Antimicrobial susceptibility of Salmonella isolated from various products, from 1999 to 2003.

Foodborne salmonellosis continues to be a major health concern worldwide; thus, detection and tracking of antimicrobial resistance in Salmonella isolates is of interest. The U.S. Food and Drug Administration initiated antimicrobial sensitivity screening of Salmonella isolates from food and related samples in 1999. This paper summarizes the antimicrobial resistance data for Salmonella isolates obtained from 1999 to 2003. A total of 22,231 imported and domestic samples were analyzed for Salmonella, of which 1,319 (5.9%) yielded the pathogen. Since more than one culture was isolated from some samples, the total number of isolates obtained and tested for antimicrobial sensitivity was 1,382. Antimicrobial sensitivity screening was performed with the disc diffusion assay on 11 antimicrobial agents. Of the 1,108 food isolates screened, 42.1% (n = 467) were serotypes Weltevreden, Newport, Lexington, Senftenberg, Typhimurium, Saint Paul, Paratyphi, Enteritidis, Thompson, and Bareilly. A total of 249 (18.0%) isolates from all sources were resistant to two or more antimicrobials. Resistance to sulfisoxazole, streptomycin, and tetracycline was most common, whereas resistance to ciprofloxacin was least common. Weltevreden (n = 148) was the most common serotype isolated from food, but only nine (6.1%) of these isolates were resistant to two or more antimicrobials. In contrast, although Derby was recovered only 19 times, 11 (57.9 %) of these isolates were resistant to two or more antimicrobials. Of the 274 isolates from animal feed, dog treats and environmental swabs, 49.6% (n = 136) belonged to serotypes Infantis, Mbandaka, Anatum, Senftenberg, Typhimurium, Montevideo, Cerro, Enteritidis, and Bredeney, with 76 (27.7%) of these isolates resistant to two or more antimicrobials. Only limited trends in antimicrobial resistance were observed over time, with resistance to sulfisoxazole increasing, resistance to tetracycline decreasing, and resistance to streptomycin fluctuating.

Validation of a PCR-based method for the detection of various rendered materials in feedstuffs using a forensic DNA extraction kit.

A method trial was initiated to validate the use of a commercial DNA forensic kit to extract DNA from animal feed as part of a PCR-based method. Four different PCR primer pairs (one bovine pair, one porcine pair, one ovine primer pair, and one multispecies pair) were also evaluated. Each laboratory was required to analyze a total of 120 dairy feed samples either not fortified (control, true negative) or fortified with bovine meat and bone meal, porcine meat and bone meal (PMBM), or lamb meal. Feeds were fortified with the animal meals at a concentration of 0.1% (wt/wt). Ten laboratories participated in this trial, and each laboratory was required to evaluate two different primer pairs, i.e., each PCR primer pair was evaluated by five different laboratories. The method was considered to be validated for a given animal source when three or more laboratories achieved at least 97% accuracy (29 correct of 30 samples for 96.7% accuracy, rounded up to 97%) in detecting the fortified samples for that source. Using this criterion, the method was validated for the bovine primer because three laboratories met the criterion, with an average accuracy of 98.9%. The average false-positive rate was 3.0% in these laboratories. A fourth laboratory was 80% accurate in identifying the samples fortified with bovine meat and bone meal. A fifth laboratory was not able to consistently extract the DNA from the feed samples and did not achieve the criterion for accuracy for either the bovine or multispecies PCR primers. For the porcine primers, the method was validated, with four laboratories meeting the criterion for accuracy with an average accuracy of 99.2%. The fifth laboratory had a 93.3% accuracy outcome for the porcine primer. Collectively, these five laboratories had a 1.3% false-positive rate for the porcine primer. No laboratory was able to meet the criterion for accuracy with the ovine primers, most likely because of problems with the synthesis of the primer pair; none of the positive control DNA samples could be detected with the ovine primers. The multispecies primer pair was validated in three laboratories for use with bovine meat and bone meal and lamb meal but not with PMBM. The three laboratories had an average accuracy of 98.9% for bovine meat and bone meal, 97.8% for lamb meal, and 63.3% for PMBM. When examined on an individual laboratory basis, one of these four laboratories could not identify a single feed sample containing PMBM by using the multispecies primer, whereas the other laboratory identified only one PMBM-fortified sample, suggesting that the limit of detection for PMBM with this primer pair is around 0.1% (wt/wt). The results of this study demonstrated that the DNA forensic kit can be used to extract DNA from animal feed, which can then be used for PCR analysis to detect animal-derived protein present in the feed sample.

Antimicrobial resistance of food-related Salmonella isolates, 1999-2000.

Salmonellosis is a major foodborne infection in the United States, and strains of Salmonella that are resistant to a variety of antimicrobial agents have become a major public health concern. To estimate the incidence of antimicrobial-resistant Salmonella in our food supply, the U.S. Food and Drug Administration (FDA) has initiated screening of foodborne isolates for sensitivity to antimicrobial agents, including several antibiotics. Salmonella cultures (n = 502) isolated by FDA laboratories during fiscal year 2000 (1 October 1999 through 30 September 2000) from domestic and imported food products and related samples were tested for susceptibility to each of 12 antimicrobial agents using a disc diffusion assay. Because all isolates were resistant to rifampin (5 or 25 microg), only results with the remaining 11 antimicrobial agents are discussed in this paper. Of the 502 isolates, 247 (49.2%) were resistant to one or more antimicrobial agents, and of these 247 isolates, 170 (68.8%) were resistant to one antimicrobial agent, 33 (13.4%) to two antimicrobial agents, 25 (10.1%) to three antimicrobial agents, 7 (2.8%) to four antimicrobial agents, 8 (3.2%) to five antimicrobial agents, and 2 (0.8%) each to six and seven antimicrobial agents. No isolates were resistant to norfloxacin, whereas only seven were resistant to sulfamethoxazole/trimethoprim, six to trimethoprim, three to gentamicin, and one to ciprofloxacin. These results, for the first time, provide a baseline of data on the incidence of antimicrobial-resistant Salmonella in the U.S. food supply, which should be useful in determining the evolution of antimicrobial resistance in the future.