USDA
on Friday issued draft guidelines on the design of sampling
and testing programs for E. coli o157:H7 for establishments
that make or use boneless beef manufacturing trim and other
raw ground beef components. (See USDA
taking aim at bench trim, retail grinding for closer scrutiny
on Meatingplace.com , August 13, 2008.)
The document, issued by USDA's Food Safety and Inspection
Service, suggests frequent trim testing, particularly during
the warmer months April through October. "Optimally, every
production lot should be sampled and tested before leaving
the supplier and again before use at the receiver." It also
suggests follow-up on each positive result with the supplier,
whether the trim was produced on site or procured from another
supplier.
FSIS recommends testing finished product even if the source
material has been tested and found negative, as there can
still be pockets of contamination in the product that were
not in the test sample, the product might have become contaminated
after it was sampled, or E. coli cells within the sample
might have gone undetected because the number of cells was
below the limit of detection.
On the days trim is cut, unless primal and sub-primal cuts
are effectively treated with antimicrobials after trimming,
these cuts and food contact surfaces should be assess for
E. coli O157:H7, FSIS suggested.
FSIS is setting a percent positive guidance value of pre-tested
trimmings at 1.5 percent. While not a regulatory limit,
FSIS is using the 1.5 percent value to identify a statistical
framework to identify when a process, during some period
of time, is producing product in a way that does not adequately
control the occurrence of E. coli O157:H7.
The agency also suggests trimmings from each supplier be
tested as a separate lot to help decrease the extent of
product that would be recalled or sent for cooking when
a positive test result is obtained.
View the full guidelines document here.
Comments are due by September 15, 2008 and can be sent to:
FSISGuidanceDocumentComments@fsis.usda.gov.
|
