Using simulations to explore the potential effect of disease and inflammation on the frequency of violative flunixin residues in cattle.

Abstract

The US Food and Drug Administration (FDA) assigns a tolerance and withdrawal period when evaluating new drugs for use in food-producing species. Because withdrawal periods are determined from data generated in normal, healthy animals, questions have been raised regarding whether disease and inflammation can be a factor associated with some residue violations. We explored this question using flunixin liver concentrations as a model situation. Using data contained in the flunixin FOI Summary (NADA 101-479) and Monte Carlo simulation, we generated sets of residue depletion data. Our mathematical model was simple linear regression containing the terms alpha (the marker residue back-extrapolated to time zero, which equals ) and beta (the elimination rate constant which equals ). By modifying alpha and beta means and variances, we determined the smallest change in these parameters that would result in the presence of violative residues above the statistically determined expected frequency of 1%. The results of this in silico study indicated that the magnitude of change in alpha and beta needed to generate violative residues exceeds that likely to occur due to disease or inflammation when flunixin is used in accordance with the approved product label.