Remarks prepared for delivery by USDA Under Secretary for Food Safety Dr. Richard Raymond, at
the ARS-FSIS workshop, on March 7, 2006, in Annapolis, MD..
Note: Slides are available in an attached PDF document; individual
pages are linked within the text.
Thank you and good afternoon. I appreciate this opportunity
to speak with you today about our efforts to combat Salmonella
and lay the foundation for a more robust risk-based inspection
I often talk about how there are science-based solutions to
nearly every food safety challenge that we face today. I want
to ask everyone here a favor, please don't make me a liar!
This year marks the 100th anniversary of the passage of the
Federal Meat Inspection Act. The Act ushered in a new era of
food safety on a national level, but after a century its age
is beginning to show.
It was based on visual examination for visible signs of disease.
The future demands that we be able to focus on the things which
the human eyes cannot see, the human nose cannot smell, and
that our fingers cannot feel.
Your future research, insight, and effort will be critical
to ensuring that our food safety system will be able to meet
these demands in the next century.
That is why I believe strengthening FSIS' relationship with
ARS, and for that matter all of our research partners, is critical
to our future success.
Simply put, FSIS' food safety regulatory policies are built
upon the strong foundation provided by your scientific data.
A foundation that has been strengthened by the contributions
of the National Alliance for Food Safety and Security (NAFSS).
In 2005, FSIS, ARS, and NAFSS developed, funded and conducted
an important nationwide risk-assessment on Listeria monocytogenes
in ready-to-eat luncheon meats. This study will be critical
to determining FSIS' future actions to combat this dangerous
Our relationship with ARS has only gotten closer since Dr.
Merle Pierson became the Deputy Under Secretary for Research,
Education and Economics. I cannot think of anyone better for
For three and a half years, Dr. Pierson served as Deputy Under
Secretary and then Acting Under Secretary for Food Safety. He
knows our issues and challenges inside and out. He also knows
where the research money is, which might prove dangerous.
I want to also recognize the important contributions made by
researchers from the CDC and FDA. Their efforts have helped
make the Foodborne Diseases Active Surveillance Network an essential
public health tool. Together we are making great strides toward
being able to link foodborne illness estimates to specific food
vehicles. This will prove to be a tremendous asset as we work
to further protect the nation's food supply
Thanks to the tools you have provided us, we have seen dramatic
declines in the prevalence of pathogens in the products that
we regulate and the numbers of foodborne illnesses stemming
from these pathogens.
4) | (SLIDE
5) | (SLIDE
Another significant measure of how our science-based policies
and control measures in plants are affecting public health can
be found in an annual report published by the CDC.
According to the CDC, there have been significant declines
from 1996 to 2004 in illnesses caused by E. coli O157:H7,
Listeria monocytogenes, Campylobacter, and
Compared to the 1996-98 baseline, illnesses caused by E.
coli O157:H7 decreased by 42 percent. I am happy to report
that we met the Department of Health and Human Services' Healthy
People 2010 objective for E. coli O157:H7 six years
In 2004, the CDC reported .9 cases of E. coli O157:H7
infections per 100,000 people. This is lower than the Department
of Health and Human Services' Healthy People 2010 objective
of one case of E. coli O157:H7 per 100,000 people.
8) | (SLIDE
9) | (SLIDE
Taken together these human health results and decreasing numbers
of pathogens in our sampling program indicate that our risk-based
approach is working. However, you deserve to share some of the
credit for these successes. After all, it was your dedication
and research that made our science-based approach possible.
All of this is good news, but we still have areas of concern.
Areas that can benefit from your work.
A specific and critical concern is Salmonella. When
FSIS reported its 2003 data, the Agency acknowledged that the
percentage of positive Salmonella tests had increased
slightly in all three poultry categories. While the 2004 data
showed more mixed results, there was a continued increase for
young chicken (or broiler) carcasses, and that number rose again
The four-year trend of rising rates on broiler carcasses does
not bode well for the public's health. It has risen nearly 50
percent in just three years, and you do not have to be a food scientist
to know that this is not a positive development.
The CDC's most recent FoodNet report is not much better. It
is clear that the overall incidence of Salmonella infections
remains far above our objective. In 2003, there were 14.5 cases
of Salmonella infections per 100,000 people. That's
43,500 people per year with culture proven Salmonellosis.
The CDC estimates that the actual number could be over one
million. That is a tremendous burden on human health.
While CDC did report that Salmonella infections
dropped eight percent, only one of the five most common strains,
which account for 56 percent of the reported Salmonella
infections in 2004, declined significantly. That strain was
Salmonella Typhimurium which declined 38 percent.
Salmonella Enteritidis and Salmonella Heidelberg
neither increased nor decreased significantly. Incidences of
Salmonella Newport increased by an alarming 41 percent.
Given the challenge we face with Salmonella, it is
imperative that we take a risk-based approach to investigating
and controlling the incidence of Salmonella in meat,
poultry and egg products. An approach that will be based upon
your scientific research.
I believe that we can leverage new technologies and cutting
edge research, not only to reach the Healthy People 2010 objective,
but to drive the numbers even lower.
Since the prevalence rate in broiler chickens seems to be a
trouble spot, we are revising the performance measure for Salmonella
on this particular product. This is in part due to three weaknesses
we have identified in the current measure.
The first one is that the measure is scientifically unsound.
The FSIS regulatory testing program that is the source of the
data used in the current performance measure does not provide
a true measure of prevalence of the pathogen.
For example, if samples from an establishment are only taken
early in the first shift, then those samples are not providing
us with an accurate understanding of the environment, or workforce's
characteristics of that establishment's second shift.
The second weakness is that the current measure is for generic
Salmonella, and includes serotypes that are not, or
are rarely attributed to foodborne illness. There are many known
serotypes of Salmonella found in broilers. Each serotype
cause human illness with varying severity. In fact, the most
common has been found not to be a significant factor in human
The third weakness is that the current testing program is not
consistent with FSIS' goal of transitioning to a more robust
risk-based inspection system.
To generate the data needed to report on the current measure,
FSIS would need to continue scheduling a sample set for every
plant each year under the current strategy. But plant process
controls for Salmonella vary widely, as do their results.
Since 2003, aggregate percent positives in sample sets have
increased each year from 11.5 percent in 2002, to 16.3 percent
in 2005 while still remaining within regulatory performance
standards. In order to improve program performance, FSIS is
working to strengthen its verification testing program by making
it more risk-based.
The image on the screen illustrates an analysis of 103 large
broiler plants where FSIS completed five or more HACCP verification
sets from 1998-2004.
Twenty-six of the 103 establishments (25%) routinely demonstrated
Salmonella control, with six or fewer Salmonella
positive tests out of 51 tests per set. Another 46 establishments
(45%) exceeded half the standard without failing at least once
and another 31 (30%) exceeded the standard at least once.
You can see that a substantial number of the plants were below
50 percent of the existing performance standard on every set.
If those plants can accomplish it then we believe others can
do the same.
15) The plants that fail one or more sets, and the plants
that fluctuate above and below 50 percent but never fail, will
frequently have a set where they perform very well.
This helps to explain why we find that the majority of set
results are below 50 percent of the existing Salmonella
prevalence standard. Even with 35 percent of the analyzed establishments
failing one or more sets, less than 25 percent of those sets
were above the standard.
We know that lowering the prevalence of Salmonella
in poultry products is possible. We have seen one plant with
a Salmonella prevalence rate of 30 percent reduce that
rate to 2 percent after a food safety assessment. Our goal as
we move forward must be to make these reductions a reality across
The Agency has also found strong evidence that plants that
have consistently achieved a percent positive rate in sample
sets at or below half the current regulatory performance standard
are less likely to produce raw product that have the serotypes
of Salmonella that are causes of human illness.
As a result, achievement of performance goals established under
the new measure would provide a better indication of process
control and relate more directly to the improved safety of broilers.
Nearly two weeks ago, we announced an initiative to reduce
Salmonella in meat and poultry products. It incorporated
11 steps, including increased sampling in plants where it is
most needed and quarterly publication of nationwide Salmonella
data by product class.
The initiative uses the old carrot and stick approach to encourage
change by offering establishments the possibility of improved
efficiency, incentives, and also disincentives based on their
progress. I am not going to get into the specifics of our 11-point
plan, today. After all that is why we keep people like Dr. Patricia
Bennett on the payroll. She will be detailing our new initiative
tomorrow. However, I am confident that this initiative will
help FSIS to be pro-active and take action before people get
Robust Risk-based Inspection Systems
FSIS must develop these critical abilities further and
that is why it is so important that we begin to lay the foundation
for a more robust risk-based inspection system.
I want to focus our time and valuable resources on prevention,
rather than on response. Command and control was the old agency
mantra. We are now after a common sense, cost-effective public
health strategy that best serves the American consumer by preventing
I did not take this job to recall ground beef, ready-to-eat
chicken tenders and sausage. As a doctor, I want to prevent
contamination and reduce the number of people getting sick each
year. I want to focus on prevention.
I know with your continued support, we can further improve
upon the food safety successes that we have already seen.
Our current system, while strong, is not suited to the future
realities of food safety and public health, and we will need
the ability to anticipate and quickly respond to food safety
challenges before they negatively affect public health offered
by an enhanced risk-based system.
This is vital, as is a system that will allow us to use our
finite resources more effectively and efficiently to further
improve food safety.
However, it is important to note that FSIS already uses a risk-based
approach to food safety. Our goal is to further enhance and
strengthen that system so that we are prepared for the food
safety challenges in the next century. The research that you
conduct will be crucial to this effort.
To continue our progress toward a fully implemented risk-based
inspection system, we need to be sure that we communicate openly
and often with all of our food safety stakeholders. I assure
you that we will use a transparent and inclusive process to
seek input on a wide range of issues related to creating a more
robust risk-based inspection system.
At the last meeting of the National Advisory Committee on Meat
and Poultry Inspection (NACMPI) in November, the Committee recommended
a third-party approach to assist us in reaching out to, and
gaining input from, our stakeholders.
We are now in the process of selecting a neutral third party
and a subcommittee of NACMPI has been established to provide
regular, ongoing guidance. I urge everyone here to take an active
role. Your experience and knowledge is invaluable.
We all know that we can save lives through sensible science-based
policies, and together I know we will do just that.
Before I close, I want to remind you that we have a strong system
in place. But we must continue to improve and enhance our food
safety system. A system not moving forward is a system falling
behind. Mother Nature is always changing, and we must change
along with her.
Remember, the state of public heath is constantly evolving.
It is your responsibility to ensure that whatever that next
evolution is that it will be grounded in science.
The entire history of public health is really just a long list
of achievements that were at one time thought impossible. The
work of dedicated scientists like you ensures that the impossible
will eventually become commonplace given enough time.
Did you know that 100 years ago, 1 in 5 coffins contained
a child under 5 years old? Today that number is only 1 in 100
coffins. That is an amazing accomplishment that has had a profound
impact on our society and everyone in this room.
Clean water, sewage, vaccines and antibiotics — but a
safe food supply has also played a role in this amazing phenomenon.
I know that together we can continue to lower the number of
After all, the bottom line is that we all have the same objectives
- safe food and healthy people. We must never lose sight of
these common goals.
Thank you again for your time and I look forward to seeing
the research that will flow from this strong collaborative relationship.