Aspecial 46. MaddalenaA,BaleS,DasS,etal;theACMG 2000;25:259e60. report:histocompatabilitytestingguidelinesfor LaboratoryQualityAssuranceCommittee. 6. UKDepartmentofHealth.ReportoftheReviewof hematopoieticstemcelltransplantationusing Technicalstandardsandguidelines:moleculargenetic NHSPathologyServicesinEngland.2006.http:// volunteerdonors.HumImmunol1999;60:347e60. testingforultra-raredisorders.GenetMed www.dh.gov.uk/prod_consum_dh/groups/ 29. IbarretaD,EllesR,CassimanJJ,etal.Towards 2005;7:571e84. dh_digitalassets/@dh/@en/documents/digitalasset/ qualityassuranceandharmonizationofgenetic 47. BelgianSocietyofHumanGenetics.Guidelinesfor dh_4137607.pdf(accessed22Aug2011). testingservicesintheEuropeanUnion.Nat PredictiveGeneticTestingforLateOnsetDisorders. 7. NedRM,SijbrandsEJ.Cascadescreeningforfamilial Biotechnol2004;22:1230e5. 2003.http://www.beshg.be/workgroups/ hypercholesterolemia(FH).PLoSCurr2011;3: 30. McGovernMM,EllesR,BerettaI,etal.Reportofan guidelinespredictivetesting.htm(accessed2Aug2011). RRN1238. internationalsurveyofmoleculargenetictesting 48. RoyalCollegeofPathologistsofAustralasia 8. SamaniNJ,TomaszewskiM,SchunkertH.The laboratories.CommunityGenet2007;10:123e31. PositionStatement.Samplerequirementsfor personalgenomedthefutureofpersonalised 31. RainenL,ArbiqueJC,AsthanaD,etal.Collection, medicalgenetictesting:dogenetictestsdemand medicine?Lancet2010;375:1497e8. Transport,Preparation,andStorageofSpecimensfor adifferentstandard?2007.http://www.rcpa.edu.au/ 9. DequekerE,RamsdenS,GrodyWW,etal.Quality MolecularMethods;ProposedGuideline.Wayne, static/File/Assetlibrary/publicdocuments/Policy controlinmoleculargenetictesting.NatRevGenet PennsylvaniaUSA:Clinicalandlaboratorystandards Manual/PositionStatements/Samplerequirementsfor 2001;2:717e23. Institute(CLSI),2005.CLSIDocumentNo.MM13-P. medicalgenetictesting.pdf(accessed2Aug2011). 10. PalomakiGE,BradleyLA,RichardsCS,etal. 32. RichardsS,GrodyWW,DequekerE,etal.Molecular 49. ThomasDL,ThioCL,MartinMP,etal.Genetic Analyticvalidityofcysticfibrosistesting:apreliminary DiagnosticMethodsforGeneticDiseases;Approved variationinIL28Bandspontaneousclearanceof estimate.GenetMed2003;5:15e20. Guideline.2ndedn.Wayne,PennsylvaniaUSA: hepatitisCvirus.Nature2009;461:798e801. JClinPatholMay2012Vol65No5 393