Salmonella Control in Poultry

Salmonella control is a top priority for regulatory agencies, public health organizations, and food production companies based on the steady number of associated foodborne illnesses. While much effort has focused on better understanding salmonellosis and managing Salmonella during food production, the estimated 1.2 million cases that occur in the U.S. each year are well above current public health goals of about 36,000 cases nationwide.

The basic ecology of agricultural and animal-derived food products results in a normal association with Salmonella. ­Accordingly, Salmonella management is a major challenge for the food production industry. Understanding salmonellosis is difficult because the number of Salmonella cells required is unclear. Scientists also do not understand whether every strain of Salmonella can cause illness. There are thousands of strains found in nature. Previous risk assessments have indicated that there is a dose-response relationship between the number of Salmonella present and the severity and number of individuals infected after consumption of contaminated poultry products. The ability to identify points in processing with higher levels of contamination would greatly assist processors in better managing Salmonella levels present in finished food products.

Public health data illustrate an important association between raw poultry and salmonellosis. Accordingly, the USDA Food Safety and Inspection Service (FSIS) and the poultry industry are working on programs to improve Salmonella control. This August, USDA-FSIS published the Final Rule for Modernization of Poultry Inspection. The new regulation is expected to have a direct impact on the number of Salmonella-associated illnesses each year with advancing inspection practices and more science-based detection methods. With a more preventive focus, this new inspection process will allow USDA-FSIS to verify safety programs and provide a more comprehensive assessment of process control by examining sanitation procedures, reviewing records, and collecting test samples for microbiological analysis.

Furthermore, poultry processors must consider Salmonella a food safety hazard. Failure to implement and manage control procedures could result in regulatory action. Therefore, processors will be expected to engage in food safety management, including sampling and testing programs for Salmonella.

The Need for New Techniques

Rapid detection assays have been used by the food processing industry for many years to detect extremely low levels of Salmonella. Based on currently available commercial technology, a sample enrichment period is required to reliably and qualitatively detect low levels of Salmonella in food products. Qualitative detection of foodborne pathogens indicates presence or absence in the test portion analyzed and is not intended to provide quantitative information on the starting levels of the pathogen. Thus, if a series of independent samples were determined to be positive, it would be unknown how many cells were originally present at the time of sampling and testing. Quantitative data, which would allow for a better understanding of contamination levels, would be useful to food processors of raw agricultural products that consistently have pathogens present. It can also be used to better understand the extent of contamination for a given sample type and point in the process or environment.

Quantitative pathogen analysis has typically been conducted using conventional microbiological methods. A most probable number (MPN) analysis is typically used to enumerate low concentrations of pathogens, while direct plating would be used for higher levels. These methods are time-consuming, labor intensive, and expensive. Their accuracy can also be confounded by the dynamics associated with enrichment and bacterial isolation. Specifically, MPN analysis relies on enrichment of a series of cultures that have been serially diluted and subjected to detection to yield a ratio of positive and negative samples. This ratio is then used to estimate the number of starting cells per gram or milliliter in the original, un-enriched sample. This approach is labor intensive and expensive due to the number of dilutions, enrichments, detections, and confirmations required per sample.