Standardising EQA submissions with automated defactoring of post-analytical correction factors

Post Analytical Correction Factors are routinely applied by clinical laboratories to align results from different instruments, methods, or because of the need to maintain consistent results in the face of lot-to-lot reagent variation, or if a new method introduces a bias compared to the technique from which the reference interval was derived. In any of these situations, some procedures must be followed to ensure that the factor used to align results from different methods, procedures, or reagents is valid and robust. In addition to the problems a laboratory faces in deriving and managing these factors, poorly described procedures exist for submitting results for External Quality Assurance (EQA) challenges. Should the participant submit the factored or unfactored result? Should the laboratory remain with its peer method group or report the result it would to a laboratory referrer? If the EQA material is commutable, the corrected bias should not be removed for the EQA purpose of laboratory assessment. However, it should be removed for metrological traceability between different instruments. If the EQA material is not commutable, the corrected bias should be removed for laboratory assessment. Non-commutable EQA material cannot be used to evaluate metrological traceability. The Royal College of Pathologists of Australasia Quality Assurance Programs (RCPAQAP) has a policy of treating the EQA sample as if it were a patient but manually defactoring the final submitted results to enable valid peer group comparisons. However, there is evidence that not all laboratories complied with this policy, making it challenging to identify and troubleshooting biases. A Factor/defector tool was developed to automatically defactor submitted results applied to survey results and allow laboratories to flag and reverse PACFs during data submission. This standardises EQA submissions across institutions, regardless of LIS or middleware configurations.