A bottom-up approach in estimating the measurement uncertainty and other important considerations for quantitative analyses in drug testing for horses.

This research paper presents a ‘bottom-up’ method developed by the authors for estimating the measurement uncertainty, an important factor in quantitative analyses, particularly while testing for threshold substances in horse doping control. The article further discusses certain points that can potentially influence the outcome of these quantitative analyses.

Understanding Measurement Uncertainty in Quantitative Analyses

• The research emphasizes the difference between qualitative identification and quantitative determination, with the latter being more challenging due to the need for precise measurements. Certain substances in biological samples are only problematic if they exceed certain thresholds. The uncertainty surrounding this measurement, dubbed ‘measurement uncertainty’ (MU), is a crucial aspect of these quantitative determinations.
• This study refers to the requirements laid out by the International Standard ISO/IEC 17025, which superseded ISO/IEC Guide 25 and is stricter concerning MU. According to the 1999 and 2005 versions of the ISO/IEC 17025 standard, all quantitative determinations must have an estimated MU.
• Following these necessities, the researchers established a protocol in their lab in 2001, which further developed through their experiences, with a refined version getting implemented in 2004.

The ‘Bottom-Up’ Approach to Evaluating MU

• The authors describe their process for establishing MU, focusing on quantifying threshold substances in biological samples for horse doping control. This form of testing is a compliance test or testing to a specified limit, implying MU should be established at the threshold level only.
• The “Bottom-Up” method developed by the researchers aligns with guidelines provided in the EURACHEM/CITAC guide, “Quantifying Uncertainty in Analytical Measurement”. It requires the specification of the measurand (the quantity intended to be measured) and its relationship with input quantities.
• The second step involves identifying uncertainty contributions through a “cause and effect” diagram. The uncertainty component for each of them is then calculated and converted into a standard uncertainty. If the method bias is significant — determined through a recovery study — then a recovery or correction factor is applied.
• The combined standard uncertainty is derived from all standard uncertainties exceeding 30% of the largest one. An expanded uncertainty at a 99% one-tailed confidence level is then computed by multiplying the standard uncertainty with an appropriate coverage factor (k). If the measurement of the threshold substance’s concentration exceeds the threshold by the expanded uncertainty, the substance is deemed ‘positive’.

Other Important Considerations

• The study also highlights additional factors that can significantly affect quantitative analyses. However, the abstract doesn’t specify what these considerations are — that information is likely detailed in the full paper.