WIXLIB

6σQMS(1a) Regulatory &Accred .Requirements(2a) Traceability &Harmonization(4b) Pre- and Post-analyticRequirementsAct(12) Improve Quality,TQC Plan [CQI, CAPA](1) Define Goals forIntended Use (TEa, Dint)(1b) Clinical andMedical Applications(2) Select AnalyticMeasurement Procedure(2b) Manufacturer’sRef Methods & MaterialsPlan(3) Validate MethodPerformance (CV,bias)(3a) Manufacturer’sClaims(4) Implement Methodand Analytic System(4a) Manufacturer’sInstallation/Training(5) Formulate“Total QC Strategy”(5a) Sigma[(TEa-Bias)/CV](6) Select/DesignSQC (rules, N)(6a) SigmaQC Selection ToolDoCheck(11) Monitor FailuresFRACAS. Quality Indicators(7) DevelopTotal QC Plan(7a) Risk Analysis(10) Measure Quality &Performance (EQA, PT, MU)(8) ImplementAnalytic TQC Plan(8a) QC Tools(9) Verify Attainment ofIntended Quality22

Simplifying ISO 15189Define Policies, Processes, Procedures The laboratory should define policies,processes, and procedures in a QualityManual to document the implementation ofits Quality Management System– Policies: Describes what is to be achieved– Processes: Describes activities to implement apolicy– Procedures: Provides step by step directions foractivities23

ISO 15189Intended Use, Intended Quality 5.5.1.1 Selection, Validation– The laboratory shall select examination procedures whichhave been validated for their intended use.– The specified requirements (performance specifications)for each examination procedure shall relate to theintended use of that examination. 5.6.2.1 Quality Control– The laboratory shall design quality control procedures thatverify the attainment of the intended quality of results.24

Example ApplicationQuality Required for Intended Use Policy - Quality requirement shall be defined for each test toidentify critical performance characteristics that must beachieved to satisfy the intended clinical use of each laboratorytest. Process – Analytical goals shall be established for each testbased on an assessment of specific clinical diagnostic andtreatment requirements for the test, goals based on biologicvariability, criteria for performance in PT and EQA programs,and “state of the art” performance Procedure – Select specific analytical goal, or qualityrequirement, by comparison of different quality requirementsusing an error grid.25

Example – Quality Goals for HbA1cComparison of CAP, NGSP, ADAHbA1c Error Grid10.00NGSP HiNew Method Value (%Hb)9.00CAP PT-Hi8.00Y X7.00CAP PT-LoNGSP 0Reference Method Value (%Hb)CAP NGSP 6%, as demanding as ADA clinicaldiagnostic and treatment criteria26

ISO 15189Validation of Examination Procedures Section 5.5.– The laboratory shall select examination procedures whichhave been validated for their intended use – The specified requirements (performance specifications)for each examination shall relate to the intended use ofthat examination– The independent verification by the laboratory shallconfirm through obtaining objective evidence that theperformance characteristics for the examination procedureare met.27

Validation of Examination Procedures Policy - The laboratory shall validate the performancecharacteristics to ensure the quality required for intended useis achieved. Process - The laboratory shall employ an evaluation studythat includes experiments for analytic range, replication,comparison of methods, confirmation of reference ranges,and when appropriate, detection limit, interference, andrecovery. Experimental results should be analyzed graphicallyand statistically and the sigma-metric determined to judge theacceptability for the intended use.28

Validation of Examination Procedures Procedure - Preparation of a Method Decision Chart todetermine quality on the sigma scale and assess theacceptability of precision and accuracy for intended use.1.2.3.4.5.6.7.Define the quality requirement for the test in the TEa format.Scale the y-axis from 0 to TEa and the x-axis from 0 to TEa/2.Draw “sigma lines” for 2, 3, 4, 5, and 6σ, e.g., 2σ line has y-intercept ofTEa and x-intercept of TEa/2; 3σ y-int of Tea, x-int of TEa/3; 4σ y-int TEa, x-int TEa/4; 5σ y-int TEa, x-int TEa/5; 6σ Y-int TEa, x-int TEa/6.Obtain %CV from the replication experiment.Obtain %Bias from the comparison of methods experiment.Plot an “operating point” that has %Bias as the y-coordinate and %CV asthe x-coordinate.Assess sigma-quality from location of the operating point.29

Simplifying Method ValidationUse a Method Decision ChartMethod Decision Chart TEa 6.0%Accuracy - observed %Bias652-sigma3 sigma4 sigma5 sigma6 sigmaRegion 511.522.53Precision - observed %CV30

Goal is Six Sigma Quality!- Tolerance TargetSpecification-6 SDsshould fitinto spec ToleranceSpecification 6 SDsshould fitinto spec-6s -5s -4s -3s -2s -1s 0s 1s 2s 3s 4s 5s 6s31

Minimum Acceptable Quality is 3σ!- Tolerance TargetSpecification-3s-2s-1s0s ToleranceSpecification1s2s3s32

Simplifying Method ValidationCalculate a oleranceLimitBias TEaSigma TEa-BiasSDSD-6s -5s -4s -3s -2s -1s 0s 1s 2s 3s 4s 5s 6s33

Simplifying Method ValidationCalculate a Sigma-metric Sigma (TEa – Bias)/SD– TEa, allowable Total Error– Bias, inaccuracy of the method Initially from comparison of methods exp. Ongoing from PT/EQA survey or Peer Comparisonprogram– SD or CV, imprecision of the method Initially from replication experiment On-going from routine SQC data– All terms in either % or concentration units34

Example Calculations of Sigma Sigma (TEa – Bias)/CV Glucose, TEa 10%– Bias 1.0, CV 1.5%– Sigma (10.0-1.0)/1.5 6.0 HbA1c, TEa 6.0%– Bias 1.0%, CV 1.5%– Sigma (6.0-1.0%)/1.5% 3.3335

ISO 15189Quality Control 5.6.2.1– The laboratory shall design quality controlprocedures that verify the attainment of theintended quality of results. Must have defined “intended quality” Must have QC selection/design process that relates thequality required with the precision and accuracy of theexamination procedure and the rejection characteristicsof different control rules and different numbers ofcontrol measurements36

Simplifying QC DesignUse quality-planning tools Select SQC procedures on basis of observedsigma quality of examination procedures– Sigma QC Selection tool Described in CLSI C24A3– Charts of Operating Specifications Available from Westgard QC publications, website– “Westgard Sigma Rules” graphic tool Available from Thermo Fisher Scientific– http://thermoscientific.com/westgard-poster37

Westgard Sigma RulesTM2 Levels of ControlsDataQCReport ResultsNo13sNo22sYesR4sYesN 2R 1NoYesN 2R 1No8X41sYesN 4R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action6σ5σ4σ3σSigma Scale (%TEa-%Bias)/%CV38

Westgard Sigma RulesTM2 Levels of ControlsDataQCReport ResultsNo13sNo22sYesR4sYesN 2R 1NoYesN 2R 1No8X41sYesN 4R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action6σ5σ4σ3σSigma Scale (%TEa-%Bias)/%CV39

HbA1c: TEa 6.0%, Bias 1.0%,CV 1.0%, Sigma (6-1)/1 5DataQCReport ResultsNo13sNo22sYesR4sYesN 2R 1NoYesN 2R 1No8X41sYesN 4R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action6σ5σ4σ3σSigma Scale (%TEa-%Bias)/%CV40

HbA1c: TEa 6.0%, Bias 1.0%,CV 1.0%, Sigma (6-1)/1 5DataQCReport ResultsNo13sNo22sYesYesN 4R 1No8XYesLocate SigmaN 2R 1No41sR4sYesN 2R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action6σ5σ4σ3σSigma Scale (%TEa-%Bias)/%CV41

HbA1c: TEa 6.0%, Bias 1.0%,CV 1.0%, Sigma (6-1)/1 5DataQCReport ResultsNo13sNo22sYesYesN 4R 1No8XYesLocate SigmaN 2R 1No41sR4sYesN 2R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action6σ5σ4σ3σSigma Scale (%TEa-%Bias)/%CV42

Glucose: TEa 10.0%, Bias 1.0%,CV 1.5%, Sigma (10-1)/1.5 6DataQCReport ResultsNo13sNo22sYesR4sYesN 2R 1NoYesN 2R 1No8X41sYesN 4R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action6σ5σ4σ3σSigma Scale (%TEa-%Bias)/%CV43

Glucose: TEa 10.0%, Bias 1.0%,CV 1.5%, Sigma (10-1)/1.5 6DataQCReport ResultsNo13s22sYesNoR4sYesLocate SigmaN 2R 16σNoYesN 2R 1No8X41sYesN 4R 1NoYesN 2R 2N 4R 2N 2R 4Take Corrective Action5σ4σ3σSigma Scale (%TEa-%Bias)/%CV44

Why a low sigma is bad!Sigma-metric 3.0DataQCReport ResultsNo13sNo22sYesR4sYesYesN 2R 1No8X41sYesN 4R 1NoYesN 2R 2N 4 N 2R 2 R 4Take Corrective Action6σ5σ4σ3σLocate SigmaN 2R 1NoSigma Scale (%TEa-%Bias)/%CV45

ISO 15189Measurement Uncertainty (MU) 5.5.1.4– The laboratory shall define the performancerequirements for the measurement uncertainty ofeach measurement procedure and regularlyreview estimates of measurement uncertainty.– The laboratory shall consider measurementuncertainty when interpreting measured quantityvalues.– Upon request, the laboratory shall make itsestimates available to laboratory users.46

Simplifying MUUse “Intermediate” QC data 5.5.1.4 Note 2– Measurement uncertainties may be calculatedusing quantity values obtained by themeasurement of quality control materials underintermediate precision conditions that include asmany routine changes as reasonably possible inthe standard operation of a measurementprocedure, e.g., changes of reagent and calibratorbatches, different operators, scheduled instrumentmaintenance.47

Intermediate Precision Conditions Should be long enough time period to includechanges of calibration, calibrator lots, reagentlots, different operators, instrumentmaintenance Recommend at least 100 data points– Confidence interval for SD is about 10%– Approximately 3 to 4 months data would be theminimum time period48

Simplifying Measurement Uncertainty Policy - The laboratory shall evaluate theuncertainty of measurement for eachexamination procedure where precision dataare available from stable control materials. Process - The laboratory shall determinestandard measurement uncertainty from SQCdata obtained under intermediate precisionconditions.49

ISO 15189Review of Estimates of MU 5.5.1.4 The laboratory shall define the performancerequirements for the measurement uncertainty ofeach measurement procedure and regularly reviewestimates of measurement uncertainty. Process - The laboratory shall compare estimates ofstandard measurement uncertainty to theperformance required for intended use of theexamination procedure by determination of sigmametrics.50

Simplifying MUCalculate Sigma-metric Procedure - Define TEa, estimate SD fromintermediate QC data Sigma (%TEa)/%CV– Where %TEa is PT or EQA requirement foracceptable performance– %CV is determined from laboratory’s intermediateterm SQC data51

ISO 15189External Quality Assessment 5.6.3.1 Interlaboratory comparison– The laboratory shall participate in anintralaboratory comparison programme [PT, EQA]appropriate to the examination andinterpretations of examination results.– The laboratory shall monitor the results andparticipate in the implementation of correctiveactions when predetermined performance criteriaare not fulfilled.52

CAP 2014 HbA1c 6.97 %Hb53

Sigma Proficiency AssessmentCAP 2014 GH2-02 6.97%Hb6.0Observed Inaccuracy 4.0-5.0-6.00.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0Observed Imprecision (%CV)54

Summary: Simplifying ISO 15189Development/Implementation Select Quality Manager Assemble Management Planning Team– Adopt PDCA for management requirements– Define policies, processes, procedures Appoint Technical Planning Team– Adopt PDCA for scientific technical requirements– Integrate Six Sigma concepts and tools Method Decision Chart, calculation of sigma-metrics QC Planning tools, such as “Westgard Sigma Rules”55

Summary: Simplifying ISO 15189Plan and DO Stages of 6σQMS Define quality requirements in form of TEa Select examination procedures with attentionto traceability Validate examination performance by– Use Method Decision Chart– Calculate sigma-metric Implement examination procedures– Training for use of SOPs56

Summary: Simplifying ISO 15189Check and Act Stages of 6σQMS Select/design SQC procedures for the sigmaquality achieved by the method Estimate Measurement Uncertainty fromintermediate QC data Assess acceptability of MU by calculation ofsigma-metric Assess acceptability of bias from PT/EQA bycalculation of sigma-metric57

Thank you for your kind attention!58

Any Questions?MUCH, MUCH MORECan be found onlineAt www.westgard.com59

Simplifying ISO 15189 Make it a PDCA process, not a list! Don't think of ISO 15189 as a long list of requirements -Document isn't organized and ordered to show the "flow" of requirements into a process -Better to understand and organize those requirements for a process and system Think 1st about process for implementing QMS