What EQAWarden knows about chemical pathology, and where every fact came from.

Method families

• Indirect ISE (pre-diluted sample; Atellica CH A-LYTE Integrated Multisensor, 1:10 dilution)
• Direct ISE (undiluted; blood gas and point-of-care analysers)
• Flame atomic emission (historical routine; basis of Siemens standardisation)
• Enzymatic (beta-galactosidase activation) photometry - rare

EQA pitfalls the engine knows

• Indirect vs direct ISE matrix (electrolyte-exclusion) effect
• Evaporation or reconstitution-volume error
• ISE calibration/slope or multisensor ageing
• Surfactant/viscosity additives in processed material

Investigation checks, in order

1. Confirm the comparator: indirect-ISE peer group (not all-method mean if direct-ISE users are pooled).
2. Look at the same sample's Cl, K, Ca, total protein and albumin deviations: a shared % shift indicates evaporation/reconstitution, not the sodium electrode.
3. Review IMT calibration (slope, Standard A/B lot), multisensor age and any diluent/reference lot change around the EQA run date.
4. Check internal QC at similar concentration on the run date and the preceding week for drift.
5. Confirm the vial was reconstituted with a calibrated volumetric pipette and analysed promptly (not re-run next day from an open cup).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Indirect ISE (valinomycin membrane; Atellica CH A-LYTE IMT, 1:10 dilution)
• Direct ISE (blood gas/POC)
• Flame atomic emission (reference/historical)
• Enzymatic (pyruvate kinase) photometry - rare

EQA pitfalls the engine knows

• Evaporation / reconstitution volume
• Indirect vs direct ISE separation
• Low-end constant error
• Haemolysis is not a valid explanation for EQA

Investigation checks, in order

1. Compare with the indirect-ISE peer group; check whether Na and Cl in the same vial show the same % shift (handling).
2. Regress the cycle's results on targets: a non-zero intercept suggests reference-electrode/low-end issues, a slope error suggests calibration/slope.
3. Review IMT calibration slope, multisensor age and Standard A/B lots; check QC on the run date.
4. Confirm prompt analysis after reconstitution (no overnight re-run).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Indirect ISE (Atellica CH A-LYTE IMT, 1:10 dilution)
• Direct ISE (blood gas/POC)
• Coulometric (amperometric) titration - reference method basis
• Colorimetric mercuric thiocyanate/ferric - legacy

EQA pitfalls the engine knows

• Anion interference on the chloride ISE (salicylate, bromide, iodide, fluoride, thiocyanate)
• Indirect vs direct ISE matrix effect
• Evaporation / reconstitution

Investigation checks, in order

1. Check the same-survey targets for salicylate/other spiked drugs: a positive Cl bias confined to the drug-rich sample indicates ISE selectivity loss.
2. Compare Cl and Na % deviations: identical shifts implicate handling or the shared reference system; Cl-only shifts implicate the Cl membrane.
3. Review multisensor age/slope and replace if selectivity is suspect.
4. Use the indirect-ISE peer group as comparator.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic PEP carboxylase / malate dehydrogenase with NADH analogue, decrease in absorbance at 410/478 nm (Atellica CH CO2_c)
• Acidification with pCO2 electrode or CO2-ISE (some analysers)
• Calculated bicarbonate from pH and pCO2 (blood gas) - not comparable with serum tCO2

EQA pitfalls the engine knows

• CO2 loss from opened/reconstituted vials and sample cups (reads low)
• Non-commutability of lyophilised material for tCO2
• Reagent CO2 absorption / onboard reagent age

Investigation checks, in order

1. Ask how long the reconstituted sample stood uncapped before CO2 was measured, and whether it was re-run (the commonest cause of low results).
2. Check whether other labs in the enzymatic PEPC peer group show a similar negative spread (material/handling) versus a lab-specific shift.
3. Review CO2_c reagent well open date, calibration age and QC trend.
4. Confirm the result is serum tCO2, not a calculated blood-gas bicarbonate entered by mistake.
5. If persistent, test a freshly reconstituted vial immediately (analyse CO2 first).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Urease + glutamate dehydrogenase (GLDH), NADH consumption, kinetic at 340/410 nm (Atellica CH UN_c)
• Urease with conductimetric or ammonium-ISE detection
• Diacetyl monoxime colourimetry - legacy

EQA pitfalls the engine knows

• Ammonium contamination of material, reagent water or sample (reads high)
• Unit/reporting convention error (urea vs BUN)
• High conjugated bilirubin in the same sample (reads low)

Investigation checks, in order

1. Confirm unit convention (urea mmol/L vs BUN) in analyser, LIS and RCPAQAP enrolment - look for ratios 2.8/0.357/6.0/2.
2. Regress across the cycle: positive intercept suggests ammonium contamination; slope error suggests calibration.
3. Check same-sample ammonia and bilirubin targets for interference.
4. Review calibration and QC at the relevant level; check reagent water.

Unit traps

• BUN (mg/dL as nitrogen) x 0.357 = urea mmol/L; urea (mg/dL) x 0.1665 = urea mmol/L - ratio 2.14 between the two mass conventions.
• 'Urea nitrogen' expressed in mmol/L of N would be 2 x urea mmol/L (rare; a factor-2 discrepancy suggests this).
• A result/target ratio of ~2.8 (or 0.357) indicates BUN mg/dL entered as mmol/L (or vice versa).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Kinetic Jaffe (alkaline picrate), rate-blanked, with fixed intercept correction (-0.3 mg/dL = -26.5 umol/L for serum/plasma) - Atellica CH Crea_2
• Enzymatic (creatininase/creatinase/sarcosine oxidase/peroxidase-Trinder) - Atellica CH ECre_2 (principle per chart 'Enzymatic'; reaction scheme textbook)
• IDMS (GC or LC) - reference methods

EQA pitfalls the engine knows

• Jaffe fixed intercept correction does not match EQA matrix
• Non-commutability of processed material for creatinine
• Interferents co-spiked in multi-analyte material (bilirubin, glucose, ketoacids, pyruvate, cefoxitin)
• ECre_2 NAPQI interference (recall Z-2283-2020)

Investigation checks, in order

1. Identify the method on the report (Crea_2 Jaffe vs ECre_2 enzymatic) and compare with that peer group, not the all-method mean.
2. Regress the cycle's results on targets: intercept (constant) error points to Jaffe matrix/chromogen or blanking issues; slope error to calibration.
3. Check same-sample bilirubin, glucose and drug targets for interference-driven sample-specific outliers.
4. Check units (umol/L vs mg/dL vs mmol/L) - ratios 88.4 or 1000.
5. Review calibration (CHEM CAL lot), QC and reagent lot changes; the 26.5 umol/L correction is fixed and cannot be adjusted by the user.

Unit traps

• Ratio ~88.4 (or 0.0113) = mg/dL vs umol/L confusion.
• Ratio 1000 = mmol/L vs umol/L (common in eGFR-related LIS fields).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Hexokinase/G6PD, NAD reduction at 340/410 nm with sample blank (Atellica CH GluH_3)
• Glucose oxidase-peroxidase (Trinder), 505/694 nm (Atellica CH GluO)
• Glucose oxidase with oxygen or H2O2 electrode; glucose dehydrogenase (POC)

EQA pitfalls the engine knows

• Microbial glycolysis in reconstituted or stored material (reads low)
• Hexokinase vs glucose oxidase method-group differences in processed material
• Calibration/proportional error

Investigation checks, in order

1. Confirm which glucose TDef (GluH_3 hexokinase vs GluO oxidase) generated the result and use the matching peer group.
2. Check time from reconstitution to analysis (negative drift suggests glycolysis/contamination).
3. Check unit (ratio ~18).
4. Review calibration, QC and reagent lots.

Unit traps

• Ratio ~18.0 (or 0.0555) = mg/dL vs mmol/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• o-Cresolphthalein complexone (CPC) with 8-hydroxyquinoline Mg masking, 545/658 nm (Atellica CH Ca)
• Arsenazo III at pH 5.9, 658/694 nm (Atellica CH CA_2)
• Atomic absorption / ICP-OES / ID-ICP-MS - reference methods

EQA pitfalls the engine knows

• Small relative errors breach tight limits
• CPC vs arsenazo III method-group differences
• Gadolinium contrast or chelator contamination (patients, not EQA)

Investigation checks, in order

1. Identify the dye (CPC 'Ca' vs arsenazo 'CA_2') and compare with that peer group.
2. Compare the % deviation with sodium, albumin and total protein in the same vial (handling/volume).
3. Check units (ratio 4.0 or 2).
4. Review calibration and QC; calcium's tight APS means small calibrator-lot shifts matter.

Unit traps

• Ratio ~4.0 (or 0.25) = mg/dL vs mmol/L.
• Ratio 2 = mEq/L vs mmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Xylidyl blue in alkaline medium with EGTA to exclude calcium, 505/694 nm (Atellica CH Mg)
• Calmagite, methylthymol blue, chlorophosphonazo III dyes; enzymatic (isocitrate dehydrogenase / glucokinase)
• Atomic absorption / ICP / ID-MS - reference

EQA pitfalls the engine knows

• Dye-method matrix differences
• Chelator (EDTA/citrate) in material (reads low)
• Handling (reconstitution/evaporation)

Investigation checks, in order

1. Compare with the xylidyl-blue peer group; check units (ratio 2.43 or 2).
2. Check co-measured Ca and Na % deviations for handling errors.
3. Review calibration and QC.

Unit traps

• Ratio ~2.43 (or 0.411) = mg/dL vs mmol/L.
• Ratio 2 = mEq/L vs mmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Unreduced phosphomolybdate UV endpoint at 340/658 nm (Atellica CH IP)
• Reduced molybdenum blue; enzymatic (purine nucleoside phosphorylase)
• Ion chromatography (DGKL reference method)

EQA pitfalls the engine knows

• Turbidity of reconstituted material at 340 nm (reads high)
• Paraprotein precipitation (patients) (reads high)
• Unit convention (as P)

Investigation checks, in order

1. Check units (ratio 3.1).
2. Look for sample-specific positive bias with turbid/lipaemic material (UV method).
3. Compare with the phosphomolybdate UV peer group.
4. Review calibration and QC.

Unit traps

• Ratio ~3.1 (or 0.323) = mg/dL (as P) vs mmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Uricase / peroxidase (Trinder: 4-aminophenazone + TOOS), endpoint 545/694 nm (Atellica CH UA)
• Uricase UV (293 nm) direct
• ID-GC/MS, ID-LC/MS - reference

EQA pitfalls the engine knows

• Trinder negative interference (reducing agents) (reads low)
• Unit scaling (mmol/L vs umol/L)
• Calibration/proportional error

Investigation checks, in order

1. Check unit (mmol/L in Australia; ratio 1000 if umol/L entered).
2. Check for concordant negative bias in other Trinder assays on the same sample.
3. Review calibration and QC; compare with the uricase-peroxidase peer group.

Unit traps

• Ratio 1000 = umol/L vs mmol/L.
• Ratio ~16.8 (or 0.0595) = mg/dL vs mmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Biuret end point (alkaline tartrate, cupric sulfate), bichromatic 545/694 nm, 37 C, on-board 1 in 5 predilution, calibrated with CHEM CAL: Atellica CH TP_2
• Biuret end point, earlier Atellica CH Total Protein II (TP); no package insert in the laboratory's library, assay chart data only; calibrated with CHEM CAL
• Pyrogallol red-molybdate dye binding with a 5-level logit calibration for urine and CSF protein (Atellica CH UCFP): a different matrix, calibrator and chemistry...
• Reference measurement procedure: INSTAND e.V. RMP for total protein in serum (JCTLM C17RMP3)

EQA pitfalls the engine knows

• Assay version: TP (Total Protein II) versus TP_2
• Shared CHEM CAL calibrator handling or lot value
• Pack calibration interval shorter than onboard life
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault
• Icterus, haemolysis, dextran or paraprotein in a particular EQA sample
• Freeze-thaw of liquid (fresh-frozen) material

Investigation checks, in order

1. Confirm the test definition (TP_2 or TP) and that the RCPAQAP method code and peer group match it.
2. Compute the percentage deviation of the other analytes in the same sample: a common shift (including Na and Cl) points to reconstitution or evaporation; a shift shared only by photometric CH tests poi...
3. Compare the same survey's albumin (BCG), calcium and creatinine: a shared proportional shift points to CHEM CAL handling or calibration.
4. Review the TP_2 calibration history (lot 180 days, pack 58 days), the pack's onboard time and IQC around the analysis date.
5. Check the sample's bilirubin target and description for icterus, haemolysis or paraprotein.

Unit traps

• Ratio 10 (or 0.1) = g/dL vs g/L confusion.
• A urine or CSF protein result (mg/L) mapped to serum total protein gives a ratio near 1000 or an implausible value.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Bromocresol green (BCG) dye binding at pH 4.2, end point 596/694 nm, calibrated with CHEM CAL, traceable to a BCG reference method using NIST SRM 927: Atellica...
• Bromocresol purple (BCP) dye binding at pH 4.9, end point 596/694 nm, own single-level AlbP CAL plus zero, traceable to ERM-DA470k: Atellica CH AlbP (OUS editio...
• Immunochemical (nephelometric or turbidimetric) albumin, calibrated to ERM-DA470k/IFCC; IFCC Committee on Plasma Proteins reference method (JCTLM C1RMP_P4)

EQA pitfalls the engine knows

• BCG versus BCP method group
• AlbP calibrator: single level, 8 hour life, fussy reconstitution
• Alb (BCG) calibrated from the shared CHEM CAL
• Haemolysis or lipaemia in a particular EQA sample (reads high)
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault
• Unit and conversion factor configuration

Investigation checks, in order

1. Identify the dye (Alb = BCG, AlbP = BCP) and confirm the RCPAQAP method code; compare with that dye's peer group.
2. Check whether the deviation is larger at the low-albumin sample (BCG globulin effect) or constant in percentage (calibration).
3. For AlbP: review the AlbP CAL reconstitution time (8 h life), lot value and pack calibration (every 8 days). For Alb: compare the other CHEM CAL assays of the same survey.
4. Compute the percentage deviation of the other analytes in the same sample for a reconstitution or predilution signature.
5. Check the sample description for haemolysis or lipaemia.

Unit traps

• Ratio 10 (or 0.1) = g/dL vs g/L confusion; check the TDef unit factor.
• A urine albumin (mg/L) result mapped to serum albumin gives a ratio near 1000.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Vanadate oxidation to biliverdin at pH 2.9 with detergent, fall in absorbance at 451/545 nm, CHEM CAL or BILI CAL, traceable to the AACC reference method with N...
• Diazo (2,4-dichloroaniline) with detergent accelerator, 545/658 nm, BILI CAL only, traceable to NIST SRM 916: Atellica CH D_TBil (OUS; adults and neonates)
• Reference measurement procedures: Doumas (JCTLM NRMeth 29), AACC (C3RMMP28) and RfB spectrophotometric (C15RMP6)
• Direct spectrophotometry (neonatal point of care) and enzymatic (bilirubin oxidase) methods on other platforms

EQA pitfalls the engine knows

• Vanadate versus diazo method group
• Light exposure of the reconstituted sample (reads low)
• Calibrator choice and short open-vial life
• Imprecision and rounding at low concentration
• Haemolysis, lipaemia or eltrombopag in a particular sample
• Freeze-thaw of liquid serum material
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Identify the chemistry (TBil_2 vanadate or D_TBil diazo) and compare with that method group; expect vanadate and diazo groups to separate.
2. Ask about light protection and time from reconstitution to analysis; compare with conjugated bilirubin in the same vial.
3. Review the calibrator used (CHEM CAL or BILI CAL), its reconstitution time (8 h life) and the calibration dates (lot 60, pack 30 days for TBil_2; lot 90, pack 7 days for D_TBil).
4. At low concentrations assess the deviation in umol/L against low-level IQC imprecision before calling it a bias.
5. Check the sample description for haemolysis or lipaemia.

Unit traps

• Ratio about 17.1 (or 0.058) = mg/dL vs umol/L confusion; TDef 1.4 added a unit conversion factor, check it is configured consistently.
• Conjugated and total bilirubin swapped in the result mapping gives a large sample-specific negative (or positive) deviation.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Vanadate oxidation of direct-reacting bilirubin at about pH 3 (tartrate buffer, detergent), fall in absorbance at 451/545 nm, CHEM CAL or BILI CAL, traceable to...
• Diazo (2,4-dichloroaniline) in strong acid without accelerator, 545/658 nm, BILI CAL only, traceable only to internal gravimetric standards: Atellica CH D_DBil...
• Direct bilirubin is method-defined (conjugated plus a variable part of delta bilirubin and, for vanadate, some photoisomers); no RMP or CRM was found in the JCT...

EQA pitfalls the engine knows

• Vanadate versus diazo: a method-defined measurand
• Light exposure: photoisomers read as direct bilirubin by vanadate
• Haemolysis strongly lowers diazo direct bilirubin (reads low)
• Low concentrations and rounding
• Calibrator and reagent lot changes
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Identify the chemistry (DBil_2 vanadate or D_DBil diazo) and compare only with that group.
2. Compare the deviation of total bilirubin in the same vial: opposite directions (direct up, total down) suggest light-derived photoisomers with vanadate.
3. Check the sample's H index and description; diazo direct bilirubin falls sharply with haemolysis.
4. Express low-level deviations in umol/L against low IQC SD and the APS absolute limit.
5. Review calibrator (CHEM CAL or BILI CAL, 8 h life), calibration dates and reagent lot changes.

Unit traps

• Ratio about 17.1 = mg/dL vs umol/L.
• Direct and total bilirubin swapped in the mapping.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• IFCC-type kinetic method: p-nitrophenyl phosphate in 2-amino-2-methyl-1-propanol (AMP) buffer pH 10.25 with Mg and Zn (HEDTA), 410/478 nm rate, 37 C, ALP_2 CAL...
• Diethanolamine (DEA) buffer methods (historical, give roughly 2-3 fold higher activities) and other buffers on other platforms
• IFCC reference measurement procedure at 37 C (JCTLM C7RMP2R); no IFCC certified enzyme reference material for ALP; JCTLM lists GBW(E)091042 (frozen human serum)

EQA pitfalls the engine knows

• ALP activity rises with standing time after reconstitution
• Reaction temperature away from 37 C
• ALP_2 CAL single-level calibrator
• Method group and isoenzyme or source differences
• Icterus or haemolysis in a particular sample
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Record the time from reconstitution to ALP measurement; ALP rises on standing.
2. Compare with the AMP/IFCC (Atellica) peer group rather than the all-method median.
3. Look at the other enzymes of the same survey for a shared temperature-type shift.
4. Review ALP_2c calibration (lot 60, pack 17 days), the ALP_2 CAL vial and pack onboard time, and IQC.
5. Compute the percentage deviation of other analytes in the same vial for a reconstitution signature.

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.
• A DEA-method peer group or target would be about 2-3 fold higher than AMP methods; this is a method mismatch, not a unit error.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• NADH-coupled kinetic method (L-alanine, alpha-ketoglutarate, lactate dehydrogenase), 340/410 nm, 37 C, WITHOUT pyridoxal-5'-phosphate (P5P), ENZ 2 CAL, traceabl...
• Same reaction WITH P5P added manually to reagent 1 (300 uL P5P_L per well) as recommended by IFCC: Atellica CH ALTPLc
• IFCC reference measurement procedure at 37 C with P5P (JCTLM NRMeth 67); certified reference material ERM-AD454k/IFCC (recombinant human cytosolic ALT)

EQA pitfalls the engine knows

• P5P versus no P5P method group
• ALTPLc pack preparation and very short pack calibration
• Shared ENZ 2 CAL calibrator
• Reaction temperature away from 37 C
• Non-human or recombinant enzyme in processed EQA material
• Haemolysis in a particular sample (reads high)
• Stability of reconstituted material
• Substrate depletion flag and range
• EQA vial reconstitution volume or evaporation

Investigation checks, in order

1. Confirm ALT (no P5P) or ALTPLc (P5P) and compare with the matching method group.
2. Compare AST of the same survey: a shared shift suggests ENZ 2 CAL or temperature; check the other enzymes for temperature.
3. For ALTPLc: check the P5P preparation record, load date and the 4-day pack calibration.
4. Record the time from reconstitution to analysis.
5. Review calibration dates (lot 131 days, pack 30 days for ALT) and IQC around the run.

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• NADH-coupled kinetic method (L-aspartate, alpha-ketoglutarate, malate dehydrogenase), 340/410 nm, 37 C, WITHOUT P5P, ENZ 2 CAL, IFCC-traceable: Atellica CH AST
• Same reaction WITH P5P added manually (300 uL P5P_L per well): Atellica CH ASTPLc
• IFCC reference measurement procedure at 37 C (JCTLM NRMeth 68); CRM ERM-AD457k/IFCC (recombinant human cytosolic AST)

EQA pitfalls the engine knows

• P5P versus no P5P method group
• Haemolysed or partly lysed EQA sample (reads high)
• ASTPLc pack preparation and pack calibration
• Shared ENZ 2 CAL calibrator
• Constant offset at low activity
• Reaction temperature away from 37 C
• Non-human or recombinant enzyme in processed EQA material
• Stability of reconstituted material
• EQA vial reconstitution volume or evaporation

Investigation checks, in order

1. Confirm AST (no P5P) or ASTPLc and compare with the matching method group.
2. Check the sample description and LD of the same sample for haemolysis.
3. Compare ALT (shared ENZ 2 CAL) and the other enzymes (temperature) in the same survey.
4. Regress the cycle for a constant low-end component.
5. Review calibration (lot 131, pack 30 days for AST; pack 11 days for ASTPLc) and the P5P preparation record.

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Kinetic L-gamma-glutamyl-3-carboxy-4-nitroanilide with glycylglycine (Szasz/Persijn and van der Slik type), 410/478 nm, 37 C, ENZ 1 CAL, IFCC-traceable: Atellic...
• IFCC reference measurement procedure at 37 C (JCTLM NRMeth 69); CRM ERM-AD452/IFCC (pig kidney GGT)

EQA pitfalls the engine knows

• Method groups for GGT
• Assay version and frequent pack changes (GGT vs GGT_2)
• Shared ENZ 1 CAL calibrator
• Haemolysis: opposite direction in the two assay versions
• Reaction temperature away from 37 C
• Non-human or recombinant enzyme in processed EQA material
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Confirm the test definition (GGT or GGT_2) and compare with the Atellica peer group.
2. Compare LD, lipase and cholinesterase (shared ENZ 1 CAL) and all enzymes (temperature) in the same survey.
3. Line up pack changes (6-day packs for GGT) and calibrations with the EQA date.
4. Check the sample description for haemolysis; direction depends on the assay version.
5. Check units (ratio 60 for ukat/L).

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Forward reaction lactate to pyruvate with NAD, Tris buffer (Richards), 340/410 nm rate, 37 C, ENZ 1 CAL, IFCC-traceable (IFCC-453): Atellica CH LDLP
• Reverse (pyruvate to lactate) methods on other platforms give different, typically higher, activities
• IFCC reference measurement procedure (lactate to pyruvate) at 37 C (JCTLM NRMeth 66); CRM ERM-AD453k/IFCC (recombinant human LD1)

EQA pitfalls the engine knows

• Reaction direction and buffer method groups
• Cold storage of the reconstituted sample (reads low)
• Haemolysis in a particular sample (reads high)
• Isoenzyme composition of calibrator and EQA material
• Shared ENZ 1 CAL calibrator
• Reaction temperature away from 37 C
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Compare with the L-P Atellica peer group.
2. Ask whether the reconstituted vial was refrigerated before analysis (LD falls in the cold).
3. Check the sample description and AST for haemolysis.
4. Compare GGT, lipase and cholinesterase (ENZ 1 CAL) and all enzymes (temperature) in the same survey.
5. Review calibration (lot 60, pack 28 days) and IQC.

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• IFCC-adapted modified Szasz: creatine phosphate plus ADP with N-acetylcysteine activation, hexokinase and G6PD coupled NADPH rise at 340/596 nm, 37 C, ENZ 3 CAL...
• IFCC reference measurement procedure at 37 C (JCTLM NRMeth 65); CRM ERM-AD455k/IFCC (recombinant human CK-MM)

EQA pitfalls the engine knows

• CK activity loss after reconstitution
• Reaction temperature away from 37 C
• ENZ 3 CAL: frozen calibrator handling
• Haemolysis in a low-activity sample (reads high)
• Very wide automatic dilution
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Record the time from reconstitution to CK measurement (activity falls on standing).
2. Look at all enzymes of the same survey: CK moves most with a temperature error.
3. Review the ENZ 3 CAL thaw and opening record and CK calibration dates.
4. Check the sample description for haemolysis at low CK.
5. Compute the percentage deviation of other analytes in the same vial.

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.
• CK (activity, U/L) and CK-MB (mass, ug/L) must not be crossed in the mapping.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (acridinium ester anti-CK-MB Lite Reagent, anti-CK-BB on paramagnetic particles), mass in ng/mL (ug/L), calibrate...
• CK-MB activity by immunoinhibition of the M subunit (U/L), which also counts CK-BB and macro-CK activity
• No RMP or CRM found in the JCTLM keyword search; recombinant CK-MB2 was proposed by an AACC committee as a common calibrator and reduced between-manufacturer bi...

EQA pitfalls the engine knows

• Method-specific calibration of CK-MB mass
• Low dilution recovery above the measuring range (reads low)
• CKMB CAL reconstitution water temperature and life
• Spiked antigen with different immunoreactivity
• Atellica IM system-wide signal or wash fault
• Imprecision at low concentration
• High-dose hook

Investigation checks, in order

1. Confirm the result is CK-MB mass (ug/L) and the RCPAQAP method group is Atellica IM (not an activity group).
2. Check units: ug/L = ng/mL; nmol/L x 80 = ug/L.
3. For a high sample check automatic dilution (recovery about 81%).
4. Review CKMB CAL reconstitution (water 20-25 C), master curve and calibration dates.
5. Compare other Atellica IM assays and IQC for a system-wide shift.

Unit traps

• Ratio 80 (or 0.0125) = nmol/L vs ug/L.
• U/L (activity) reported against a mass group, or vice versa: no conversion exists.
• CK and CK-MB crossed in the result mapping.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Ethylidene-blocked 4-nitrophenyl maltoheptaoside with alpha-glucosidase, 410/694 nm rate, 37 C, SPCL CHEM CAL, traceable to IRMM/IFCC-456 and stated commutable...
• Pancreatic amylase (PAMY_2): same reaction after antibody inhibition of salivary amylase; a separate measurand, not mapped to this analyte
• Original Atellica CH amylase test 'Amylas' listed on the 2020 assay chart without a calibrator (factor-based)
• IFCC reference measurement procedure at 37 C (JCTLM NRMeth 70); CRM ERM-AD456/IFCC (purified human pancreatic amylase)

EQA pitfalls the engine knows

• Between-method offset
• Shared SPCL CHEM CAL calibrator
• Total versus pancreatic amylase mapping
• Reaction temperature away from 37 C
• Icterus in a particular sample (reads low)
• Fall with storage (reads low)
• Non-human or recombinant enzyme in processed EQA material
• EQA vial reconstitution volume or evaporation

Investigation checks, in order

1. Confirm the reported value is total amylase (AMY_2), not pancreatic amylase, and compare with the Atellica peer group.
2. Compare lactate and TIBC (shared SPCL CHEM CAL) and all enzymes (temperature) in the same period.
3. Review SPCL CHEM CAL reconstitution (7-day life) and AMY_2 calibration dates.
4. Check the sample's bilirubin target for icteric interference.
5. Check units (ratio 60 for ukat/L).

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.
• Pancreatic amylase submitted as total amylase.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Colorimetric rate with DGGMR (1,2-o-dilauryl-rac-glycero glutaric acid 6'-methylresorufin ester), colipase and deoxycholate, 571/694 nm, 37 C, ENZ 1 CAL, tracea...
• Other routine principles (turbidimetric triolein, 1,2-diglyceride coupled enzymatic, titrimetric reference-type)
• No IFCC reference procedure; no RMP or CRM found in the JCTLM keyword search

EQA pitfalls the engine knows

• Method-specific lipase activity
• Short reagent life and pack calibration
• ENZ 1 CAL onboard residence limited to 48 hours for lipase
• Loose imprecision
• Lipaemia or haemolysis in a particular sample (reads high)
• Carry-over from preceding tests
• Reaction temperature away from 37 C
• Non-human lipase in calibrator and EQA material
• EQA vial reconstitution volume or evaporation

Investigation checks, in order

1. Compare only with the Atellica lipase peer group.
2. Check pack age (9 days), pack calibration and calibrator onboard residence (48 h limit).
3. Compare GGT, LD and cholinesterase (ENZ 1 CAL) and all enzymes (temperature).
4. Judge the deviation against lipase IQC imprecision.
5. Check the sample description for lipaemia or haemolysis.

Unit traps

• Ratio 60 (or 0.0167) = ukat/L vs U/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Butyrylthiocholine hydrolysis with hexacyanoferrate(III) reduction, absorbance fall at 410/658 nm, 37 C, ENZ 1 CAL, standardised to the molar absorptivity of th...
• Earlier Atellica CH cholinesterase assay (CHE); no insert in the laboratory's library
• Other substrates and principles (acetylthiocholine or propionylthiocholine with Ellman DTNB) on other platforms

EQA pitfalls the engine knows

• Substrate and assay version define the result
• Activity depends on sample dilution
• Shared ENZ 1 CAL calibrator
• HIL is an unlikely explanation
• Non-human or recombinant enzyme in processed EQA material
• Reaction temperature away from 37 C
• EQA vial reconstitution volume or evaporation
• On-board 1 in 5 predilution fault

Investigation checks, in order

1. Confirm the test definition (CHE_2 or CHE) and compare with the matching peer group.
2. Check units (U/L vs kU/L, ratio 1000).
3. Compare GGT, LD and lipase (ENZ 1 CAL) and all enzymes (temperature) in the same survey.
4. Review CHE_2 calibration (lot 180, pack 90 days) and IQC.
5. Compute the percentage deviation of other analytes in the same vial.

Unit traps

• Ratio 1000 = kU/L vs U/L.
• Ratio 60 = ukat/L vs U/L (if reported in ukat/L).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic cholesterol esterase, cholesterol oxidase and peroxidase (Trinder: 4-aminoantipyrine and phenol) end point at 505/694 nm after an on-board 1 in 5 pre-...
• Abell-Kendall and isotope dilution GC/MS or LC/MS - reference measurement procedures

EQA pitfalls the engine knows

• Shared CHEM CAL calibrator vial or lot (lyophilised, 48 hours after reconstitution)
• Pack calibration is a zero (C0) adjustment, so a slope error persists until the next lot calibration
• Peroxidase (Trinder) end-point interference from reductants (reads low)
• Bilirubin in icteric EQA material
• Haemolysis in the EQA material (reads high)
• Non-commutable lyophilised material
• Reagent carry-over from a preceding test (reads low)
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check transcription and units (mmol/L versus mg/dL, ratio about 38.7) against the instrument printout.
2. Compare with the Atellica peer group (method code for Chol_2) before the all-method or reference target.
3. Review the Chol_2 calibration history: last lot calibration (50 days), pack calibrations (7 day C0 adjustment), CHEM CAL lot and vial reconstitution date.
4. Look at the other CHEM CAL analytes on the same analyser and at the other peroxidase end-point analytes of the same sample for a shared shift.
5. Check H and I indices and the bilirubin target of the deviating sample.

Unit traps

• Ratio about 38.7 (or 0.0259) = mg/dL versus mmol/L confusion.
• The IFU uses 0.0259 to convert mg/dL to mmol/L; the app factor 0.025862 differs only in rounding.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic lipoprotein lipase, glycerol kinase, glycerol-3-phosphate oxidase and peroxidase (Trinder: 4-aminoantipyrine and 4-chlorophenol) end point at 505/694...
• Earlier Atellica CH Trig (GPO) test definition
• Total glycerol by isotope dilution GC/MS or LC/MS - reference measurement procedures

EQA pitfalls the engine knows

• Calibration scheme changed: 14-day interval, pack calibration only
• Shared CHEM CAL calibrator vial or lot (lyophilised, 48 hours after reconstitution)
• Free glycerol in the material read as triglyceride (reads high)
• Peroxidase (Trinder) end-point interference from reductants (reads low)
• Haemolysed material (reads high)
• Non-commutable processed material
• Lipoprotein stratification after freeze-thaw or incomplete mixing
• Reagent carry-over (triglyceride as donor and recipient)
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check transcription and units (mmol/L versus mg/dL, ratio about 88.5).
2. Compare with the Atellica peer group for Trig_2.
3. Read the Trig_2 calibration intervals configured (IFU Rev 06: lot 14 days, pack 7 days) and the calibration and CHEM CAL history.
4. Look for concordant shifts in the other CHEM CAL analytes and the other peroxidase end-point analytes of the same sample.
5. Check H and I indices, vial mixing and whether the result came from an auto-repeat dilution.

Unit traps

• Ratio about 88.5 (or 0.0113) = mg/dL versus mmol/L confusion (factor as triolein).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Homogeneous direct HDL cholesterol, accelerator selective detergent method with two reagents, peroxidase colour at 596/694 nm - Atellica CH HDLC
• Earlier Atellica CH Direct HDL Cholesterol (D_HDL, elimination/catalase)
• Ultracentrifugation with heparin-manganese precipitation and Abell-Kendall cholesterol - CDC reference method

EQA pitfalls the engine knows

• Direct HDL method on processed or lyophilised material
• Change between Atellica D-HDL and HDLC assays
• HDLC CAL handling (separate calibrator, 14 days after reconstitution)
• Conjugated bilirubin at low HDL-C (reads low)
• Peroxidase (Trinder) end-point interference from reductants (reads low)
• Low-level imprecision
• HDL-C rise on standing at room temperature (reads high)
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Confirm the tdef (HDLC or D_HDL) and method code; compare with that peer group.
2. Review the HDLC calibration (HDLC CAL lot and reconstitution date) and IQC.
3. Check whether total cholesterol and triglyceride of the same sample also deviate (shared vial or CHEM CAL issue) or HDL alone (HDLC calibrator, reagent or matrix).
4. Check the bilirubin target and I index of the sample, and the concentration (low HDL-C scatters more).
5. If the laboratory calculates LDL, remember an HDL error passes into calculated LDL with the opposite sign.

Unit traps

• Ratio about 38.7 = mg/dL versus mmol/L confusion (same factor as cholesterol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Homogeneous direct LDL cholesterol, two detergents, peroxidase colour (DSBmT and 4-aminoantipyrine) at 545/694 nm - Atellica CH LDLC
• Calculated LDL cholesterol: Friedewald (TC - HDL-C - TG/5 in mg/dL; TG/2.2 in mmol/L) or the NIH (Sampson) equation
• Beta-quantification (ultracentrifugation) - CDC reference method

EQA pitfalls the engine knows

• Direct LDL on lyophilised or frozen processed material
• Calculated versus direct LDL in the same comparison
• Formula, unit or constant error in a calculated LDL (ratio test, CentraLink or LIS)
• LDLC CAL used beyond 48 hours after reconstitution
• Reagent lot change at high LDL-C
• HIL effects near the 10% limit
• Peroxidase (Trinder) end-point interference from reductants (reads low)
• Siemens platforms differ for direct LDL
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Establish whether the submitted LDL is direct (LDLC) or calculated, and which equation; check the registered method.
2. If calculated, recompute from the submitted TC, HDL-C and TG and examine each input's deviation first.
3. If direct, compare with the Atellica LDLC peer group; review LDLC CAL reconstitution date, calibration dates and reagent lot changes.
4. Check H, I and L indices of the sample.
5. Check for a correlation factor, ratio-test formula or middleware calculation change.

Unit traps

• Ratio about 38.7 = mg/dL versus mmol/L confusion.
• Friedewald divisor: TG/5 applies in mg/dL and TG/2.2 in mmol/L; a formula using the wrong divisor gives a triglyceride-dependent error.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Ferrozine colourimetric, direct (no deproteinisation; acid release and ascorbate reduction with guanidine and detergent), 571/658 nm - Atellica CH Iron_2
• Ferene colourimetric, direct, with thiourea to suppress copper, 596/694 nm - Atellica CH Iron3
• Other chromogens (for example TPTZ) and deproteinised methods on other platforms

EQA pitfalls the engine knows

• Two Atellica iron chemistries in different peer groups
• Haemolysis releases red cell iron (reads high)
• Lipid emulsion lowers iron (reads low)
• Shared CHEM CAL calibrator vial or lot (lyophilised, 48 hours after reconstitution)
• Iron contamination of the vial, diluent water or pipette tips (reads high)
• Low-iron imprecision
• Liquid Serum Chemistry material from haemochromatosis donors
• Reagent carry-over before software mitigation
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check units (umol/L versus ug/dL, ratio about 5.6) and transcription.
2. Confirm the tdef (Iron_2 or Iron3) and peer group.
3. Check the H and L indices and appearance of the sample.
4. Review the calibration (CHEM CAL vial and date), pack age and IQC.
5. Compare other CHEM CAL analytes and, for LSC samples, transferrin, TIBC and ferritin of the same sample.

Unit traps

• Ratio about 5.6 (or 0.179) = ug/dL versus umol/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• PEG-enhanced immunoturbidimetry (goat anti-human transferrin), 340/596 nm, six-level LSP CAL multi-point curve - Atellica CH Trf
• Immunonephelometry on other platforms; IFCC C-PP reference method (optimised immunoturbidimetry or immunonephelometry)

EQA pitfalls the engine knows

• Shared LSP CAL for 13 specific-protein assays
• Fortnightly pack calibration
• Lipaemia or turbidity of reconstituted material
• Antigen excess (hook) (reads low)
• Antibody reactivity to the material's transferrin
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check units (g/L versus mg/dL, ratio 100) and transcription.
2. Compare with the Atellica peer group.
3. Review the LSP CAL opening date and lot and whether other protein assays moved.
4. Check the L index and vial appearance.
5. If TIBC is calculated from transferrin, check the factor and the TIBC result together.

Unit traps

• Ratio 100 = mg/dL versus g/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Direct two-step colourimetric TIBC (Chromazurol B and ferric chloride at acid pH, then neutral bicarbonate buffer so transferrin takes up iron; fall in absorban...
• Calculated TIBC: UIBC plus iron, or transferrin x about 25
• Saturation and precipitation (for example magnesium carbonate) methods

EQA pitfalls the engine knows

• Direct versus calculated TIBC
• Seven-day reagent and seven-day pack calibration
• Shared SPCL CHEM CAL (lactate, TIBC, amylase)
• Lipaemia raises TIBC (reads high)
• Design offset against other methods
• Measuring interval without automatic dilution
• Chimney pack and CO2 uptake
• Formula error in a calculated TIBC
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Establish whether TIBC is measured (Atellica TIBC) or calculated; check the registered method.
2. Check pack age (7 days) and weekly calibration history.
3. Check SPCL CHEM CAL reconstitution date and whether lactate or amylase also moved.
4. Check the L index of the sample.
5. Check units (umol/L versus ug/dL) and transcription.

Unit traps

• Ratio about 5.6 = ug/dL versus umol/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (acridinium-labelled goat polyclonal detection antibody, mouse monoclonal capture antibody covalently coupled to...
• Other sandwich immunoassays (chemiluminescence, electrochemiluminescence) and latex immunoturbidimetry on other platforms, calibrated to one of the WHO ferritin...

EQA pitfalls the engine knows

• Method-group target: ferritin assays are not harmonised
• CAL C is shared with vitamin B12
• New reagent lot and its master curve (Fer)
• Two-point recalibration or calibrator handling (Fer, CAL C)
• Chemiluminescent signal or wash fault in a sandwich assay
• Plasma-serum difference and haemolysis
• Turbid or lipaemic reconstituted material
• High ferritin samples: onboard dilution and hook
• Low ferritin samples and imprecision
• Test definition update left operator-defined parameters unchanged
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica ferritin group rather than the all-method mean (ferritin methods are not harmonised).
2. Check vitamin B12 on the same analyser: a joint shift points to the shared CAL C calibration.
3. Review calibration and lot dates (lot 50 days, pack 28 days) and IQC across them.
4. For a high sample, check dilution factor and recovery; for a low sample, use the APS absolute limit.
5. Inspect the vial for haemolysis or turbidity; check units (ug/L and ng/mL are equal; pmol/L = ug/L x 2.2).

Unit traps

• ug/L and ng/mL are numerically identical.
• pmol/L = ug/L x 2.2 (Atellica alternate unit); a ratio near 2.2 suggests pmol/L entered as ug/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Lactate oxidase with peroxidase colour (4-aminoantipyrine), ascorbate oxidase in R1, two-point reading at 571/596 nm - Atellica CH Lac_3
• Earlier Atellica CH Lac and Lac_2 (lactate oxidase)
• Amperometric lactate oxidase sensors on blood gas analysers; lactate dehydrogenase and NAD methods

EQA pitfalls the engine knows

• Pack calibration interval differs between Siemens documents
• Shared SPCL CHEM CAL (lactate, TIBC, amylase)
• Lac_2 and Lac_3 in different peer groups
• Internal-standard traceability and method-principle differences
• Icterus, haemolysis and etamsylate lower Lac_3 (reads low)
• Peroxidase (Trinder) end-point interference from reductants (reads low)
• Change in lactate after reconstitution
• Between-instrument scatter at low lactate
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check units (mmol/L versus mg/dL, ratio about 9.0) and transcription.
2. Confirm the tdef (Lac_3, Lac_2 or Lac) and the peer group.
3. Check the configured pack calibration interval and calibration history; check SPCL CHEM CAL age and whether TIBC or amylase moved.
4. Check the time from reconstitution to analysis.
5. Check H and I indices.

Unit traps

• Ratio about 9.0 (or 0.111) = mg/dL versus mmol/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Freezing point depression osmometry
• Vapour pressure osmometry (does not register volatile solutes such as ethanol and methanol)

EQA pitfalls the engine knows

• Evaporation or reconstitution volume error
• Vapour pressure versus freezing point with volatile solutes
• Osmometer calibration standards
• Freezing artefacts and sample handling
• Calculated osmolarity reported instead of measured osmolality
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check transcription and units (mmol/kg; mOsm/kg is numerically equal).
2. Confirm the osmometer principle (freezing point or vapour pressure) and peer group.
3. Check the osmometer calibration and standard ages.
4. Compare with the other analytes of the vial for a shared percentage shift.

Unit traps

• mmol/kg and mOsm/kg are numerically equal; mmol/L (osmolarity) is a different quantity.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic: proteinase K digestion, fructosaminase oxidation of the ketoamine, peroxidase colour (4-aminoantipyrine and TOOS) at 596/694 nm - Atellica CH Fruc
• Nitroblue tetrazolium (NBT) reduction on other platforms

EQA pitfalls the engine knows

• Enzymatic versus NBT method groups
• Correlation term built into a test definition version
• Fruc CAL reconstitution (1.0 mL, human serum albumin matrix)
• Bilirubin and ascorbate act in opposite directions
• Peroxidase (Trinder) end-point interference from reductants (reads low)
• Protein composition of the material
• Low-level imprecision and range without auto-dilution
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Confirm the method code (enzymatic Fruc) and compare with that peer group.
2. Check the Fruc test definition version and any version change near the survey.
3. Review Fruc CAL reconstitution and calibration dates.
4. Check bilirubin target and I index; ask about ascorbate or other additives.
5. Check transcription; fructosamine is reported only in umol/L.

Unit traps

• Reported only in umol/L; a result in mmol/L would differ by 1000.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Colourimetric chromoionophore end point in strongly alkaline medium, 505/694 nm - Atellica CH LITH_2
• Earlier Atellica CH Li (colourimetric)
• Ion-selective electrodes; flame emission, atomic absorption and ICP methods (reference procedures)

EQA pitfalls the engine knows

• Lithium heparin contamination (reads high)
• Reagent packaging change from lot 130031
• Li and LITH_2 in different peer groups
• Shared CHEM CAL calibrator vial or lot (lyophilised, 48 hours after reconstitution)
• Nine-day pack calibration
• Haemolysis and icterus at low lithium
• Colourimetric versus ISE and reference methods on processed material
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check transcription and units (mmol/L; mEq/L is numerically equal).
2. Confirm tdef (LITH_2 or Li) and reagent lot relative to lot 130031.
3. Review CHEM CAL and pack calibration history (9 days) and IQC.
4. Check aliquoting containers for lithium heparin.
5. Check H and I indices.

Unit traps

• mEq/L and mmol/L are numerically equal for lithium; a ratio of about 6.94 suggests mg/L versus mmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay: acridinium-labelled cortisol competes with sample cortisol for rabbit polyclonal anti-cortisol held by anti-rabbit IgG...
• LC-MS/MS and ID-MS reference measurement procedures

EQA pitfalls the engine knows

• Cross-reacting steroids in EQA material (reads high)
• Method-group target and commutability of the EQA material
• Frequent recalibration and short reagent life
• New reagent lot and its master curve (Cor)
• Chemiluminescent signal or wash fault in a competitive assay
• Dilution and range
• Imprecision and urine cortisol
• Sample age
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica immunoassay group and note the LC-MS/MS group; a sample-specific immunoassay excess suggests cross-reacting steroids.
2. Check the cortisol calibration date (pack 10 days, lot 19 days) and CAL E reconstitution date (14 days at 2-8 C).
3. Review reagent lot changes and IQC; check progesterone on the same analyser for a shared CAL E effect.
4. For urine cortisol, confirm direct or extracted procedure and the collection-volume calculation.
5. Check units: ug/dL x 27.59 = nmol/L.

Unit traps

• Ratio about 27.6 = ug/dL versus nmol/L; urine results need the collection volume (ug/24 h x 2.76 = nmol/24 h).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Third-generation two-site sandwich chemiluminescent immunoassay (acridinium ester label; FITC-labelled capture antibody on anti-fluorescein paramagnetic particl...
• Other sandwich immunoassays (electrochemiluminescence, chemiluminescent microparticle) on other platforms, each calibrated to a WHO international standard

EQA pitfalls the engine knows

• Method-group target and commutability of the EQA material
• TSH3-UL and TSH3ULII are different assays
• New reagent lot and its master curve (TSH3ULII or TSH3-UL)
• Two-point recalibration or calibrator handling (TSH3ULII, CAL 3A (kit calibrator))
• Chemiluminescent signal or wash fault in a sandwich assay
• Measuring range, onboard dilution and hook
• Low-TSH samples and the lower limit
• Heterophile or human anti-animal antibodies
• Test definition update left operator-defined parameters unchanged

Investigation checks, in order

1. Confirm the method code (TSH3ULII or TSH3-UL) in myQAP matches the assay each analyser runs, and compare with that Atellica peer group, not the all-method mean.
2. Check for a reagent lot change or recalibration (lot and pack interval 63 days) near the analysis date; review IQC and, if needed, the TSH3ULII master curve material.
3. For a high sample, check above-range flags, dilution factor and Multi-Diluent 15 onboard age; for a low sample, compare the absolute difference with the APS absolute limit.
4. If several IM sandwich assays on one analyser are low (or competitive assays high) together, review trigger reagents, wash and luminometer records.
5. Check units (mIU/L and uIU/mL are numerically equal) and any correlation factor in CentraLink or the LIS.

Unit traps

• mIU/L and uIU/mL are numerically identical (factor 1); a ratio of 1000 suggests mU/L entered as uU/L or similar.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive direct (analogue-free labelled hormone) chemiluminescent immunoassay: acridinium-labelled T4 competes with sample free T4 for biotinylated rabbit an...
• Other one-step or two-step direct immunoassays on other platforms
• Equilibrium dialysis with ID-LC-MS/MS: IFCC C-STFT reference measurement procedure

EQA pitfalls the engine knows

• Method-group target and commutability of the EQA material
• Binding-protein behaviour of processed material
• CAL A is shared by FT4, FT3, T4 and T3
• Two-point recalibration or calibrator handling (FT4, CAL A)
• New reagent lot and its master curve (FT4)
• Chemiluminescent signal or wash fault in a competitive assay
• Sample age at room temperature
• Range limits and low-end imprecision
• Biotin with an avidin-biotin capture link
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica FT4 group; FT4 is not harmonised between methods.
2. Check FT3, T4 and T3 on the same analyser for a shared CAL A or signal effect.
3. Review the FT4 calibration dates (pack 7 days, lot 21 days) and reagent lot changes against the analysis date; compare IQC before and after.
4. Check the time the reconstituted vial stood at room temperature (IFU limit 8 hours).
5. Check units: ng/dL x 12.9 = pmol/L; do not treat FT4 as dilution-linear.

Unit traps

• Ratio about 12.9 = ng/dL entered as pmol/L or the reverse (IFU factor 12.9; record factor 12.871 from the molar mass).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay: sample free T3 competes with a T3 analogue on paramagnetic particles for acridinium-labelled mouse monoclonal anti-T3...
• Other direct immunoassays on other platforms

EQA pitfalls the engine knows

• Method-group target and commutability of the EQA material
• CAL A is shared by FT3, FT4, T4 and T3
• Two-point recalibration or calibrator handling (FT3, CAL A)
• New reagent lot and its master curve (FT3)
• Chemiluminescent signal or wash fault in a competitive assay
• Low-end imprecision and range
• Anticoagulant and sample-age effects
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica FT3 group.
2. Check FT4, T4 and T3 on the same analyser for a shared CAL A or signal effect.
3. Review FT3 calibration (pack 28 days, lot 82 days) and lot changes against the analysis date.
4. For a low sample, weigh the 9% within-lab CV near 1.2 pmol/L.
5. Check units: pg/mL x 1.54 = pmol/L.

Unit traps

• Ratio about 1.54 = pg/mL entered as pmol/L or the reverse.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay after release from binding proteins by an alkaline ancillary reagent (T3/T4/VB12 ANC); T4 on particles competes with sa...
• ID-LC-MS/MS reference measurement procedures

EQA pitfalls the engine knows

• Method-group target and commutability of the EQA material
• CAL A shared with the other thyroid hormones
• Shared alkaline ancillary reagent (T3/T4/VB12 ANC)
• Two-point recalibration or calibrator handling (T4, CAL A)
• New reagent lot and its master curve (T4)
• Chemiluminescent signal or wash fault in a competitive assay
• Manual dilution above 30 ug/dL (387 nmol/L)
• Imprecision and anticoagulant effects
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica T4 group.
2. Check T3, FT4, FT3 (CAL A) and T3 and B12 (shared ancillary reagent) on the same analyser.
3. Review T4 calibration and lot dates (lot 33, pack 21 days) and the T3/T4/VB12 ANC onboard age.
4. For a high sample, check the manual dilution and entered factor.
5. Check units: ug/dL x 12.9 = nmol/L.

Unit traps

• Ratio about 12.9 = ug/dL versus nmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay after alkaline release (T3/T4/VB12 ANC); T3 analogue on particles competes with sample T3 for acridinium-labelled anti-...
• ID-LC-MS/MS reference measurement procedures

EQA pitfalls the engine knows

• Method-group target and commutability of the EQA material
• CAL A shared with the other thyroid hormones
• Shared alkaline ancillary reagent (T3/T4/VB12 ANC)
• Two-point recalibration or calibrator handling (T3, CAL A)
• New reagent lot and its master curve (T3)
• Chemiluminescent signal or wash fault in a competitive assay
• Imprecision, dilution and anticoagulants
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica T3 group (not a mixed ADVIA Centaur and Atellica group).
2. Check T4, FT4 and FT3 on the same analyser (CAL A, ancillary reagent, signal).
3. Review calibration and lot dates and the T3/T4/VB12 ANC onboard age.
4. Check units: ng/mL x 1.54 = nmol/L; ng/dL x 0.0154 = nmol/L.

Unit traps

• Ratio about 1.54 = ng/mL versus nmol/L; ratio about 65 = ng/dL versus nmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (acridinium-labelled goat polyclonal and particle-bound mouse monoclonal antibodies to the beta subunit), measuri...
• Other total or intact hCG immunoassays with different isoform recognition

EQA pitfalls the engine knows

• hCG isoform content of the EQA material
• Method-group target and commutability of the EQA material
• Constant offset at low hCG
• CAL B shared with prolactin, LH, FSH and digoxin
• Two-point recalibration or calibrator handling (ThCG, CAL B)
• New reagent lot and its master curve (ThCG)
• Chemiluminescent signal or wash fault in a sandwich assay
• High-dose hook effect (sandwich format)
• Onboard dilution of high samples
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica total hCG group; note the programme's measurand (total versus intact hCG).
2. For a low sample, weigh the constant offset and 13% CV near 2.6 IU/L and use the APS absolute limit.
3. Check dilution factor, flags and diluent age for a high sample (no hook up to 400,000 IU/L).
4. Check prolactin, LH and FSH on the same analyser (CAL B) and calibration and lot dates (lot 49, pack 35 days).
5. If sandwich assays on one analyser are low together, review trigger, wash and luminometer records.

Unit traps

• IU/L and mIU/mL are numerically identical; a ratio of 1000 suggests IU/mL confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Three-site sandwich chemiluminescent immunoassay, high sensitivity (two biotinylated monoclonal capture antibodies, mouse and sheep, on streptavidin paramagneti...
• Contemporary (not high-sensitivity) three-site sandwich chemiluminescent assay with streptavidin capture, reporting ng/mL: Atellica IM TnI-Ultra (TnI_UL), 2017...
• Many other cTnI immunoassays from several manufacturers, each with its own antibodies and calibrator

EQA pitfalls the engine knows

• Method-group target: cardiac troponin I assays are not harmonised
• TnIH and TnI-Ultra are different assays with 1000-fold different units
• Low-end imprecision near the 99th percentile
• New reagent lot and its master curve (TnIH)
• Two-point recalibration or calibrator handling (TnIH, kit CAL L and CAL H)
• Commutability and troponin forms in the EQA material
• Standing time after reconstitution
• Chemiluminescent signal or wash fault in a sandwich assay
• High-dose hook effect (sandwich format)
• Haemolysis and biotin
• Test definition update left operator-defined parameters unchanged
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Confirm the method code is Atellica TnIH (high sensitivity) and compare only with that peer group.
2. Check the unit path: TnIH reports ng/L; RCPAQAP's APS is stated in ug/L (a ratio of 1000 is a unit error).
3. For a low sample, judge the absolute difference in ng/L against low-level IQC SD and the APS absolute limit.
4. Check the calibration record: calibrators are usable for only 4 hours once opened; lot 47, pack 31 days.
5. If several IM sandwich assays on one analyser moved together, review trigger, wash and luminometer records.

Unit traps

• TnIH reports ng/L (= pg/mL); RCPAQAP's APS is stated in ug/L; ug/L x 1000 = ng/L.
• TnI-Ultra reports ng/mL (= ug/L): switching between TnI-Ultra and TnIH changes the number 1000-fold.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• No Atellica assay: the laboratory's Atellica IM menu (IFU extraction and assay chart) contains no troponin T assay
• Roche Elecsys Troponin T high sensitive: sandwich electrochemiluminescence immunoassay on cobas analysers, the single commercial cTnT source; 5th generation (ra...

EQA pitfalls the engine knows

• Result not produced on Atellica
• Roche calibration or lot adjustment moves the whole peer group
• Gen 5 versus Gen 6 assay
• Low-end imprecision near the 99th percentile
• Haemolysis lowers troponin T (reads low)
• Biotin sensitivity of older reagent (reads low)
• Commutability of lyophilised serum material
• Unit path (ng/L versus ug/L)

Investigation checks, in order

1. Identify the (non-Atellica) analyser, assay generation and lot that produced the troponin T result.
2. Compare with the Roche troponin T peer group; check for a lot or generation change around the survey.
3. For a low sample, judge the absolute difference in ng/L against the APS absolute limit.
4. Check the vial for haemolysis (hs-cTnT reads low with haemolysis).
5. Check units: ng/L versus ug/L (factor 1000).

Unit traps

• hs-cTnT reports ng/L; RCPAQAP's APS is stated in ug/L; ug/L x 1000 = ng/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Latex-enhanced immunoturbidimetry, rabbit anti-CRP, 571 nm - Atellica CH CRP_2 (4-304 mg/L), wrCRP (0.5-156 mg/L) and hsCRP (0.16-10 mg/L)
• Particle-enhanced immunoturbidimetry, mouse monoclonal, 1 in 4 pre-dilution, PROT2 CAL - Atellica CH h_CRP2 (0.16-9.5 mg/L)
• Particle-enhanced turbidimetric immunoassay in two cuvettes, goat anti-CRP, 340/694 nm - Atellica CH RCRP (0.5-250 mg/L)
• Immunonephelometry on other platforms; IFCC C-PP reference method (optimised immunoturbidimetry or immunonephelometry)

EQA pitfalls the engine knows

• Five Atellica CRP test definitions with different calibrators and ranges
• RCPAQAP CRP material is not whole human serum
• RCRP reagent lot steps at low CRP
• RCRP constant offset against Dimension RCRP (reads low)
• Low-CRP samples near the quantitation limit
• Automatic dilution of high samples on high-sensitivity tdefs
• Calibrator open-vial life and pack calibration timing
• Antigen excess (hook) (reads low)
• Rheumatoid factor, lipaemia and heterophile antibodies at low CRP
• Reconstitution volume error or evaporation of the EQA vial
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check units (mg/L versus mg/dL, ratio 10) and transcription.
2. Confirm the tdef (CRP_2, wrCRP, hsCRP, h_CRP2 or RCRP) and compare with that peer group.
3. For low samples, check reagent lot changes (RCRP) and the tdef's low-level imprecision.
4. Check the calibrator opening date against the calibration in force.
5. Check for Autorepeat flags on high results from high-sensitivity tdefs.

Unit traps

• Ratio 10 = mg/dL versus mg/L confusion (CRP_2, wrCRP and RCRP report either; hsCRP and h_CRP2 report mg/L only).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic HbA1c: on-board lysis (sodium nitrite), protease release of the beta-chain N-terminal fructosyl dipeptide, fructosyl peptide oxidase and peroxidase co...
• Cation-exchange HPLC, boronate affinity, capillary electrophoresis and immunoassay on other platforms
• IFCC reference method (HPLC with ESI-MS or CE after endoproteinase Glu-C digestion) and the NGSP/DCCT designated comparison method

EQA pitfalls the engine knows

• Lyophilised haemolysate EQA material outside the validated specimen
• Unit system confusion (mmol/mol versus %)
• A1c_E CAL levels 2 and 3 (lyophilised whole blood, 11 days) and light-sensitive level 1
• Pretreatment (lysing) reagent
• Haemoglobin F and variants in the material's donor blood (reads low)
• Low-level imprecision in IFCC units
• Small design bias against the NGSP reference (reads low)
• Sample mixing before aspiration
• Carry-over before software mitigation
• Correlation factor, unit conversion or LIS mapping in the result path

Investigation checks, in order

1. Check the unit (mmol/mol versus %) and transcription.
2. Confirm the tdef (A1c_E or A1c_H) and compare with the Atellica enzymatic peer group.
3. Compare performance in the lyophilised and whole-blood programmes.
4. Review A1c_E CAL reconstitution dates, pretreatment vial or bottle changes and calibration dates.
5. Check for HbF or variant notes on the sample.

Unit traps

• mmol/mol and % are related by an equation with an intercept; a value near 6 read as mmol/mol, or near 50 read as %, is a unit error.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (biotinylated sheep monoclonal capture on streptavidin particles, acridinium-labelled sheep monoclonal F(ab')2 de...
• Other NT-proBNP sandwich immunoassays (Roche Elecsys proBNP II electrochemiluminescence and others) with different antibodies and calibrators

EQA pitfalls the engine knows

• Method-group target: NT-proBNP assays are not standardised
• PBNP and PBNPII are different assays
• proBNP and glycosylated forms in the EQA material
• New reagent lot and its master curve (PBNPII or PBNP)
• Two-point recalibration or calibrator handling (PBNPII or PBNP, kit calibrators)
• Onboard dilution of high samples under-recovers (reads low)
• Low samples, the reporting limit and a constant offset
• Chemiluminescent signal or wash fault in a sandwich assay
• High-dose hook effect (sandwich format)
• Stability and material handling
• Biotin (first-generation PBNP only) (reads low)
• Test definition update left operator-defined parameters unchanged
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Confirm which Atellica NT-proBNP assay (PBNPII or PBNP) each analyser runs and that the RCPAQAP method code matches; compare with that Atellica group.
2. Check calibration and lot dates (PBNPII lot 91, pack 36 days; PBNP lot 90, pack 28 days) and calibrator age.
3. For a high sample, check dilution (onboard dilution recovers about 84-89%); for a low sample, check for a censored '< 35' entry.
4. If BNP from the same material is also low, consider BNP-specific degradation or material handling.
5. If several IM sandwich assays on one analyser moved together, review trigger, wash and luminometer records.

Unit traps

• pg/mL and ng/L are numerically identical.
• pmol/L = pg/mL x 0.118 (Atellica alternate unit); a ratio near 8.5 or 0.118 suggests pmol/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (acridinium-labelled mouse monoclonal F(ab')2 to the BNP ring structure, biotinylated mouse monoclonal to the C-t...
• Other BNP immunoassays (Abbott, Beckman and others) with different antibodies and calibrators

EQA pitfalls the engine knows

• Method-group target: BNP assays are not standardised
• Matrix: EDTA plasma assay, serum-based EQA material
• BNP loss on standing (proteolysis) (reads low)
• New reagent lot and its master curve (BNP)
• Two-point recalibration or calibrator handling (BNP, BNP CAL)
• Onboard dilution of high samples (reads low)
• Low-end imprecision
• Chemiluminescent signal or wash fault in a sandwich assay
• High-dose hook effect (sandwich format)
• Haemolysis and biotin limits are low
• Test definition update left operator-defined parameters unchanged
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica BNP group; BNP methods are not interchangeable.
2. Check time and temperature between reconstitution and analysis and the use of plastic: BNP is labile and loss reads low.
3. Check calibration and lot dates (lot 61, pack 42 days) and BNP CAL age (5 days at 2-8 C).
4. For a high sample, check the dilution factor and Multi-Diluent 15 onboard age (7 days).
5. If several IM sandwich assays on one analyser moved together, review trigger, wash and luminometer records.

Unit traps

• pg/mL and ng/L are numerically identical.
• pmol/L = pg/mL x 0.289 (Atellica alternate unit); a ratio near 3.46 or 0.289 suggests pmol/L confusion.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent binding assay: after release with DTT/Releasing Agent, sample B12 competes with acridinium-labelled B12 for purified hog intrinsic...
• Other intrinsic-factor competitive binding assays on other platforms

EQA pitfalls the engine knows

• DTT/Releasing Agent preparation and age
• Light exposure and standing time
• CAL C shared with ferritin
• Two-point recalibration or calibrator handling (VB12, CAL C)
• New reagent lot and its master curve (VB12)
• Chemiluminescent signal or wash fault in a competitive assay
• Method-group target and commutability of the EQA material
• Low-end imprecision, dilution and haemolysis
• Intrinsic factor antibodies (patient samples) (reads high)

Investigation checks, in order

1. Compare with the Siemens Atellica B12 group.
2. Check the DTT/Releasing Agent preparation date, lot and onboard time (108 hours) for the run.
3. Check ferritin on the same analyser (CAL C) and B12 calibration and lot dates (pack 18, lot 30 days).
4. Check light protection and standing time of the reconstituted vial; inspect for haemolysis.
5. Check units: pg/mL x 0.738 = pmol/L.

Unit traps

• Ratio about 1.36 (or 0.738) = pg/mL versus pmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent binding assay: after DTT/NaOH release, sample folate competes with acridinium-labelled folate for biotinylated folate binding prote...
• Microbiological assay and LC-MS/MS for folate forms

EQA pitfalls the engine knows

• Biotin with avidin-biotin capture (reads high)
• Haemolysis raises serum folate (reads high)
• Light and standing time (reads low)
• DTT/Releasing Agent preparation
• Two-point recalibration or calibrator handling (FolSerum, Fol kit calibrator)
• New reagent lot and its master curve (Folate)
• Chemiluminescent signal or wash fault in a competitive assay
• Method-group target and commutability of the EQA material
• Test definition update left operator-defined parameters unchanged

Investigation checks, in order

1. Compare with the Siemens Atellica folate group.
2. Inspect for haemolysis and consider biotin for a single high sample.
3. Check light protection and standing time.
4. Check the DTT/REL preparation date, calibrator age (7 days at 2-8 C), calibration (pack 7, lot 14 days) and lot changes.
5. Confirm that the serum test definition (FolSerum) was used, not the whole-blood Fol definition.

Unit traps

• Ratio about 2.27 = ng/mL versus nmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay after release from vitamin D binding protein; acridinium-labelled monoclonal anti-25(OH)D and a fluorescein-labelled vi...
• ID-LC-MS/MS reference measurement procedures (Ghent, NIST, CDC) and routine LC-MS/MS

EQA pitfalls the engine knows

• VitD and VitDII are different assays
• 24,25-dihydroxyvitamin D3 cross-reactivity (VitDII) (reads high)
• 25(OH)D2-rich material (reads high)
• Haemolysis limit (reads low)
• Releasing reagent lot matched to the primary pack
• Two-point recalibration or calibrator handling (VitD or VitDII, kit calibrator)
• New reagent lot and its master curve (VitDII)
• Chemiluminescent signal or wash fault in a competitive assay
• Low-concentration imprecision (VitDII)

Investigation checks, in order

1. Confirm the assay (VitD or VitDII) and method code; compare with that Atellica group and with the LC-MS/MS group.
2. For a sample-specific high result, consider 24,25(OH)2D3 (VitDII) or 25(OH)D2 content.
3. Inspect the vial for haemolysis (limit 155 mg/dL haemoglobin).
4. Check calibration dates (VitDII weekly), calibrator reconstitution, reagent lot and ancillary lot pairing.
5. Check units: ng/mL x 2.5 = nmol/L.

Unit traps

• Ratio about 2.5 = ng/mL versus nmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (acridinium-labelled N-terminal antibody; biotinylated C-terminal antibody on streptavidin particles), direct RLU...
• Other second- and third-generation PTH immunoassays with different epitope specificity

EQA pitfalls the engine knows

• Method-group target and commutability of the EQA material
• PTH lability before analysis (reads low)
• Calibrator has no refrigerated stability
• New reagent lot and its master curve (PTH)
• Chemiluminescent signal or wash fault in a sandwich assay
• Biotin with streptavidin capture (sandwich)
• High-dose hook effect (sandwich format)
• Specimen type and low-end imprecision
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica PTH group; between-method differences of 20 to 50% are expected.
2. Check the time from reconstitution to analysis (PTH is labile; serum 8 hours).
3. Check calibrator storage (no 2-8 C storage once reconstituted) and calibration and lot dates (lot 91, pack 60 days); run PTH MCM if a calibration is suspected.
4. If sandwich assays on one analyser are low together, review trigger, wash and luminometer records.
5. Check units: pg/mL x 0.106 = pmol/L.

Unit traps

• Ratio about 9.4 (or 0.106) = pg/mL versus pmol/L; ng/L equals pg/mL.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay, equimolar for free PSA and PSA-ACT: Atellica IM PSA (goat polyclonal acridinium label, mouse monoclonal covalent...
• Other total PSA immunoassays calibrated to WHO 96/670 or to the manufacturer's traditional (Hybritech) calibration, about 25% higher

EQA pitfalls the engine knows

• Method-group differences despite WHO calibration
• PSA and tPSAII are different assays with different WHO standards
• Seminal-fluid PSA in control or EQA material
• High-dose hook effect is reached earlier on Atellica PSA (reads low)
• New reagent lot and its master curve (PSA or tPSAII)
• Two-point recalibration or calibrator handling (PSA or tPSAII, CAL Q or tPSAII kit calibrators)
• Low PSA samples (post-prostatectomy range)
• Chemiluminescent signal or wash fault in a sandwich assay
• Onboard dilution
• Test definition update left operator-defined parameters unchanged
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Confirm which Atellica total PSA assay (PSA or tPSAII) is in use and compare with that Atellica group.
2. Check calibration and lot dates (PSA lot 29, pack 28 days; tPSAII lot 42, pack 14 days) and calibrator age.
3. For a very high sample reading implausibly low, rerun diluted (Atellica PSA hook from about 2,500 ug/L).
4. For a low sample, judge the absolute difference against the APS absolute limit.
5. If PSA and ferritin (water-wash assays) or several IM sandwich assays moved together on one analyser, review wash, probe wash and signal records.

Unit traps

• ug/L and ng/mL are numerically identical; a ratio of 1000 suggests ng/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Two-site sandwich chemiluminescent immunoassay (acridinium-labelled goat polyclonal and particle-bound mouse monoclonal anti-prolactin), direct RLU: Atellica IM...
• Other sandwich immunoassays; methods differ widely in macroprolactin detection

EQA pitfalls the engine knows

• Macroprolactin content of the EQA material
• Method-group target and commutability of the EQA material
• CAL B shared by prolactin, hCG, LH, FSH and digoxin
• Two-point recalibration or calibrator handling (PRL, CAL B)
• New reagent lot and its master curve (PRL)
• Chemiluminescent signal or wash fault in a sandwich assay
• High-dose hook effect (sandwich format)
• Unit conversion between ng/mL and mIU/L
• Heterophile or human anti-animal antibodies

Investigation checks, in order

1. Compare with the Siemens Atellica PRL group; note whether the sample shows unusual between-method spread (macroprolactin).
2. Check units and the 21.2 conversion factor (ng/mL versus mIU/L).
3. Check hCG, LH and FSH on the same analyser for a shared CAL B effect; review PRL calibration (pack 28, lot 34 days) and lot dates.
4. If sandwich assays on one analyser are low together, review trigger, wash and luminometer records.
5. For a very high sample, check dilution factor and flags (no hook up to 30,000 ng/mL).

Unit traps

• Ratio about 21.2 = ng/mL versus mIU/L (WHO 3rd IRP 84/500).
• mIU/L and uIU/mL are numerically identical.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay after steroid release: acridinium hapten competes with sample testosterone for biotinylated sheep monoclonal antibody c...
• ID-LC-MS/MS (CDC HoSt reference measurement procedure)

EQA pitfalls the engine knows

• Biotin in a competitive streptavidin-biotin format (reads high)
• Immunoassay versus LC-MS/MS at low (female-range) concentrations
• Releasing agent lot matched to the primary reagent
• Two-point recalibration or calibrator handling (TSTII, TSTII kit calibrator)
• New reagent lot and its master curve (TSTII)
• Chemiluminescent signal or wash fault in a competitive assay
• Cross-reacting androgens and bilirubin
• Low-end imprecision and dilution

Investigation checks, in order

1. Compare with the Siemens Atellica group and the LC-MS/MS group; separate male-range from female-range samples.
2. For a female-range sample, compare the absolute deviation with the APS absolute limit.
3. Check calibration (pack 69, lot 91 days), calibrator reconstitution date (14 days at 2-8 C) and the Releasing Agent lot pairing.
4. If a single sample is high, consider biotin or exogenous androgens in a donor-derived material and the sample's bilirubin.
5. Check units: ng/dL x 0.0347 = nmol/L; ng/mL x 3.47 = nmol/L.

Unit traps

• Ratio about 28.8 = ng/dL versus nmol/L; ratio about 3.47 = ng/mL versus nmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Competitive chemiluminescent immunoassay (three-step): after release by an ancillary reagent, acridinium-labelled sheep monoclonal anti-oestradiol binds sample...
• LC-MS/MS and ID-GC/MS reference procedures

EQA pitfalls the engine knows

• Low-concentration samples (male, postmenopausal, child range)
• Method-group target and commutability of the EQA material
• CAL 30 short stability
• Two-point recalibration or calibrator handling (eE2, CAL 30)
• New reagent lot and its master curve (eE2)
• Chemiluminescent signal or wash fault in a competitive assay
• Ancillary (releasing) reagent
• Dilution of high samples
• Fulvestrant and steroid drugs (reads high)

Investigation checks, in order

1. Separate low-concentration samples: judge them on the APS absolute limit and against the Atellica group.
2. Check CAL 30 reconstitution date (14 days) and calibration dates (pack 21, lot 22 days); review lot changes and IQC.
3. Check the ancillary pack onboard age and that serum, not plasma, was the specimen type set.
4. If competitive assays on one analyser are high together, review trigger, wash and luminometer records.
5. Check units: pg/mL x 3.67 = pmol/L.

Unit traps

• Ratio about 3.67 = pg/mL versus pmol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Latex-enhanced turbidimetric inhibition immunoassay (competitive), rate at 694 nm, sample pre-diluted 1 in 5: Atellica CH Dgn
• Chemiluminescent competitive immunoassay: an Atellica IM digoxin assay also exists (IFU part 10995341 Rev.
• Other immunoassays (CMIA, ECLIA, EMIT, CEDIA, KIMS, FPIA historically) on other platforms
• ID-LC-MS/MS reference measurement procedures (DGKC, INSTAND)

EQA pitfalls the engine knows

• Method-group offset at low and therapeutic concentrations
• Shared calibrator TDM CAL: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Low-concentration imprecision near the APS absolute limit
• Glycoside metabolites or preservatives in EQA material
• Haemolysis, icterus or turbidity of the reconstituted EQA material
• Reagent carry-over before the software 1.29 mitigation
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Confirm which analyser and assay reported digoxin (Atellica CH Dgn or Atellica IM) and that the RCPAQAP method registration matches.
2. Compare with the Atellica CH (or IM) peer group; regress the cycle on targets to separate an intercept (method offset, low-level effects) from a slope (calibration).
3. Review TDM CAL: vial reconstitution date (7 day life, light-protected), calibrator lot, last lot and pack calibration dates.
4. Check reagent lot and pack onboard age, and IQC at the lowest level against the 0.2 ug/L APS limit.
5. Check the analyser HIL indices for the EQA sample and the software version (digoxin carry-over mitigation in 1.29).

Unit traps

• ng/mL and ug/L are numerically equal.
• Ratio about 1.28 (or 0.78): nmol/L versus ug/L (M = 780.94 g/mol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic colourimetric: aryl acylamidase hydrolysis to p-aminophenol, coupling with 8-hydroxyquinoline-5-sulfonic acid (Mn2+), 596/805 nm: Atellica CH Acet
• Immunoassays (EMIT, DRI and others) on other platforms
• HPLC and LC-MS/MS (a Roche LC-MS/MS RMP is JCTLM-listed)

EQA pitfalls the engine knows

• Enzymatic versus immunoassay method groups
• Bilirubin or lipaemia in the EQA sample
• Shared calibrator TOX CAL: vial age, handling or calibrator lot
• Daily C0 (zero) adjustment and reagent onboard life
• Reagent lot change or pack near the end of onboard stability
• N-acetylcysteine, rifampicin or metamizole in the sample
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Check the same sample's bilirubin (total and conjugated) and triglyceride targets and the analyser HIL indices: bilirubin raises and lipaemia lowers Acet.
2. Compare with the enzymatic (Siemens) peer group, not immunoassay or chromatography groups.
3. Review the C0 adjustment before the run, TOX CAL vial age (3 days open) and lot, reagent onboard age (14 days).
4. If salicylate also deviated in the same direction on the same run, suspect TOX CAL handling first.
5. Check dilution flags for high (overdose-range) samples.

Unit traps

• Ratio 10: mg/dL (Atellica default) versus mg/L.
• Ratio about 6.6 (or 0.151): umol/L versus mg/L (M = 151.16 g/mol).
• Ratio about 66: mg/dL versus umol/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Particle-enhanced turbidimetric inhibition immunoassay (PETINIA), 545/694 nm rate, 1 in 5 pre-dilution: Atellica CH Vanc
• Other immunoassays (CMIA, KIMS, EMIT, CEDIA, historical FPIA)
• LC-MS/MS (a Roche/Univ. Munich RMP is JCTLM-listed)

EQA pitfalls the engine knows

• Manufacturer and method-group bias and drift
• Crystalline degradation product (CDP-1) in aged or warm material (reads high)
• Shared calibrator DRUG CAL II: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Protein interference in turbidimetric inhibition assays (IgM, light chains, rheumatoid factor)
• Low-range limit and trough-range precision
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Haemolysis, icterus or turbidity of the reconstituted EQA material (reads high)
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Compare with the Atellica CH or Siemens PETINIA peer group; look at that group's bias trend (Siemens groups have shown drift in EQA).
2. Check DRUG CAL II: vial opening date (30 days), calibrator lot, last lot (30 day) and pack (7 day) calibrations; look at carbamazepine, gentamicin and valproate on the same run for a shared shift.
3. Check reagent lot and pack onboard age and IQC before and after changes.
4. Check standing time and temperature of the reconstituted EQA sample (CDP-1 formation).
5. Check below-range flags for trough-level samples.

Unit traps

• ug/mL and mg/L are numerically equal.
• Ratio about 0.69 (or 1.45): umol/L versus mg/L (M = 1449.3 g/mol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• PETINIA (competitive turbidimetric inhibition), 545/694 nm rate, 1 in 5 pre-dilution: Atellica CH Gent
• Other immunoassays (CMIA, KIMS, EMIT, CEDIA, historical FPIA)
• LC-MS/MS (a JCTLM-listed RMP, C13RMP1)

EQA pitfalls the engine knows

• Low-level imprecision against a tight absolute APS
• Shared calibrator DRUG CAL II: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Related aminoglycosides and component mix in the material
• Method group against the Dimension system
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Haemolysis, icterus or turbidity of the reconstituted EQA material (reads high)
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. For trough-level targets, compare the deviation with low-level IQC SD before treating it as bias.
2. Check DRUG CAL II vial age, lot and calibrations; compare vancomycin, carbamazepine and valproate on the same run.
3. Check reagent lot, pack onboard age and IQC around changes.
4. Read the survey sample description for other aminoglycosides.
5. Confirm the submitted value and unit against the analyser record (CentraLink/LIS) and the RCPAQAP unit; test the result-to-target ratio against the unit factors.

Unit traps

• ug/mL and mg/L are numerically equal.
• Ratio about 2.16: umol/L (IFU convention) versus mg/L.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• PETINIA (competitive), 545/694 nm rate, sample pre-diluted 1 in 25: Atellica CH VPA
• Other immunoassays (CMIA, EMIT, CEDIA, KIMS)
• GC and LC-MS/MS

EQA pitfalls the engine knows

• Bovine versus human serum matrix (commutability)
• Method group against the Dimension system
• Shared calibrator DRUG CAL II: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Small primary sample volume (1 in 25 pre-dilution)
• Metabolites in the material
• Spiked, lyophilised EQA material and commutability
• Haemolysis, icterus or turbidity of the reconstituted EQA material (reads high)
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Compare with the Atellica or Siemens PETINIA peer group; regress the cycle for slope and intercept.
2. Check DRUG CAL II vial age, lot and calibrations; compare vancomycin, gentamicin and carbamazepine on the same run.
3. Check reagent lot and pack onboard age and IQC around changes.
4. Check sample aspiration and pre-dilution flags for the run (1 in 25 pre-dilution from 10 uL).
5. Confirm the submitted value and unit against the analyser record (CentraLink/LIS) and the RCPAQAP unit; test the result-to-target ratio against the unit factors.

Unit traps

• ug/mL and mg/L are numerically equal.
• Ratio about 6.93 (or 0.144): umol/L versus mg/L (M = 144.21 g/mol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Potentiometric glass or polymer-membrane pH electrode against a reference electrode with liquid junction, measured at 37 degrees C in a blood gas analyser (benc...
• Primary pH standards traceable to the IFCC reference method for pH in blood (1986) and NIST SRM 186 I/II phosphate buffers
• Temperature-corrected pH (to an entered patient temperature) is a calculated result, not a separate method

EQA pitfalls the engine knows

• Gas exchange after the ampoule is opened (CO2 loss)
• Ampoule temperature before opening
• Aqueous bovine-albumin EQA material is not blood
• pH or reference electrode drift, junction or membrane deposits
• Temperature correction applied to the EQA result
• IQC material type differs from EQA material

Investigation checks, in order

1. Confirm the result was entered as measured at 37 degrees C (no temperature correction) and without transcription error; pH is unitless and needs no conversion.
2. Look at pCO2 (and pO2) of the same ampoule: pH high with pCO2 low points to CO2 loss after opening; all three shifted with a temperature entry points to temperature correction.
3. Ask about ampoule handling: equilibration temperature and time, shaking, time from opening to aspiration, transfer into a syringe or cup, repeat aspiration.
4. Compare with the same-analyser peer group, not the all-method target (aqueous bovine-albumin material).
5. Review pH sensor calibration, sensor or cartridge age, maintenance and IQC around the run; if a second blood gas analyser exists, compare.

Unit traps

• pH has no unit; an H+ result in nmol/L entered as pH (e.g. 40) or pH at a non-37 C temperature are the realistic entry errors.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Severinghaus-type pCO2 electrode (a pH electrode behind a CO2-permeable membrane with a bicarbonate electrolyte), 37 degrees C, blood gas analyser
• Other sensor designs (planar or optical) on cartridge-based analysers; principle per the analyser's documentation
• Accuracy established by tonometry with certified gas mixtures (IFCC 1988 tonometry reference method)

EQA pitfalls the engine knows

• CO2 loss after the ampoule is opened (reads low)
• Ampoule temperature before opening
• Barometric pressure or calibration gas error
• pCO2 electrode membrane or electrolyte problem
• kPa and mmHg confusion
• Temperature correction applied to the EQA result
• Aqueous bovine-albumin material and method groups

Investigation checks, in order

1. Check the unit (mmHg versus kPa; ratio 7.50) and that the result is the 37 degrees C value.
2. Look at pH of the same ampoule: pCO2 low with pH high suggests CO2 loss after opening; review ampoule handling (equilibration, time open, transfer, bubbles).
3. Look at pO2 of the same ampoules: pCO2 and pO2 off by a similar percentage on one analyser suggests barometer or calibration gas; check the barometer and gas lot history.
4. Compare with the same-analyser peer group.
5. Review pCO2 sensor calibration, membrane or cartridge age and IQC around the run.

Unit traps

• 1 kPa = 7.5006 mmHg (1 mmHg = 0.133322 kPa): a result/target ratio near 7.5 or 0.133 means kPa and mmHg were confused.
• Serum total CO2 (mmol/L, Atellica CO2_c) is not pCO2; a value near 25 entered for pCO2 (target near 40 mmHg) may be a field mix-up.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Amperometric Clark-type pO2 electrode (platinum cathode behind an O2-permeable membrane), 37 degrees C, blood gas analyser
• Optical (luminescence quenching) pO2 sensors on some cartridge-based analysers
• Accuracy established by tonometry with certified gas mixtures (IFCC 1988 tonometry reference method)

EQA pitfalls the engine knows

• Gas exchange with room air after opening
• Ampoule temperature before opening
• Low oxygen-capacity material and analyser carry-over
• Barometric pressure or calibration gas error
• pO2 sensor membrane, cathode or optical spot ageing
• kPa and mmHg confusion
• Temperature correction applied to the EQA result

Investigation checks, in order

1. Check the unit (mmHg versus kPa; ratio 7.50) and that the result is the 37 degrees C value.
2. Check whether the deviation points towards room air pO2 (about 150 mmHg): up at low targets, down at high targets, with pCO2 low from the same ampoule; review opening-to-aspiration time and transfer.
3. Review ampoule temperature equilibration (Ong 1983: about 5 mmHg for 5 degrees C in aqueous buffers).
4. Compare with the same-analyser peer group, especially at low pO2 levels (material effect).
5. Check the barometer, calibration gas and pO2 sensor history and IQC; pCO2 and pO2 off together on one analyser suggests barometer or gas.

Unit traps

• 1 kPa = 7.5006 mmHg (1 mmHg = 0.133322 kPa): a ratio near 7.5 or 0.133 means kPa and mmHg were confused.
• Oxygen saturation (%) is not pO2; a value near 95 to 100 entered against a lower pO2 target may be a field mix-up.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Direct (undiluted) calcium ion-selective electrode with a neutral-carrier membrane against a reference electrode, blood gas or electrolyte analyser; result repo...
• pH-adjusted ionised calcium (calculated to pH 7.4 from actual ionised calcium and pH)

EQA pitfalls the engine knows

• Heparin binding or dilution when the EQA fluid is transferred into a syringe or capillary (reads low)
• pH change of the material alters ionised calcium
• pH-adjusted (pH 7.4) value reported instead of actual ionised calcium
• Aqueous material, ionic strength convention and method groups
• Calcium ISE membrane ageing or interference
• Total calcium or wrong unit entered

Investigation checks, in order

1. Confirm the actual (not pH 7.4-adjusted) ionised calcium was reported, in mmol/L, and not total calcium.
2. Ask how the ampoule was presented to the analyser: any heparinised syringe or capillary lowers ionised calcium.
3. Look at pH and pCO2 of the same ampoule: ionised calcium low with pH high suggests CO2 loss after opening.
4. Compare with the same-analyser peer group (aqueous bovine-albumin material).
5. Review calcium sensor calibration slope, sensor or cartridge age, maintenance and IQC.

Unit traps

• mg/dL x 0.2495 = mmol/L (ratio about 4.0); mEq/L x 0.5 = mmol/L (ratio 2); a ratio near 2 can also mean total calcium was entered.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Automated serum index by multi-wavelength photometry of a diluted sample; lipaemia is read as turbidity at long wavelengths where lipoproteins scatter light
• Atellica CH: bichromatic L index at 658 and 694 nm (stand-alone HIL test or within a donor assay), reported as index levels 0 to 6
• Some analysers report a quantitative lipaemia value (in their own units) rather than an index level
• Visual assessment (poor agreement with the automated index)

EQA pitfalls the engine knows

• Index is analyser-specific (not standardised)
• Atellica L index is a semi-quantitative level, not a concentration
• Laboratory-changed index intervals or thresholds
• Supplemented lipid in EQA material behaves differently from native lipaemia
• Turbidity not caused by lipid (incomplete reconstitution, particles, paraprotein) (reads high)
• Reconstitution volume error

Investigation checks, in order

1. Confirm the reporting option (quantitative, numerical index or qualitative) matches what the Atellica L index produces, and that no conversion from index level to g/L was applied.
2. Compare with the Siemens Atellica peer group only; between-manufacturer differences are expected.
3. Check the H, I and L test definitions and the Audit Trail Log for changed index intervals or thresholds.
4. Inspect the vial for undissolved material and review reconstitution (volume, pipette, standing, mixing).
5. Compare with H and I from the same vial: a shared percentage shift points to reconstitution; L alone points to turbidity or the lipaemia channel.

Unit traps

• g/L versus mg/dL: ratio 100 (1 g/L = 100 mg/dL).
• An Atellica index level (0 to 6) entered in the quantitative field gives values far below any g/L target in the upper levels and cannot be converted with a fact...

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Reversed-phase HPLC with post-column alkaline ferricyanide derivatisation (thiochrome) and fluorescence detection of TDP in deproteinised whole blood
• LC-MS/MS of TDP in whole blood with a stable-isotope internal standard
• HPLC of total thiamine (after dephosphorylation) rather than TDP: a different measurand

EQA pitfalls the engine knows

• Light exposure of the material (reads low)
• Standing at room temperature after reconstitution (reads low)
• Measurand and method differences (TDP versus total thiamine; HPLC versus LC-MS/MS)
• Whole-blood material handled as plasma or incompletely mixed
• Calibration without a reference system
• Units and haemoglobin normalisation

Investigation checks, in order

1. Confirm measurand (TDP or total thiamine), method code and unit (nmol/L, ug/L, per g Hb).
2. Compare with the same-measurand, same-principle peer group (HPLC or LC-MS/MS).
3. Review handling: complete reconstitution and mixing of the lyophilised whole blood, light protection, time before deproteinisation, no centrifugation of the material.
4. Check calibrator and kit lot changes and IQC around the analysis date.
5. If only one vial is low, suspect handling of that vial (light, standing time) before calibration.

Unit traps

• ug/L x 2.3568 = nmol/L (ratio about 2.36 or 0.42).
• ng/g Hb results need the haemoglobin of the material and cannot be converted by a factor.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• PETINIA (competitive), 545/694 nm rate, 1 in 5 pre-dilution: Atellica CH Carb; measures carbamazepine-10,11-epoxide at more than 90%
• Other immunoassays (EMIT, CEDIA, CMIA) with low or variable epoxide cross-reactivity
• HPLC and LC-MS/MS measuring parent and epoxide separately (JCTLM-listed RMPs)

EQA pitfalls the engine knows

• Carbamazepine-10,11-epoxide in the material (metabolite cross-reactivity)
• Matrix of the EQA material (animal or bovine serum)
• Shared calibrator DRUG CAL II: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Method group against the Dimension system and serum versus plasma
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Haemolysis, icterus or turbidity of the reconstituted EQA material (reads high)
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Read the sample description for carbamazepine-10,11-epoxide and compare PETINIA, other immunoassay and chromatography groups on that sample.
2. Compare with the Atellica or Siemens PETINIA peer group.
3. Check DRUG CAL II vial age, lot and calibrations; compare vancomycin, gentamicin and valproate on the same run.
4. Check reagent lot and pack onboard age and IQC around changes; check the L index of the sample.
5. Confirm the submitted value and unit against the analyser record (CentraLink/LIS) and the RCPAQAP unit; test the result-to-target ratio against the unit factors.

Unit traps

• ug/mL and mg/L are numerically equal.
• Ratio about 4.23 (or 0.236): umol/L versus mg/L (M = 236.27 g/mol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• PETINIA (competitive), 545/694 nm rate, 1 in 5 pre-dilution, total phenytoin: Atellica CH Phny
• Other immunoassays (EMIT, CEDIA, CMIA, KIMS)
• HPLC, GC and LC-MS (NIST RMP C10RMP5 JCTLM-listed)
• Free phenytoin by ultrafiltration then immunoassay or LC-MS (not the RCPAQAP GSC measurand)

EQA pitfalls the engine knows

• No automatic dilution above 40 mg/L
• Shared calibrator DRUG CAL: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Method group against the Dimension system
• Protein binding and matrix (total versus free drug)
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Haemolysis, icterus or turbidity of the reconstituted EQA material (reads high)
• Unit or transcription error in the submitted result

Investigation checks, in order

1. For targets above 40 mg/L, check the manual dilution (DRUG CAL Level 1, 2-fold) and the factor applied.
2. Compare with the Atellica or Siemens PETINIA peer group.
3. Check DRUG CAL vial age (90 days open), lot and calibrations (lot 28 days, pack 7 days); compare theophylline (and phenobarbital) on the same run.
4. Check reagent lot and pack onboard age and IQC around changes.
5. Confirm the submitted value and unit against the analyser record (CentraLink/LIS) and the RCPAQAP unit; test the result-to-target ratio against the unit factors.

Unit traps

• ug/mL and mg/L are numerically equal.
• Ratio about 3.96 (or 0.252): umol/L versus mg/L (M = 252.27 g/mol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• Enzymatic UV: salicylate hydroxylase consumes NADH, fall in absorbance at 340/410 nm: Atellica CH Sal
• Trinder colour reaction with ferric ions (older, less specific) on some platforms
• Immunoassay and HPLC on other platforms

EQA pitfalls the engine knows

• Enzymatic versus Trinder method groups
• Haemolysis, icterus or turbidity of the reconstituted EQA material (reads low)
• Shared calibrator TOX CAL: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Sulfasalazine, sulfapyridine or temozolomide in the sample
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Check the sample's haemoglobin, bilirubin and lipaemia burden: all HIL effects on Sal are negative.
2. Compare with the enzymatic (Siemens) peer group, not Trinder or chromatography groups.
3. Review TOX CAL vial age (3 days open) and lot, last lot and 21 day pack calibration, reagent onboard age.
4. If paracetamol deviated in the same direction on the same run, suspect TOX CAL handling first.
5. Check whether chloride on the same sample was high: salicylate raises indirect-ISE chloride (cross-analyte clue, C3 chloride monograph).

Unit traps

• Ratio 10: mg/dL (Atellica default) versus mg/L.
• Ratio about 138 (or 0.0072): mmol/L versus mg/L (M = 138.12 g/mol).
• Salicylate may be reported as mmol/L in Australia; confirm the unit registered with RCPAQAP.

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.

Method families

• PETINIA (competitive), 545/694 nm rate, 1 in 5 pre-dilution: Atellica CH Theo
• Other immunoassays (EMIT, CEDIA, CMIA; historical FPIA)
• HPLC and ID-LC-MS reference methods (Ghent, INSTAND, DGKL)

EQA pitfalls the engine knows

• Caffeine and methylxanthines in the material
• Haemolysis, icterus or turbidity of the reconstituted EQA material
• Shared calibrator DRUG CAL: vial age, handling or calibrator lot
• Reagent lot change or pack near the end of onboard stability
• Method group against the Dimension system
• Shared CH sample pre-dilution (primary aspiration, diluent volume)
• Spiked, lyophilised EQA material and commutability
• Measuring interval limits and dilution
• Unit or transcription error in the submitted result

Investigation checks, in order

1. Check the L index and triglyceride target of the sample (lipaemia +9% to +10%).
2. Compare with the Atellica or Siemens PETINIA peer group; check for co-spiked caffeine.
3. Check DRUG CAL vial age (90 days open), lot and calibrations; compare phenytoin on the same run.
4. Check reagent lot and pack onboard age and IQC around changes.
5. Confirm the submitted value and unit against the analyser record (CentraLink/LIS) and the RCPAQAP unit; test the result-to-target ratio against the unit factors.

Unit traps

• ug/mL and mg/L are numerically equal.
• Ratio about 5.55 (or 0.180): umol/L versus mg/L (M = 180.16 g/mol).

Every fact in the record cites its source: the analyser's instructions for use, JCTLM, the EFLM biological variation database and the published literature. The full record ships with the application.