Skip to main content
ARS Home » Southeast Area » Athens, Georgia » U.S. National Poultry Research Center » Poultry Microbiological Safety and Processing Research Unit » Research » Publications at this Location » Publication #295691

Title: Sampling naturally contaminated broiler carcasses for Salmonella by three different methods

item Cox, Nelson - Nac
item Buhr, Richard - Jeff
item SMITH, D - North Carolina State University
item CASON, J - Former ARS Employee
item Rigsby, Luanne
item Bourassa, Dianna
item Cray, Paula
item Cosby, Douglas

Submitted to: Journal of Food Protection
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/8/2013
Publication Date: 3/1/2014
Citation: Cox Jr, N.A., Buhr, R.J., Smith, D.P., Cason, J.A., Rigsby, L.L., Bourassa, D.V., Cray, P.J., Cosby, D.E. 2014. Sampling naturally contaminated broiler carcasses for Salmonella by three different methods. Journal of Food Protection. 77(3):493-495.

Interpretive Summary: While Salmonella is widespread in nature and can be found in many wild and domestic animals, raw poultry is a significant source of human salmonellosis. Postchill whole carcass rinsing (WCR) and neck skin (NS) maceration are the regulatory methods used to detect salmonellae from commercially processed broilers in the United States and the European Union, respectively. In 1975, a whole carcass enrichment method (WCE) was developed as a research tool and can detect as few as 8 Salmonella cells per broiler carcass. This study was to compare each of the three methods to recover naturally occurring Salmonella from the same carcass. WCR and NS were very similar in detection of Salmonella, but Salmonella were detected 4 times more often with WCE than either of the other two methods. Most frequently used sampling methods are selected for ease of performance, expense and other factors, but one should remember that along with these advantages comes the disadvantage of producing many false negatives, which also means they are ineffective to evaluate an intervention strategy to produce Salmonella-free broilers.

Technical Abstract: Postchill neck skin (NS) maceration and whole carcass rinsing (WCR) are frequently used methods to detect salmonellae from commercially processed broilers. These are practical, nondestructive methods, but they are insensitive and may result in frequent false negatives (20 to 40%). NS samples only 4% of the skin from the broiler carcass by weight while WCR will not detect firmly attached Salmonella with only 7.5% of the rinsate utilized. Whole carcass enrichment (WCE) involves incubation of the whole carcass overnight in a preenrichment broth and can recover as few as 8 inoculated Salmonella cells per carcass. The objective of this study was to use NS, WCR, and WCE sampling to detect naturally occurring Salmonella from the same commercially processed broiler carcass either prechill or postchill. Ten carcasses were obtained prechill and another 10 postchill on each of two replicate days from each of two commercial processing plants. From each carcass, 8.3g of neck skin was sampled (NS) and then the carcass was rinsed with 400 ml of 1% buffered peptone water. Thirty ml were removed and incubated (WCR), and the remaining 370 ml of broth and the carcass were incubated at 37oC for 24 h (WCE). Overall Salmonella were detected on prechill carcasses 21/40 (52.5%), 24/40 (60%) and 32/40 (80%) with the NS, WCR, and WCE, respectively, while postchill carcasses were 2/40 (5%), 2/40 (5%) and 19/40 (48%) for NS, WCR and WCE, respectively. The 240 samples (40 carcasses X 3 methods X 2 plants), prechill carcasses were 77/120 (64%) positive for Salmonella while postchill carcasses were 23/120 (19%) positive. Therefore the commercial chilling process reduced the positive sample prevalence by 45%. Salmonella were detected on 20% (24/120) of the samples from plant 1 and 63% (76/120) of the carcasses from plant 2. This study demonstrates significant differences in Salmonella prevalence among sampling methods, both before and after immersion chilling, and between processing plants on days that samples were taken.