Getting the Right Answer –the Importance of Traceability
Graham JonesMBBS, BSc, DPhil, FRCPA, FAACB
Department of Chemical PathologySt Vincent’s Hospital, Sydney
Contents
• Interpreting laboratory results• Performing measurements• The Science of Metrology• Turning Science Into Caring
Laboratory Medicine
• Our goal: To improve patient health• Our tools: Laboratory tests• Our mechanism: Support medical decisions
• To optimise medical decisions we need toproduce accurate results
• If our work aversely affects medical decisionmaking, we are not achieving our goals
Numerical laboratory results
Example:Mr Bill Bloggs (DoB 1 Jul 1950)Sample Collected: 21 Aug 2012, 10:00 am
Test Result UnitsSerum sodium: 140 mmol/L
How is this number interpreted?
Interpreting laboratory results
All results are interpreted by comparison.Comparison may be with:• A clinical decision point
• A population reference interval
• A previous result from the patient
5-Aug 1-AugSodium: 140 145 mmol/L
Professor Per-Hyltoft Peteresen, Sydney 2005
Valid comparisons
• Your results must be unbiased relative to thecomparator
• Clinical Decision Point– Method used to perform the study
• Population reference interval– Method used for the reference interval study*
• Previous result on the patient– Method used for the previous result*
*May be your laboratory or elsewhere
Adverse clinical effects - example
• LDL Cholesterol target is 2.5 mmol/L (frompatient outcome studies)
• Laboratory method has positive bias relativeto trial methods
• Patient does not get treatment
Adverse financial effects
• Unnecessary testing costs due to analyticalfactors (patient recalls, follow-up, treatment):
• Germany 1.5 Billion US$ per year– German Health Report 1998
• USA 7.5 Billion US$ per year– Willie May, Chief Analytical Chemistry NIST
• Australia? (0.5 Billion A$)
Murphy KE et al. J. Anal. At. Spectrom., 2002, 17, 469–477
Laboratory measurements
• All numerical laboratory measurements aremade by comparison
• Analyte concentration in the sample iscompared with concentrationin the assay calibrators.
• Done using a standard curve
• Value assignment ofcalibrators establishes assay trueness (bias)
Routine method compares patient samplewith calibrator value to give result.
Calibrator value “sets” assay trueness / bias
Value assignment of Calibrators
• Calibrators have values assigned bycomparison with manufacturer mastercalibrators
• This process uses a superior assay to fieldmethod
The top of the traceability chain
• All assays are “anchored” in one of thefollowing– A Material– A Method (eg Enzymes)
Reference Materials
• Certified Reference Materials– Produced by National Measurement Institutes– Highly purified– Purity verified (and certified)– Very accurately weighed (and certified)– Reconstituted very accurately
• May also be “Matrix matched” eg urine, serum– Values assigned by comparison with pure materials
Other Reference materials
• International conventional calibrator– Eg WHO standards
• Other suppliers– Eg US Pharmacopoeia, commercial suppliers
• Manufacturer’s In-house materials
Reference Methods
• For some analytes the a method defines thetrue result
• Examples: IFCC methods for AST, ALT, ALP• Assays NOT calibrated with pure material
• For most analytes reference methods arecalibrated by a material
• Examples: Isotope Dilution Mass Spectrometry
Who decides?
The top of the chain is vital to accuracy.What Reference Material or Method is the top
of the Traceability Chain?
Joint Committee on Traceability inLaboratory Medicine
• JCTLM - Joining of:– Metrology Community (BIPM)– Laboratory Medicine Community (IFCC)– Accreditation Community (ILAC)
• Different languages, different journals, differenttraditions, different history
• Aim to bring rigour and processes of metrology tolaboratory medicine
Metrology - BIPM
Bureau International de Poids et Mesures(International Bureau of Weights and Measures)
(Pont de Sevres, Paris)
Metre Convention
• The Metre Convention (1875)• Treaty to oversee the keeping of metric
standards.• In 1960, the system of units it established was
renamed the "International System of Units"(Système international d'unités or SI).
• 56 signatory countries in 2012• “..to promote world wide uniformity in units of
measurement..”
Metrology in practice
• International network of Laboratories– National Measurement Institutes
• International Treaties– Recognition of measurements
• Metrology Research– All aspects
• Legal Metrology
Measurements in general
• Ordering a machine part to fit• Selling tonnes of wheat• Knowing the time• GPS navigation
• These are all possible because of metrology!
The kilogram
This international prototype, made ofplatinum-iridium, is kept at the BIPMunder conditions specified by the 1stCGPM in 1889
Achim Leistner at the AustralianCentre for Precision Optics holds
a 1-kg silicon crystal sphere.
Metrology
• In the lab – already responsible for:• Volumes, weights, light measurements,
lengths, current …• The building itself, the trucks that make
deliveries …
• Now through the JCTLM involved in assayaccuracy!
JCTLM Output
• Database of “higher Order … ”
• Reference Materials• Reference Methods• Reference Laboratories
JCTLM Review Teams
• Blood Cell Counting– Dr Mingting Peng
• Blood Groupings• Coagulation Factors• Metabolites and Substrates• Microbial Serology• Non-electrolyte Metals• Non-peptide Hormones• Vitamins and Micronutrients• Electrolytes and Blood Gases
• Drugs• Nucleic Acids• Proteins• Enzymes
Each team has volunteer scientists assessing materials and methods
Relevant ISO Standards• ISO 17511 In vitro diagnostic medical devices -
– Measurement of quantities in biological samples– Metrological traceability of values assigned to calibrators
and control materials
• ISO 18153 Metrological traceability of values forcatalytic concentration of enzymesassigned to calibrators and control materials
• ISO 15193 Reference measurement procedures• ISO 15194 Description of reference materials• ISO 15195 Reference Measurement Laboratories
Progress: 2002 - 2012
• 250 materials listed for 141 analytes• 157 methods for 80 analytes• 63 laboratory services for 32 analytes
• Still more to do!
ElectrolytesCalcium RM (P,S) RMP RMSChloride RM (P,S) RMP RMSLithium RM (P,S) RMP RMSMG RM (P,S) RMP RMSPhosphate RMPPotassium RM (P,S) RMP RMSSodium RM (P,S) RMP RMSEnzymesALT RMP RMSAST RMP RMSALP RMP RMSAmylase RMP RMSGGT RMP RMSLD RMP RMSCK RMP RMS
Metabolites and SubstratesBilirubin RM (P) RMPCholesterol RM (P,M) RMP RMSCreatinine RM (P,M) RMP RMSEthanol RM (P)5-tetrahydrofolic acid RMPGlucose RM (P,M) RMP RMSHDL-Cholesterol RMPHomocysteine RM (P,M) RMPLDL-Cholesterol RMP25-OH Vitamin D3 RM (M) RMPTriglycerides RM (M) RMP RMS(total glycerides)Urea RM (P,M) RMP RMSUrate RM (P,M) RMP RMS
Metabolites and SubstratesBilirubin RM (P) RMPCholesterol RM (P,M) RMP RMSCreatinine RM (P,M) RMP RMSEthanol RM (P)5-tetrahydrofolic acid RMPGlucose RM (P,M) RMP RMSHDL-Cholesterol RMPHomocysteine RM (P,M) RMPLDL-Cholesterol RMP25-OH Vitamin D3 RM (M) RMPTriglycerides RM (M) RMP RMS(total glycerides)Urea RM (P,M) RMP RMSUrate RM (P,M) RMP RMS
Creatinine Assays - 2013
• 43 manufacturers; 58 assays• 21 IDMS aligned* (others maybe or can be?)• Sources: IFCC survey responses; Internet
search for IFUs; personal knowledge(incomplete)
• Examples:– Calibration with aqueous calibrator may cause a
systematic error in automated procedures. Inthese cases it is recommended to use a serumcalibrator (Sentinal)
Creatinine Assays - Comments
• Many non-IDMS aligned assays available• Manufacturer’s PI / IFU often not helpful*
– Reference “NIST SRM”, not “IDMS”, not stated asassays traceable
– Some note need for Serum calibrators
• Assays in developing countries poorlyunderstood
Cystatin C
• 16 assays from 16 companies• 4 Aligned to IFCC reference material ERM
®DA471. (7 – A Grubb +3)• Dako and Abbott (only outside USA)• Roche for 2014• Options
– Advice to labs (formulae, which assays)– List of assays and formulae
Manufacturers
We want them to:• Select appropriate reference materials /
methods• Transfer values to calibrators accurately
(with minimal uncertainty)• Tell us what they are doing
– Traceability statement
Serum Creatinine Calibrators
Traceable to SRM 967 using IDMS (both JCTLM listed)
Value within 2.5% of true value
Calibrator lots should not change >+/- 2.5%
Traceability Statements
Laboratories
• Choose methods which are:– Traceable to good references (JCTLM listed)– Have low uncertainties for calibrators– Minimise changes over time
• Select and promote unbiased comparators– Common decision points, reference intervals– “4th Pillar”
• Confirm performance with EQA, standards– “5th Pillar”
Proposal
• The AACB and RCPA should adopt thefollowing principle:
• “Laboratories should use assays traceable toJCTLM-listed materials, methods andlaboratories where possible and feasible”
• Thus: LDH LPAST and ALT with P5PGGT to IFCC not Szasz
Personal Observations…
• If labs can do it differently, they will• Unless EQA can show labs are the same, assume
they are not• If EQA shows lab differences, it will not fix itself• Action requires people talking, setting standards
and doing• Action can be local, national or international
Conclusions
• The science of metrology backs most of whatwe do– Weights, volumes, currents, lights
• Assay traceability is being added to this list• Ongoing work by metrologists, manufacturers
and laboratories is required• We are part of the traceability chain in our
labs with the choices we make: choice ofmethod/supplier, how it is run; QC;QA…
• As professional organisations we fill a vital roleallowing actions as a group.