Skip to main content
ARS Home » Research » Publications at this Location » Publication #261088

Title: Testing of duplicate rinse aliquots for presence of Salmonella

item Cason Jr, John
item SMITH, DOUGLAS - North Carolina State University
item Buhr, Richard - Jeff
item Hinton, Jr, Arthur
item Cox, Nelson - Nac

Submitted to: International Poultry Scientific Forum
Publication Type: Abstract Only
Publication Acceptance Date: 12/1/2011
Publication Date: 1/24/2011
Citation: Cason Jr, J.A., Smith, D.P., Buhr, R.J., Hinton Jr, A., Cox Jr, N.A. 2011. Testing of duplicate rinse aliquots for presence of Salmonella [abstract]. International Poultry Scientific Forum. p. 75-76.

Interpretive Summary:

Technical Abstract: Testing of chicken carcass rinses for Salmonella prevalence is often performed in duplicate because of the potential importance of the results, but anecdotal reports indicate that duplicate samples often disagree. This might be due to normal variation in microbiological methods or to the testing of aliquots where the original sample volume contains low numbers of Salmonella cells. To test the latter possibility, a spreadsheet simulation was developed to test the probability that paired aliquots (each 7.5% of the original sample volume) contain at least one cell. Probabilities were calculated for 100,000 pairs of aliquots from original samples containing 1 to 50 cells each, using a random number generator to determine whether test aliquots contain at least one cell. When the original volume contains one cell, for example, each aliquot containing 7.5% of the total volume will have a 7.5% probability of containing that cell. One of the paired aliquots will be positive in 15% of cases and the pair will disagree in all of those cases. Results of the calculation demonstrate that more than 50% of all paired samples will disagree when the original volume contains from 7 to 12 cells. More than 20% of paired samples disagree until the original sample contains at least 29 cells. The original sample volume must contain 48 cells before there is a 95% probability of agreement between paired aliquots. Even if laboratory methods are 100% effective in recovering Salmonella from test aliquots, there will be substantial disagreement between paired aliquots in cases where the original sample contains relatively low numbers of Salmonella cells.