Feasibility for aggregation of commutable external quality assessment results to evaluate metrological traceability and agreement among results

Abstract

Abstract Objectives External quality assessment (EQA) with commutable samples is used for assessing agreement of results for patients’ samples. We investigated the feasibility to aggregate results from four different EQA schemes to determine the bias between different measurement procedures and a reference target value. Methods We aggregated EQA results for creatinine from programs that used commutable EQA material by calculating the relative difference between individual participant results and the reference target value for each sample. The means and standard errors of the means were calculated for the relative differences. Results were partitioned by methods, manufacturers and instrument platforms to evaluate the biases for the measurement procedures. Results Data aggregated for enzymatic methods had biases that varied from −8.2 to 3.8% among seven instrument platforms for creatinine at normal concentrations (61–85 μmol/L). EQA schemes differed in the evidence provided about the commutability of their samples, and in the amount of detail collected from participants regarding the measurement procedures which limited the ability to sub-divide aggregated data by instrument platforms and models. Conclusions EQA data could be aggregated from four different programs using different commutable samples to determine bias among different measurement procedures. Criteria for commutability for EQA samples as well as standardization of reporting the measurement methods, reagents, instrument platforms and models used by participants are needed to improve the ability to aggregate the results for optimal assessment of performance of measurement procedures. Aggregating data from a larger number of EQA schemes is feasible to assess trueness on a global scale.