A non-parametric framework for evaluating total analytical error in in vitro diagnostic medical devices in transfusion medicine

Abstract

The performance assessment of quantitative measurements is predominantly based on evaluating Total Analytical Error (TAE). This evaluation encompasses several key objectives critical to ensuring accurate, reliable, and clinically relevant test results. Traditional parametric methods often fall short due to data normality assumptions in the performance assessment of in vitro diagnostic medical devices (IVD-MDs). This study presents a non-parametric approach to estimating and evaluating the TAE in transfusion medicine, aiming to enhance the reliability and patient safety of IVD-MDs. A protocol to estimate TAE over diverse data distributions is suggested, employing a robust statistical definition and comparative measurement procedures. Results from 200 samples indicate that non-parametric methods provided a more accurate reflection of TAE. The findings assert that non-parametric TAE estimation is vital for ensuring the 'fitness for purpose' of clinical tests in transfusion medicine, directly impacting post-transfusion outcomes and patient care. The study concludes that adopting non-parametric methods in transfusion services can significantly improve test accuracy, aligning with the highest laboratory practice standards.