Arsenic content in fish and associated measurement uncertainty as a metrological parameter for Total Diet Studies

Total Diet Studies (TDS) are the most important tools to assure that chemical contaminants presence in foodstuffs remains within safety levels. Reliability of data provided by laboratories to assess the effectiveness of measurements aiming at reducing the risk of exposure to chemical hazards is of paramount importance. Evaluation of measurement uncertainty associated with a result is an important parameter in assessing the sources of analytical data variability. Two methods to estimate uncertainty in analytical measurement of arsenic in fish are discussed. Measurement uncertainty was estimated for the determination of arsenic content in fish samples by Inductively Coupled Mass Spectrometry (ICP-MS). The work addressed both approaches accepted by Eurolab and NIST: modeling (bottom up) and empirical (top down). The Ishikawa diagram was used to identify the most significant sources of uncertainty. The mathematical modeling techniques to assess uncertainty components based on a classical model that accounts for all recognized significant sources of errors was similar to the top down based on validation data from interlaboratory comparison studies. Both methods showed that analytical variability in arsenic content was less than 10%. This work was completed on behalf of Project GOODFISH PTDC/SAU-ESA/103825/2008