A "Step-Up" approach for harmonization.

According to the measurement paradigm, assays claiming the same measurand should give equivalent results within clinically relevant constraints. The key to accomplish this paradigm is standardization of assays to an SI-traceable reference measurement system. However, measurement results for complex analytes often are not expressed in SI-, but in arbitrary units, defined, for example, by an international conventional measurement standard. Traceability to such a standard mostly does not lead to among assay equivalence. To achieve this, the concept of harmonization has been proposed. We describe here the practical aspects involved with the "Step-Up" design. It essentially comprises a sequence of method comparisons with selected sets of commutable samples. The outcome of each phase allows to decide whether the step-up to the next phase can be set. The harmonization process itself uses a statistically valid location measure as surrogate reference measurement procedure. The design also foresees a protocol for sustaining the first-established harmonization status. For the design to be successful, it is essential that as many assays as possible are involved in the method comparison leading to the harmonization target, and that it can be shown that they sufficiently correlate to that target with consistency of performance over the covered measurement range.