EQA SCHEME FOR HORMONE IMMUNOASSAY: EVALUATION OF LABORATORY PERFORMANCE THROUGH A SCORING SYSTEM

The External Quality Assessment (EQA) scheme Immunocheck for hormone immunoassay has been organised by our Institute since 1996 in co-operation with ProBioQual (Lyon, France); at present time more than 1000 laboratories are involved in the scheme which includes 18 analytes. Participants assay 18 control samples every year; statistics of the collected results are circulated sending back to the laboratories periodic and cumulative reports. To allow participants an easy evaluation of their own analytical performance, a new scoring system has been adopted. Each result is scored according to its deviation (from target value) expressed in SD unita (Z-score). In detail the scores are: 4(excellent) if Z <0.5, 3(good) if 0.53. The sum of scores for all samples assayed in a control cycle, normalised by the maximum achievable total score, describes the analytical performance of the laboratory. Z-score is computed as the ratio of percent deviation from target and "state-of-the-art" imprecision. Method mean is assumed as target; state-of-the-art imprecision is computed as mean imprecision of the methods most used in the survey. To take account that imprecision depends on the analyte concentration, 3 CVs are computed corresponding to 3 concentration range. The table, in the document attached, reports for each analyte the concentration ranges and the state-of-the-art CVs (within-method, between-laboratories imprecision) computed from the Immunocheck data-base (results collected during 2001 EQA cycle).