Evaluation of Methods Used to Detect Antibiotic Residues in Milk ¹

Over nine replicate trials, five methods for detecting antibiotic residues were evaluated for sensitivity to various levels of penicillin, erythromycin and chloramphenicol in milk. The methods included (a) the Charm test, (b) Delvotest P, (c) disc assay with Bacillus subtilis on whey agar, (d) B. subtilis on Antibiotic Medium #1 and (e) Difco disc assay method using Bacillus stearothermophilus var. calidolactis on PMI agar. Specific to penicillin only, the Charm test detected the antibiotic 100% of the time down to and including 0.01 unit/ml, dropping to 56% detectability at 0.0025 unit/ml. Delvotest P detected penicillin 100% of the time at 0.025 unit/ml, but at varying degrees of sensitivity through 0.0025 unit/ml, and with evidence of showing false-positive results. Methods (c), (d) and (e) detected penicillin at 0.05 unit/ml 78, 89 and 100% of the time, respectively. Methods (d) and (e) were generally more sensitive to erythromycin and chloramphenicol than either (b) or (c).