White Paper on
Escherichia coli O157:H7
Food Safety and Inspection Service
U.S. Department of Agriculture
The Food Safety and Inspection Service (FSIS) is exploring whether
further changes are needed in its policy regarding E. coli
O157:H7 in light of new information that is emerging about the pathogen
and its relation to human health. This paper presents background
information on FSIS’ current policy regarding E. coli O157:H7 and
discusses new information and developments that will likely have a
bearing on the Agency’s policy. FSIS will re-evaluate its policy on
E. coli O157:H7 through an open, participatory process, soliciting
input from all of its various constituents. FSIS plans to hold a public
meeting in the near future on this issue, but a date has not yet been
In 1994, FSIS declared that E. coli O157:H7 is an adulterant
in ground beef and instituted a testing program for the pathogen. FSIS
established an end-product sampling program for ground beef to keep
contaminated product from reaching consumers and to spur industry to
institute pathogen reduction and HACCP-associated verification programs
to reduce the risk of this pathogen in beef products. The Agency’s
policy, as reflected in the attached FSIS Directive 10,010.1, dated
February 1, 1998, was shaped in five major ways.
First, because E. coli O157:H7 was an emerging pathogen and
not very well understood, the Agency was forced to design its policy
based largely on the limited data available at the time. The fact that
the pathogen produces severe and sometimes fatal consequences at what
was believed to be a very low infectious dose (fewer than 10 organisms)
spurred FSIS to take quick action.
Second, the pathogen was difficult to find in meat products. The lack
of information about the pathogen made it difficult to determine whether
the low rate at which it was found was attributable to the rarity of the
pathogen, or to the fact that the method for detecting it was not very
sensitive. Because the pathogen was found at a low rate, it has
generally not been considered a hazard "reasonably likely to occur" and
thus establishments have not considered it necessary to address it in
their HACCP plans.
Third, the fact that the pathogen was difficult to find had an impact
on the design of FSIS’ testing program. Because FSIS could not have
confidence in a single negative result, once a positive is found by the
Agency from a plant that grinds meat, there must be 15 consecutive
negative findings from daily sampling before FSIS returns the plant to
routine testing. Moreover, once such a plant has a positive finding,
FSIS inspection personnel are instructed to select a sample if the
positive finding occurred within the preceding six months--regardless of
whether the plant has its own routine sampling program. Furthermore,
sampling is conducted at retail locations that grind meat as well as at
plants that do so. Sampling at retail, the point in the distribution
chain closest to the consumer, provides an additional opportunity to
intercept adulterated product.
Fourth, the fact that outbreaks attributed to E. coli O157:H7
were associated with ground beef led FSIS to focus its regulatory
program on comminuted (ground) products. The first case control study
for E. coli O157:H7 implicated ground beef as the risk factor
most significantly associated with illnesses. The focus on ground
product made sense because grinding moves the pathogen away from the
surface of the product to the inside, so thorough cooking is needed to
kill the pathogen. Americans were not used to cooking their ground beef
patties, including hamburgers, thoroughly, and ground beef is extremely
popular among young children who are affected most severely by the
pathogen. These facts were used to make the initial decision in 1994
that E. coli O157:H7 adulterates ground beef as well as the
decision in January 1999 that the presence of E. coli O157:H7
would adulterate not just ground beef, but any non-intact product or
intact product intended for use as a non-intact product (Docket No.
97-068N, 64 Fed. Reg. 2803, January 19, 1999).
Fifth, cooking product found to be contaminated with E. coli
O157:H7 was the only intervention known to FSIS at the time. That is why
FSIS’ consistent advice has been that when this pathogen is found in
non-intact meat or intact meat intended for non-intact use, the meat
must be diverted for cooking under the control of FSIS.
Several major developments that will greatly enhance FSIS’ knowledge
and understanding of E. coli O157:H7 are now in the offing and
are likely to have a bearing on the Agency’s policy.
First, new information is emerging suggesting that E. coli
O157:H7 is not as rare as previously thought. This new
information comes from several sources.
In September 1999, FSIS began using a method for analyzing samples of
products for E. coli O157:H7 that is four times more sensitive
than the previous method. As a result of the new method, 40 percent (21
out of 53) of the positive samples found since the testing program began
in 1994 have been found since the new test was initiated. This suggests
that the low rate of positive findings may have had more to do with the
sensitivity of the method being used than with the rarity of the
Further evidence that E. coli O157:H7 may occur more
frequently than was thought to be the case is provided by recent
estimates of foodborne illness released by the Centers for Disease
Control and Prevention (CDC). Those data show a much higher rate of
illness from E. coli O157:H7 than had been previously reported by
CDC, due mainly to milder cases of illness revealed by surveillance
data. [Mead, Paul S., et al., "Food-Related Illness and Death in
the United States," Journal of Emerging Infectious Diseases,
Vol. 5, No. 5, 1999]. While not all of these illnesses are attributable
to beef, they do support the observation that this pathogen occurs more
frequently than was previously thought, and that the great majority of
cases are limited to relatively mild symptoms that are unlikely to
result in an infected person seeking medical care or to be reported to
public health officials.
In addition to these data, the American Meat Institute is conducting
a study in slaughter plants that should provide further insight into
whether E. coli O157:H7 is a rare pathogen; particularly, whether
it occurs on hides and freshly slaughtered carcasses of beef with some
regularity. Under its protocol, 1 in 300 carcasses are tested for E.
coli O157:H7 before hide removal, after hide removal, and after
pathogen reduction interventions have been applied. The protocol was to
run for 30 days, starting in early September. Twelve plants are involved
in the protocol, although the specific plants involved have not been
identified to FSIS. It is FSIS’ understanding that most of the testing
is complete, and that the data are being analyzed. FSIS expects to
receive the results of the testing by the end of November 1999.
The Agriculture Research Service, in Clay Center, Nebraska, also is
carrying out research related to prevalence, and FSIS itself plans to
conduct some sampling to assess the feasibility of identifying E.
coli O157:H7 on carcasses and of establishing a routine,
Agency-directed sampling program to supplement or replace FSIS' ongoing
ground beef testing.
Second, a soon-to-be completed FSIS risk assessment for E. coli
O157:H7 in ground beef and trimmings will better enable both the Agency
and industry to identify interventions that can lead to public health
improvements and weigh available options. FSIS has based this
risk assessment on a comprehensive review of the available literature
and data (see 63 Fed. Reg. 44232, August 18, 1998). Although needed peer
reviews make it unlikely that the risk assessment will be completed
before spring 2000, FSIS is planning to make a presentation on the risk
assessment to the National Advisory Committee on Microbiological
Criteria for Foods in December 1999. At that time, the risk assessment
team from FSIS' Office of Public Health and Science will seek the
Committee’s support for the assessment’s format and design. When the
risk assessment on ground beef and trimmings is complete, FSIS expects
to expand it to cover all meat products, as well as other products that
may be affected by E. coli O157:H7.
Third, FSIS has received new information that could affect the scope
of products affected by its policy. Specifically, FSIS has received data
and information on blade tenderized roasts and steaks. As stated above,
FSIS' policy is that non-intact cuts of meat are adulterated if
contaminated by E. coli O157:H7. During a March 1999 public
meeting on the Agency's E. coli O157:H7 policy, preliminary
findings of research being done at Kansas State University on blade
tenderized beef steaks were presented to the Agency. The researchers
stated that the blade tenderization process transfers approximately 3-4
percent of surface contamination to the interior of the muscle. The
researchers pointed out that proper cooking to a specified
time/temperature combination resulting in rare steaks could reliably
result in safe product. Industry members have further stated that muscle
systems from which steaks are derived could be removed from larger
primal or sub-primal cuts hygienically. The beef industry has been
persistent in encouraging FSIS to exempt blade tenderized product,
especially when derived hygienically or with reduced possibilities for
becoming contaminated, from the scope of the E. coli O157:H7
As of fall 1999, FSIS has tentatively determined that there is
insufficient information regarding the hygienic processing of muscle
systems to narrow the scope of products affected by the E. coli
O157:H7 policy. FSIS expects its planned effort to broaden the risk
assessment will address some of the issues raised by the industry.
Meanwhile, FSIS has encouraged industry to label their intact and
non-intact primal and sub-primal cuts with appropriate cooking
statements. The 1999 Food Code (section 3-401.11) prescribes appropriate
cooking instructions for intact versus non-intact steaks for destruction
of organisms of public health concern.
Fourth, there may be interventions other than cooking product under
FSIS control that could be used to address a positive E. coli
O157:H7 finding. For example, irradiation offers the possibility of
treating raw meat products to eliminate E. coli O157:H7. In
addition, FSIS is willing to consider whether other alternatives to
cooking product within an FSIS-inspected establishment could be used to
address a positive finding.
Several other considerations are likely to be important as the Agency
reviews its policy on E. coli O157:H7. For example, the fact that
on January 25, 2000, all meat and poultry plants will be operating under
the Pathogen Reduction and HACCP rule will have a positive impact on
food safety and may affect the Agency’s approach to the regulation of
E. coli O157:H7. In addition, actions taken by the meat industry to
reduce the pathogen at the production level also would be considered by
FSIS in reviewing its policy.
Areas for Consideration
These new developments could have a significant impact on FSIS’
policy on E. coli O157:H7. Following are some areas for
- If FSIS finds that E. coli O157:H7 occurs with some
regularity on hides and carcasses of feedlot cattle but not on cattle
raised under different production practices, should the pathogen be
considered a hazard "reasonably likely to occur" only in slaughter and
processing operations that use feedlot cattle, and thus be addressed in
their HACCP plans?
- Should FSIS re-design its testing program? Specifically:
Are any changes needed in the proportion of
samples taken in-plant and at retail?
Should FSIS alter its policy that 15
consecutive samples be negative after a positive finding?
Should FSIS continue selecting a sample if a
plant has a positive finding within the last 6 months, or should the
Agency defer to plant routine testing completely and remove the 6-month
restriction? If FSIS sampling is continued under these circumstances,
should the rules for the random selection of samples be changed?
Should FSIS sampling of carcasses replace or
supplement ground beef sampling at slaughter plants?
- Should FSIS consider a plant's generic E. coli and
Salmonella results in making its decision on whether to target a
plant’s products for sampling?
- What effect should a plant’s testing or verification program have
on whether and how FSIS targets its testing in that plant?
- How should FSIS treat non-intact product? Specifically, should
blade-tenderized beef steaks and roasts--with specific cooking
instructions for destroying the pathogen and handling instructions for
preventing cross-contamination and temperature abuse-- be treated the
same as other non-intact beef with regard to the FSIS policy?
- How effective are voluntary producer actions in providing animals
with reduced levels of E. coli O157:H7 to plants, and should
these voluntary activities, if effective, affect slaughter plants'
strategies and FSIS’ policy?
FSIS anticipates a number of developments in the coming months that
will have a significant bearing on its policy on E. coli O157:H7.
It is FSIS' intent to be guided in its policy development process by the
best available science and to reevaluate its policy through an open,
participatory process. FSIS will be soliciting input on these and other
questions at its upcoming public meeting on E. coli O157:H7.