On Nov. 19, 2014, the Food Safety and Inspection Service (FSIS) published a Notice in the Federal Register announcing the availability of its cost-benefit analyses on the current non-O157 STEC testing of beef trimmings and the potential expansion of the agency’s testing program to raw ground beef and ground beef components.

In 2011, after FSIS published its determination that six Shiga toxin-producing Escherichia coli:O26, O45, O103, O111, O121, and O145 (non-O157 STEC) were adulterants, the agency began testing beef manufacturing trimmings. FSIS also announced that when it completed a cost-benefit analysis, FSIS would consider expand its testing to include other raw ground beef components and ground product.

Although FSIS will accept comments on the cost-benefit analysis until Jan. 20, 2015, it appears FSIS ultimately will expand the non-O157 testing to the other raw products.

In estimating the costs of the non-O157 testing of trim, ground and other components, FSIS focused almost exclusively on the costs attributed to the increase in the number of non-O157 samples likely analyzed by the industry and the number analyzed by the agency; an annual total of only $2.37 million.

•  For industry non-O157 testing, FSIS calculated a cost of approximately $985,000 per year; the cost of analyzing an additional 41,457 trim samples, 2,552 ground samples, 1,366 other raw ground components and 192 bench trim samples.

•  For FSIS sample costs, FSIS expects an added cost of approximately $700,000 for ground and approximately $200,000 for other components/bench trim. When totaled with the existing non-O157 sampling of trim and follow-up samples, FSIS’ total testing costs would be slightly less than $1.38 million.

In determining costs, FSIS excluded many costs related to industry testing, such as the costs attributed to the higher screen positive rate for the non-O157 STEC analysis (2 percent screen positive rate for FSIS), the costs of holding product, the costs of diverting positive product (screened or confirmed), and the cost of testing when required by customers.

FSIS did not include these costs because FSIS does not mandate industry testing for E. coli O157:H7 or non-O157 STEC.

On benefits, FSIS conceded it could not quantify the benefits, but asserted that there would be “reduced illnesses and deaths, reduced outbreak-related recalls, and improved business practices.” If FSIS persists with its low cost estimate of $2.37 million, even unquantifiable benefits would exceed the costs and support FSIS’ expansion of non-O157 STEC testing to raw ground and raw ground components.

Moving forward

Even though FSIS likely will expand its non-O157 STEC sampling, it will not happen overnight.  Given the comment period, comment review, finalization of the cost-benefit analysis, and a 3-6 month effective date, we have at least a year.

As stated repeatedly in the cost-benefit analysis, FSIS is not requiring establishments to conduct non-O157 STEC testing. Let us take FSIS at its word. When implementing its non-O157 trim testing, FSIS did not mandate routine industry testing.

That said, trim producers were responsible for validating their HACCP plans for non-O157 STEC. This meant they conducted non-O157 validation sampling to demonstrate their interventions are equally effective for all STEC. Once done, these trim producers use the more robust and accurate E. coli O157:H7 method to monitor the process and only conduct limited, periodic non-O157 STEC verification analysis.

Advice to those producing raw ground beef/ground beef components: Start generating data now on non-O157 incidence on your products. To minimize the costs of obtaining preliminary data, begin by conducting non-O157 analysis on a lot that:

•  Has tested positive for E. coliO157;

•  Has tested negative for E. coliO157, but is implicated by a positive, such as with a high event period; or

•  Is sent to cooking under inspection.

Ultimately, some random sampling would be required to validate that no matter what “flavor” of STEC is present in the raw materials, the non-O157 is controlled. This could be done by option three above, sending a random lot to cooking. Otherwise the plant would need to hold the lot sampled. Once validated, a plant can limit non-O157 STEC analysis to periodic verification sampling. If FSIS is going to expand its testing, we need to get ready now. If we do, the costs identified by FSIS may not seem as unrealistic as they first appear.


Dennis R. Johnson is a principal with Olsson Frank Weeda Terman Matz PC in Washington, D.C.  Johnson has 30 years experience in food-safety law and regulation, representing large and small meat and poultry companies.