BACKGROUND: There is a need for systematic evaluation of methods before their release to the market. We addressed this problem in novel homocysteine assays as part of an European Demonstration Project involving six centers in four countries. METHODS: Two immunological methods for measurement of plasma total homocysteine (P-tHcy), the fluorescence polarization immunoassay (FPIA) and the enzyme immunoassay (EIA), were compared with two comparison methods, HPLC and gas chromatography-mass spectrometry (GC-MS). All laboratories performed the following procedures: (a) familiarization; (b) determination of linearity and precision by analyzing five plasma samples with interrelated concentrations for 20 days; (c) correlation using patients' samples; and (d) assessment of long-term performance. RESULTS: Both immunological methods were linear for P-tHcy between 5 and 45 micromol/L. The intralaboratory imprecision (CV) was <5% for FPIA and <9% for EIA used with a sample processor. The bias was -2% to 3% for FPIA and 2-4% for EIA used with a sample processor. CONCLUSIONS: The immunological methods provide results with little bias compared with HPLC and GC-MS. The imprecision of the assays must be considered in the context of their intended use(s).
BACKGROUND: There is a need for systematic evaluation of methods before their release to the market. We addressed this problem in novel homocysteine assays as part of an European Demonstration Project involving six centers in four countries. METHODS: Two immunological methods for measurement of plasma total homocysteine (P-tHcy), the fluorescence polarization immunoassay (FPIA) and the enzyme immunoassay (EIA), were compared with two comparison methods, HPLC and gas chromatography-mass spectrometry (GC-MS). All laboratories performed the following procedures: (a) familiarization; (b) determination of linearity and precision by analyzing five plasma samples with interrelated concentrations for 20 days; (c) correlation using patients' samples; and (d) assessment of long-term performance. RESULTS: Both immunological methods were linear for P-tHcy between 5 and 45 micromol/L. The intralaboratory imprecision (CV) was <5% for FPIA and <9% for EIA used with a sample processor. The bias was -2% to 3% for FPIA and 2-4% for EIA used with a sample processor. CONCLUSIONS: The immunological methods provide results with little bias compared with HPLC and GC-MS. The imprecision of the assays must be considered in the context of their intended use(s).