“…The variability associated with the use of different ELISAs or matrices (serum or meat juice) has been also reported (Farzan et al., ; Mainar‐Jaime, Atashparvar, Chirino‐Trejo, & Blasco, ; Mejía, Casal, Mateu, & Martin, ; Vico & Mainar‐Jaime, ). Thus, the bias assumed when these ELISAs are used on individuals should be translated to the herd level as well, which probably make them unsuitable for proper herd risk characterization (Gradassi et al., ; Sørensen, Alban, Nielsen, & Dahl, ; Vico et al., ). The small sample size usually considered (between 36 and 100 pigs/year) for an infection whose presentation varies among batches and years, and the lack of representativeness of the on‐farm animal distribution, as pigs are usually sampled at slaughter, are factors that also may have contributed to increase that bias.…”