Limits...
A Progress Report of the IFCC Committee for Standardization of Thyroid Function Tests.

Thienpont LM, Van Uytfanghe K, Van Houcke S, Das B, Faix JD, MacKenzie F, Quinn FA, Rottmann M, Van den Bruel A, IFCC Committee for Standardization of Thyroid Function Tests (C-STF - Eur Thyroid J (2014)

Bottom Line: The IFCC Committee for Standardization of Thyroid Function Tests aims at equivalence of laboratory test results for free thyroxine (FT4) and thyrotropin (TSH).The impact of recalibration on the numerical results was particularly high for FT4.Because of the impact on the numerical values, the implementation needs careful preparation with the stakeholders.

View Article: PubMed Central - PubMed

Affiliation: Laboratory for Analytical Chemistry, Faculty of Pharmaceutical Sciences, Gent University, Gent, Bruges, Belgium.

ABSTRACT

Background: The IFCC Committee for Standardization of Thyroid Function Tests aims at equivalence of laboratory test results for free thyroxine (FT4) and thyrotropin (TSH).

Objectives: This report describes the phase III method comparison study with clinical samples representing a broad spectrum of thyroid disease. The objective was to expand the feasibility work and explore the impact of standardization/harmonization in the clinically relevant concentration range.

Methods: Two sets of serum samples (74 for FT4, 94 for TSH) were obtained in a clinical setting. Eight manufacturers participated in the study (with 13 FT4 and 14 TSH assays). Targets for FT4 were set by the international conventional reference measurement procedure of the IFCC; those for TSH were based on the all-procedure trimmed mean. The manufacturers recalibrated their assays against these targets.

Results: All FT4 assays were negatively biased in the mid- to high concentration range, with a maximum interassay discrepancy of approximately 30%. However, in the low range, the maximum deviation was approximately 90%. For TSH, interassay comparability was reasonable in the mid-concentration range, but worse in the pathophysiological ranges. Recalibration was able to eliminate the interassay differences, so that the remaining dispersion of the data was nearly entirely due to within-assay random error components. The impact of recalibration on the numerical results was particularly high for FT4.

Conclusions: Standardization and harmonization of FT4 and TSH measurements is feasible from a technical point of view. Because of the impact on the numerical values, the implementation needs careful preparation with the stakeholders.

No MeSH data available.


Related in: MedlinePlus

a-d Plots showing the %-difference before (FT4: a, TSH: b) and after recalibration (FT4: c, TSH: d). The most discrepant assays before recalibration are highlighted by special symbols (FT4: K, circles; M, triangles; TSH: I, circles; K, triangles); all other assays are indicated with the same symbol X. e, f Interassay CV (FT4: e, TSH: f) before (squares) and after (triangles) recalibration by IVD manufacturers.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4109515&req=5

Figure 1: a-d Plots showing the %-difference before (FT4: a, TSH: b) and after recalibration (FT4: c, TSH: d). The most discrepant assays before recalibration are highlighted by special symbols (FT4: K, circles; M, triangles; TSH: I, circles; K, triangles); all other assays are indicated with the same symbol X. e, f Interassay CV (FT4: e, TSH: f) before (squares) and after (triangles) recalibration by IVD manufacturers.

Mentions: Figure 1A and table 3 show that, compared to the cRMP targets, all FT4 assays were strongly negatively biased (beyond the arbitrary limit of −10%) for concentrations >27 pmol/l (median bias: −37%, range: −21 to −48), and between 9 and 27 pmol/l (median bias: −24%, range: −14 to −42). In contrast, they were negatively as well as positively biased in the range <9 pmol/l (median bias: −9%, range: −28 to 62). IVD manufacturers were able to eliminate the observed bias of their assays by recalibration to the cRMP target values (fig. 1c).


A Progress Report of the IFCC Committee for Standardization of Thyroid Function Tests.

Thienpont LM, Van Uytfanghe K, Van Houcke S, Das B, Faix JD, MacKenzie F, Quinn FA, Rottmann M, Van den Bruel A, IFCC Committee for Standardization of Thyroid Function Tests (C-STF - Eur Thyroid J (2014)

a-d Plots showing the %-difference before (FT4: a, TSH: b) and after recalibration (FT4: c, TSH: d). The most discrepant assays before recalibration are highlighted by special symbols (FT4: K, circles; M, triangles; TSH: I, circles; K, triangles); all other assays are indicated with the same symbol X. e, f Interassay CV (FT4: e, TSH: f) before (squares) and after (triangles) recalibration by IVD manufacturers.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4109515&req=5

Figure 1: a-d Plots showing the %-difference before (FT4: a, TSH: b) and after recalibration (FT4: c, TSH: d). The most discrepant assays before recalibration are highlighted by special symbols (FT4: K, circles; M, triangles; TSH: I, circles; K, triangles); all other assays are indicated with the same symbol X. e, f Interassay CV (FT4: e, TSH: f) before (squares) and after (triangles) recalibration by IVD manufacturers.
Mentions: Figure 1A and table 3 show that, compared to the cRMP targets, all FT4 assays were strongly negatively biased (beyond the arbitrary limit of −10%) for concentrations >27 pmol/l (median bias: −37%, range: −21 to −48), and between 9 and 27 pmol/l (median bias: −24%, range: −14 to −42). In contrast, they were negatively as well as positively biased in the range <9 pmol/l (median bias: −9%, range: −28 to 62). IVD manufacturers were able to eliminate the observed bias of their assays by recalibration to the cRMP target values (fig. 1c).

Bottom Line: The IFCC Committee for Standardization of Thyroid Function Tests aims at equivalence of laboratory test results for free thyroxine (FT4) and thyrotropin (TSH).The impact of recalibration on the numerical results was particularly high for FT4.Because of the impact on the numerical values, the implementation needs careful preparation with the stakeholders.

View Article: PubMed Central - PubMed

Affiliation: Laboratory for Analytical Chemistry, Faculty of Pharmaceutical Sciences, Gent University, Gent, Bruges, Belgium.

ABSTRACT

Background: The IFCC Committee for Standardization of Thyroid Function Tests aims at equivalence of laboratory test results for free thyroxine (FT4) and thyrotropin (TSH).

Objectives: This report describes the phase III method comparison study with clinical samples representing a broad spectrum of thyroid disease. The objective was to expand the feasibility work and explore the impact of standardization/harmonization in the clinically relevant concentration range.

Methods: Two sets of serum samples (74 for FT4, 94 for TSH) were obtained in a clinical setting. Eight manufacturers participated in the study (with 13 FT4 and 14 TSH assays). Targets for FT4 were set by the international conventional reference measurement procedure of the IFCC; those for TSH were based on the all-procedure trimmed mean. The manufacturers recalibrated their assays against these targets.

Results: All FT4 assays were negatively biased in the mid- to high concentration range, with a maximum interassay discrepancy of approximately 30%. However, in the low range, the maximum deviation was approximately 90%. For TSH, interassay comparability was reasonable in the mid-concentration range, but worse in the pathophysiological ranges. Recalibration was able to eliminate the interassay differences, so that the remaining dispersion of the data was nearly entirely due to within-assay random error components. The impact of recalibration on the numerical results was particularly high for FT4.

Conclusions: Standardization and harmonization of FT4 and TSH measurements is feasible from a technical point of view. Because of the impact on the numerical values, the implementation needs careful preparation with the stakeholders.

No MeSH data available.


Related in: MedlinePlus