Validation of dog treats and bones for the presence of Salmonella through in-plant analysis and inoculation challenge studies

Abstract

Salmonella is a continual problem within the pet food industry. Salmonella in pet food and treats can lead to a potential human Salmonella infection through the handling of these products. The objective of this experiment was to determine the prevalence of Salmonella in various dog treats through an in-plant analysis study and an inoculation challenge study. For the in -plant analysis study, thirty raw and thirty cooked samples of lamb lung cubes, weasands (weasands), pizzles (steer pizzles), and jumbo pizzles were delivered to the Texas Tech University Food Microbiology Laboratory on ice and analyzed by the FDA approved BAX system method for detection of Salmonella. When testing raw products, 7 of 30 weasands, 0 of 30 lamb lung cubes, 26 of 30 pizzles, and 29 of 30 jumbo pizzles were positive for Salmonella. For the cooked products, 1 of 30 taffy’s was positive for Salmonella¸ while the other products had no positive cooked samples. During the inoculation challenge, two different lots of raw samples of liver tripe steaks, lamb lung cubes, taffy’s, and pizzles were delivered on ice to Texas Tech University’s Experimental Sciences Building. Fifteen samples from each lot of products were inoculated with a four-strain Salmonella cocktail at 102-5 CFU/g. After a 30-min attachment period, five control samples were sampled for initial concentration. Remaining samples were cooked in accordance with the standard cooking procedure established by the company. Before and after cooking, samples were serially diluted and plated onto xylose lysine deoxycholate agar (XLD) with a tryptic soy agar overlay for injured cells. If no colonies were present on XLD plates, detection was performed by the BAX system and immunomagnetic separation (IMS) with Remel Salmonella agglutination kit for confirmation. Statistical analysis was performed using SAS program with a significance level of α = 0.05. Cooking reduced (P < 0.01) Salmonella of lamb lung cubes by 2.85 log with one positive sample after cooking (<10 CFU/g present). In addition, cooking reduced (P < 0.01) Salmonella of liver tripe steaks by 3.11 log, taffy’s by 2.85 log, and pizzles by 5.47 log, leaving no positive samples after cooking. A baseline study was performed for pizzles and compared between two groups (ice vs. normal). No difference between the groups (P > 0.05) at 2.59 logs for normal with no BAX positives and 2.46 logs for iced with 2 BAX positives. The processing procedures were validated by inoculation challenge study with an initial inoculum of 6.95 logs resulting in a complete reduction after cooking (P < 0.01). A similar inoculation challenge was performed on bones used for dog treats including femurs, white knuckle bones, large shank bones, and twin hooves. Cooking reduced (P < 0.01) Salmonella of femurs by 3.86 log, large shank bones by 5.77 log, and twin hooves by 4.53 log, and no positive samples were detected after cooking. Furthermore, cooking reduced (P < 0.01) Salmonella of white knuckle bones by 3.62 log, but one positive sample was detected after cooking. Based on these results, modification to the processing and cooking procedures for taffy’s and white knuckle bones was recommended to ensure safeness during times of high contamination. The results from the inoculation challenge suggest that the processing and cooking procedures would be effective if a high level of Salmonella contamination occurred.