The role of measurement uncertainty in the conformity assessment of the chemical composition of feeds

Official feed inspection has been directly related to the dispersion of the measures achieved by control laboratories, which can be due to several analytical techniques, calculation models and estimation approaches. The present study aims to demonstrate the importance of calculating the measurement uncertainty (MU) in the analysis of the chemical composition of several feedingstuffs, in order to avoid making wrong decisions due to incomplete measures. Samples of feedingstuffs (n = 690), collected from 321 establishments, were analyzed. Complete measurements were compared to the labeling information declared by the processor. Animal feeds and mineral mixtures were the most frequent non-compliant samples and, among them, crude protein (n = 94), calcium (n = 50) and phosphorus (n = 40) were the most frequent violations. Method performance was assessed in terms of in-house precision and recovery and by international proficiency testing. MU was calculated by using a bottom-up approach, combined with random effects inherited from the performance evaluation. A top-down approach based on the interlaboratory deviation was also carried out. The MU values calculated by the combined bottom-up approach were mostly lower than those calculated by interlaboratory experimental data. This latter approach overestimated the MU in most results, becoming inadequate. Titrimetric and spectrometric methods presented the higher and the lower MU among all techniques, respectively. A detailed study of the MU plays a key role in the conformity assessment of feedingstuffs, because the use of complete measures avoided 314 wrong decisions taking and prevented the disposal of approximately 1,000,007 tons of animal feed.