Identification of determinants of the between-operator variation of the total protein S antigen assay: a collaborative study of the Dutch Working Group on Haemostasis Laboratory Diagnosis

Abstract

In a previous study, a between-operator variability (CVO BETWEEN ) of 9.6% and 15.0% was observed for total protein S antigen assays in 11 laboratories using a frozen or lyophilized reference plasma, respectively, and the need to standardize the use of lyophilized reference plasma was identified. The aim of the present study was to identify further determinants of this CVO BETWEEN in order to improve between-laboratory comparisons of test results for one method for protein S antigen assay. Two protocols were carried out: the first again involving local execution but using a joint standardized and detailed prescription of the technical performance in each laboratory; the second using a central session for all operators with the same prescription but with joint reagent and equipment. In the present study, improved handling of lyophilized reference plasma was included and resulted in comparable CVO BETWEEN of 10.9% and 9.6% for the use of frozen and lyophilized reference plasma for the local test performance. An improvement was found in the CVO BETWEEN in the central session compared with the standardized local performance, showing lower values for the central performance of 8.5 and 6.6% for frozen and lyophilized reference plasma, respectively. Further analysis of the difference between the local and central test performance identified the use of different curve fit options of data evaluation software as a significant source of this difference. Interestingly, the withinoperator variability in the central performance was around 2% lower (5.9 and 6.0% for frozen and lyophilized plasma, respectively) than that in the local performance (8.1 and 8.0% for frozen and lyophilized plasma, respectively). Although the reduction is not statistically significant, it suggests an effect of reduction of the workload and simplification of procedures for individual operators on the within-operator variability. In this study, in which 11 operators/laboratories participated, the lowest variability between operators and within laboratories was obtained in the central test performance, which is suggested to be the lowest attainable variability for the measurement of total protein S antigen. The practical factors involved in local performance that require attention to reach similar levels of variability are mainly liquid handling, curvefit procedures and simplicity of practical procedures.