This is the accessible text file for GAO report number GAO-14-744 
entitled 'Food Safety: USDA Needs to Strengthen Its Approach to 
Protecting Human Health from Pathogens in Poultry Products' which was 
released on October 20, 2014. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as 
part of a longer term project to improve GAO products' accessibility. 
Every attempt has been made to maintain the structural and data 
integrity of the original printed product. Accessibility features, 
such as text descriptions of tables, consecutively numbered footnotes 
placed at the end of the file, and the text of agency comment letters, 
are provided but may not exactly duplicate the presentation or format 
of the printed version. The portable document format (PDF) file is an 
exact electronic replica of the printed version. We welcome your 
feedback. Please E-mail your comments regarding the contents or 
accessibility features of this document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

United States Government Accountability Office: 
GAO: 

Report to Congressional Requesters: 

September 2014: 

Food Safety: 

USDA Needs to Strengthen Its Approach to Protecting Human Health from 
Pathogens in Poultry Products: 

GAO-14-744: 

GAO Highlights: 

Highlights of GAO-14-744, a report to congressional requesters. 

Why GAO Did This Study: 

USDA is responsible for ensuring the safety of poultry products. The 
Centers for Disease Control and Prevention (CDC) report the U.S. food 
supply is one of the safest in the world, yet estimate that Salmonella 
and Campylobacter contamination in food causes more than 2 million 
human illnesses per year. Poultry products contaminated with pathogens 
cause more deaths than any other commodity. GAO was asked to examine 
USDA's approach to reduce these pathogens in poultry products. 

GAO's objectives were to (1) describe actions USDA has taken since 
2006 to reduce Salmonella and Campylobacter contamination in poultry 
products, (2) evaluate USDA's efforts to assess the effects of these 
actions on the incidence of human illnesses from Salmonella and 
Campylobacter in poultry products, and (3) determine challenges USDA 
faces in reducing these pathogens in poultry products. GAO reviewed 
relevant regulations and documents and interviewed officials from USDA 
and CDC, as well as 11 industry, consumer, and government employee 
stakeholder groups selected based on knowledge of USDA's poultry 
slaughter inspections and food safety. 

What GAO Found: 

Since 2006, the U.S. Department of Agriculture (USDA) has taken a 
number of actions to reduce contamination from Salmonella and 
Campylobacter (disease-causing organisms, i.e., pathogens) in poultry 
(chicken and turkey) products. USDA's actions to reduce these 
pathogens include, for example, tightening existing standards limiting 
the allowable amount of Salmonella contamination in young poultry 
carcasses, implementing the first standards limiting Campylobacter 
contamination in young poultry carcasses in 2011, and developing an 
action plan detailing a priority list of actions, such as developing 
new enforcement strategies, to reduce Salmonella. More recently, in 
August 2014, USDA published its final rule to modernize poultry 
slaughter inspections, which according to the agency, will play a role 
in reducing Salmonella and other poultry pathogen contamination by 
allowing better use of agency resources, among other things. 

To help assess the effects of these actions on the incidence of human 
illness from Salmonella and Campylobacter, USDA conducted research on 
the effects of agency actions to reduce these pathogens and developed 
performance measures for certain poultry products to help monitor 
progress toward agency goals. For example, USDA developed a measure to 
indicate whether agency actions to ensure compliance with the standard 
for Salmonella contamination in young chicken carcasses are helping 
the agency achieve its goal of maximizing domestic compliance with 
food safety policies. However, USDA has not developed measures for 
Salmonella contamination in ground poultry or young turkey carcasses, 
even though standards for such contamination have been in place since 
1996 and 2005, respectively, or for Campylobacter contamination in 
young poultry carcasses. USDA believes it is not appropriate to 
establish measures for ground poultry until the agency has revised 
standards, or for Campylobacter contamination until the agency has 
obtained more information on compliance levels—both of which the 
agency expects to do by the end of 2014. USDA officials stated that 
they will review the agency's strategic plan to determine what 
performance measures, if any, are needed. USDA does not believe a 
measure for young turkey carcasses is needed since historically data 
have shown that plants are meeting the standard but, in calendar year 
2013, two plants did not meet it; USDA officials told GAO that these 
plants are no longer noncompliant. Without performance measures for 
these standards, USDA is not publicly reporting performance 
information and cannot assess the effects of its actions related to 
these standards in meeting the goal of maximizing domestic compliance 
with food safety policies and, ultimately, protecting public health. 

GAO identified several challenges-—based, in part, on the views of 11 
stakeholder groups-—that could hinder USDA's ability to reduce 
contamination in poultry products. For example, contamination of 
poultry products can be affected by practices on poultry farms. To 
help overcome this challenge, the agency developed guidelines in 2010 
on practices for controlling Salmonella and Campylobacter on farms, 
but the guidelines did not include information on the effectiveness of 
each of these practices, consistent with a recommendation from an 
agency advisory committee. USDA did not confirm that it plans to 
include this information in future guidelines. Without providing this 
information in future guidelines, USDA is not fully informing the 
poultry industry of the potential benefits of adopting these practices 
and encouraging their implementation. 

What GAO Recommends: 

GAO recommends that USDA develop additional performance measures for 
Salmonella and Campylobacter contamination in poultry products and 
ensure future guidelines for controlling Salmonella and Campylobacter 
on farms include information on the effectiveness of each practice. 
USDA agreed with GAO's recommendations. 

View [hyperlink, http://www.gao.gov/products/GAO-14-744]. For more 
information, contact J. Alfredo Gómez at (202) 512-3841 or 
gomezj@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

USDA Actions Have Largely Sought to Reduce Salmonella in Poultry, with 
Campylobacter More Recently Addressed: 

In Assessing Its Actions, USDA Does Not Have Measures for All Poultry 
Pathogen Standards to Track Progress in Protecting Human Health and 
Relies on Data with Limitations: 

USDA Faces Several Challenges in Reducing Salmonella and Campylobacter 
in Poultry Products across the Farm-to-Table Continuum: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Case Study Outbreak Summary Tables: 

Appendix III: Comments from the U.S. Department of Agriculture: 

Appendix IV: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: USDA Salmonella Standards for Poultry Products: 

Table 2: USDA Campylobacter Standards for Poultry Products: 

Table 3: 2013 - 2014 Multistate Outbreak of Salmonella Linked to 
Chicken Products Produced in California: 

Table 4: 2012 - 2013 Multistate Outbreak of Salmonella Linked to 
Chicken Products: 

Table 5: 2012 Campylobacter Outbreak Linked to Chicken Livers in 
Vermont: 

Table 6: 2011 Multistate Outbreak of Salmonella Linked to Broiled 
Chicken Liver Products: 

Table 7: 2011 Multistate Outbreak of Salmonella Linked to Ground 
Turkey: 

Figures: 

Figure 1: Farm-to-Table Continuum for Poultry Products: 

Figure 2: Foods Associated with One-Third of Salmonella Outbreaks, 
2004 through 2008: 

Abbreviations: 

APHIS: Animal and Plant Health Inspection Service: 

CDC: Centers for Disease Control and Prevention: 

FDA: Food and Drug Administration: 

FSIS: Food Safety and Inspection Service: 

GPRA: Government Performance and Results Act of 1993, as amended: 

HACCP: Hazard Analysis and Critical Control Point: 

USDA: U.S. Department of Agriculture: 

[End of section] 

United States Government Accountability Office: 
GAO:
441 G St. N.W. 
Washington, DC 20548: 

September 30, 2014: 

The Honorable Kirsten E. Gillibrand: 
Chairman: 
Subcommittee on Livestock, Dairy, Poultry, Marketing and Agriculture 
Security: 
Committee on Agriculture, Nutrition, and Forestry: 
United States Senate: 

The Honorable Dianne Feinstein: 
United States Senate: 

The U.S. Department of Agriculture (USDA) is responsible for ensuring 
the safety of poultry (chicken and turkey) products. According to the 
Department of Health and Human Services' Centers for Disease Control 
and Prevention (CDC), the U.S. food supply remains one of the safest 
in the world.[Footnote 1] Nevertheless, foodborne pathogens,[Footnote 
2] such as Salmonella and Campylobacter, cause human illnesses in the 
United States. Specifically, CDC estimates that Salmonella causes more 
than 1 million illnesses per year and over 20,000 hospitalizations and 
that Campylobacter causes more than 1 million illnesses annually. 
Poultry products contaminated with pathogens--including Salmonella--
cause more deaths than any other food product.[Footnote 3] Pathogens 
in poultry products are especially concerning because Americans 
consume considerably more poultry products than beef or pork. 

The Federal Meat Inspection Act and Poultry Products Inspection Act 
gives USDA overall responsibility for ensuring the safety and 
wholesomeness of meat and poultry products that enter into commerce. 
[Footnote 4] USDA's Food Safety and Inspection Service (FSIS) inspects 
and regulates the production of most domestic poultry products sold 
for human consumption. FSIS proscribes standards for poultry slaughter 
plants under its jurisdiction for limiting Salmonella and 
Campylobacter.[Footnote 5] FSIS also verifies that such plants have 
process controls that are intended to ensure food safety. Congress 
appropriated $1 billion dollars in fiscal year 2014 to FSIS for its 
activities including the food safety inspection program. 

In 2006, FSIS launched an initiative to reduce Salmonella in raw meat 
and poultry products. According to CDC's food safety progress report 
for 2013, however, Salmonella infections in humans have not decreased, 
and the incidence of Campylobacter in humans increased by 13 percent 
in 2013 compared with previous years from 2006 through 2008 based on 
CDC's long-term comparison.[Footnote 6] In this context, we were asked 
to examine USDA's approach to protecting human health by reducing 
these pathogens in poultry products. Our objectives for this report 
were to (1) describe actions USDA has taken since 2006 to reduce 
Salmonella and Campylobacter contamination in poultry products that it 
regulates; (2) evaluate USDA's efforts to assess the effects of these 
actions on the incidence of human illnesses from Salmonella and 
Campylobacter in poultry products; and (3) determine what challenges, 
if any, USDA faces in reducing these pathogens in poultry products. 

To describe actions USDA has taken to reduce Salmonella and 
Campylobacter contamination in poultry products since 2006, we 
reviewed FSIS documents, including Federal Register notices, 
directives, and budget documents, among other things. We also 
interviewed officials at FSIS headquarters about such actions. To 
evaluate USDA's efforts to assess the effects of these actions on the 
incidence of human illnesses from Salmonella and Campylobacter in 
poultry products, we reviewed USDA and FSIS strategic plans, as well 
as USDA annual performance reports to identify relevant performance 
measures, targets, and goals.[Footnote 7] We also evaluated the source 
data, statistical methodology, and results of FSIS research articles 
and a study to determine whether the conclusions drawn about the 
effects of actions taken were adequately supported by the evidence. In 
addition, we interviewed FSIS headquarters officials about the effects 
of actions taken on the incidence of human illnesses from these 
pathogens, and we interviewed FSIS and CDC officials about an 
interagency group to improve data and coordinate analyses. To 
determine any challenges USDA faces in reducing Salmonella and 
Campylobacter in poultry products, we reviewed reports by FSIS, the 
USDA Office of Inspector General, and the National Academy of Sciences 
on FSIS's inspections and management challenges. We interviewed FSIS 
and CDC officials; two food safety researchers we selected based on 
recent academic work; and 12 industry, consumer, and government 
employee stakeholder groups we selected based on our previous 
experience with large national groups with food safety and slaughter 
inspection knowledge from our August 2013 report on poultry and hog 
inspections.[Footnote 8] We conducted a two-stage interview process 
with stakeholder groups. First, we conducted exploratory interviews 
with representatives from 12 stakeholder groups to identify potential 
challenges. After analyzing the results of these interviews to 
identify commonly cited potential challenges, we then conducted 
structured interviews with representatives from 11 of the 12 
stakeholder groups,[Footnote 9] using questions that covered the most 
common potential challenges cited in the exploratory interviews. The 
stakeholders we selected reflected a broad range of stakeholder 
groups, but the views of these stakeholders cannot be generalized to 
other stakeholder groups, and it is possible that their views did not 
include opinions some experts may have. Additionally, we reviewed five 
Salmonella and Campylobacter outbreaks attributed to poultry products 
with the highest number of illnesses to better understand any 
challenges that may have contributed to the outbreaks; these outbreaks 
started during the time period from fiscal year 2011 through fiscal 
year 2013.[Footnote 10] Appendix I contains more detailed information 
on the objectives, scope, and methodology of our review. 

We conducted this performance audit from July 2013 to September 2014 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

Background: 

Salmonella and Campylobacter are bacteria that can cause disease in 
humans and animals. More than 2,500 different types of Salmonella--
known as serotypes[Footnote 11]--exist, and 17 different types of 
Campylobacter exist--known as species.[Footnote 12] Salmonella live in 
the intestinal tracts of humans and animals, while Campylobacter live 
in the intestinal tracts of animals. Some serotypes cause illness in 
humans or in animals. According to CDC officials, Salmonella 
Enteritidis is a common serotype frequently associated with poultry. 
Salmonella Enteritidis causes the most human illnesses among other 
Salmonella serotypes, according to CDC's 2011 Salmonella annual 
report.[Footnote 13] For Campylobacter, most human illness is caused 
by one species of the pathogen, called Campylobacter jejuni. 

Foodborne illness occurs when bacteria or other harmful substances are 
ingested. Poultry is an important source of human Salmonella and 
Campylobacter infections, but Salmonella and Campylobacter 
transmission is not limited to poultry products. Contact with infected 
animals and consumption of contaminated water and foods, including 
milk, eggs, and produce can also transmit the bacteria to humans. 
According to CDC's website, typical symptoms of illness from 
Salmonella or Campylobacter are abdominal cramps, fever, and diarrhea. 
Salmonella infections are more likely than Campylobacter infections to 
lead to bloodstream infections, particularly for infants, the elderly, 
and people with weak immune systems. Salmonella bloodstream infections 
can lead to life-threatening conditions including meningitis. In rare 
cases, both Salmonella and Campylobacter infections can result in long-
term secondary complications such as reactive arthritis,[Footnote 14] 
according to CDC's website. 

To improve its food safety approach, FSIS has moved to an increasingly 
science-based, data-driven, risk-based approach. In 1996, FSIS adopted 
the risk-based Pathogen Reduction: Hazard Analysis and Critical 
Control Point (HACCP) regulations.[Footnote 15] Under the HACCP 
approach, industry--rather than federal inspectors--is responsible for 
(1) identifying food safety hazards, such as fecal material, that are 
reasonably likely to occur and (2) establishing controls that prevent 
or reduce these hazards. As part of this approach, slaughter plants 
must develop plans that identify the point (known as the critical 
control point) where they will take steps to prevent, eliminate, or 
reduce each hazard identified. Under FSIS regulations, all plants must 
also have site-specific standard operating procedures for sanitation. 
FSIS inspectors at slaughter plants routinely check records to verify 
a plant's compliance with those procedures. FSIS also has a 
verification testing program in which FSIS inspectors at slaughter 
plants collect samples of poultry products to determine whether a 
pathogen is present (known as positive rate). Test results help FSIS 
inspectors to verify that plant sanitation procedures are working and 
to identify and assist plants whose process controls may be 
underperforming. 

FSIS coordinates with numerous federal agencies, state agencies, and 
local entities to help ensure a safe poultry product from the farm to 
the consumer (known as the farm-to-table continuum--see figure 1). 
[Footnote 16] For example, on the farm, USDA's Animal and Plant Health 
Inspection Service (APHIS) administers voluntary programs to evaluate 
and certify that poultry are free of certain diseases. FSIS 
coordinates with APHIS to share information when investigating 
foodborne illnesses. FSIS also works with the Department of Health and 
Human Services' Food and Drug Administration (FDA) to, for example, 
approve chemical interventions that slaughter plants use to reduce or 
eliminate Salmonella and Campylobacter.[Footnote 17] FSIS also 
coordinates with CDC and state health departments to respond to 
foodborne illness outbreaks, among other things. 

Figure 1: Farm-to-Table Continuum for Poultry Products: 

[Refer to PDF for image: illustration] 

Farm 1 - hatchery; 
Farm 2 - production house; 
Slaughter plant; 
Retail: Home preparation; Home consumers; 
Restaurant: Restaurant preparation; Restaurant consumers. 

Sources: GAO and Centers for Disease Control and Prevention. GAO-14-
744. 

[End of figure] 

CDC coordinates with FSIS and others on investigations of multistate 
foodborne outbreaks. To identify an outbreak, CDC monitors data 
voluntarily reported from state health departments on cases of 
laboratory confirmed illness and conducts analyses to identify 
elevated rates of disease that may indicate an outbreak, according to 
CDC officials. CDC also analyzes and summarizes foodborne illness 
information on a national basis. According to CDC's website, 
determining the food source of human illness is an important part of 
improving food safety (see figure 2). Moreover, analyses to estimate 
the number or proportion of specific foodborne illness associated with 
each food and nonfood source responsible for causing illness (known as 
foodborne illness source attribution) relies heavily on outbreak data, 
according to CDC officials. 

Figure 2: Foods Associated with One-Third of Salmonella Outbreaks, 
2004 through 2008: 

[Refer to PDF for image: illustration] 

Poultry: 29%; 
Other (Includes sprouts, leafy greens, roots, fish, grains-beans, 
shellfish, oil-sugar, and dairy): 20%; 
Eggs: 18%; 
Vine vegetables, fruits, and nuts: 13%; 
Pork: 12%; 
Beef: 8%. 

Source: GAO figure derived from Centers for Disease Control and 
Prevention data. GAO-14-744. 

[End of figure] 

In addition to these coordination efforts, FSIS was a partner in the 
President's Food Safety Working Group. In 2009, this group recommended 
that FSIS develop new standards to reduce the prevalence of Salmonella 
in poultry and establish a goal of having 90 percent of poultry 
slaughter plants meeting the new Salmonella standards by the end of 
2010.[Footnote 18] In consultation with the working group, FSIS also 
committed to developing Campylobacter standards. 

USDA Actions Have Largely Sought to Reduce Salmonella in Poultry, with 
Campylobacter More Recently Addressed: 

USDA's actions in recent years to reduce Salmonella and Campylobacter 
contamination in poultry products have largely focused on reducing 
Salmonella, with the agency addressing Campylobacter more recently and 
to a lesser degree. 

For Salmonella, USDA Took Actions, Including Revising Existing 
Standards, Promoting Information Sharing, and Developing an Action 
Plan: 

Since 2006, USDA's FSIS has taken a number of actions intended to 
reduce Salmonella contamination in poultry products. These actions 
include revising existing Salmonella standards; taking steps to 
develop standards; promoting enhanced information sharing with 
industry; publicizing noncompliance for chicken slaughter plants not 
meeting the agency's Salmonella standards; developing a Salmonella 
Action Plan; and finalizing a rule to modernize the poultry slaughter 
inspection process. 

Revising Standards: 

In March 2011, FSIS finalized revisions to the agency's Salmonella 
standards for young chicken and turkey carcasses to further limit the 
amount of allowable contamination.[Footnote 19] Specifically, the 
revised standards set the expectation that no more than 7.5 percent of 
a plant's young chicken carcasses (reduced from 20 percent) and 1.7 
percent of a plant's young turkey carcasses (reduced from 19.6 
percent) will be contaminated with Salmonella.[Footnote 20] When FSIS 
revised the standards, it also made corresponding changes to the 
maximum number of samples allowed to test positive for contamination 
during agency testing; the results from the agency's testing are used 
to determine whether or not a plant is in compliance with the agency's 
standards. Specifically, for young chicken carcasses, FSIS now allows 
a maximum of 5 out of 51 samples collected by the agency to test 
positive for Salmonella, compared with the previous allowed positive 
rate of 12 out of 51.[Footnote 21] Similarly, for young turkey 
carcasses, FSIS now allows a maximum positive rate of 4 out of 56 
samples tested by the agency, compared with the previous allowed 
positive rate of 13 out of 56. To help industry meet these revised 
standards, FSIS issued an update to its compliance guideline for 
controlling Salmonella and Campylobacter.[Footnote 22] Specifically, 
the guidelines articulate how industry can meet FSIS expectations 
regarding control of food safety hazards, including control points for 
Salmonella and Campylobacter. 

To verify a chicken slaughter plant's compliance with the agency's 
Salmonella standard, FSIS inspectors collect and test one young 
chicken carcass per day for 51 consecutive days and determine whether 
the number of positive results in that sample set of 51 is above the 
maximum allowed. For turkey slaughter plants, FSIS inspectors collect 
and test one young turkey carcass per day for 56 consecutive days and 
likewise determine whether the number of positive results is above the 
maximum allowed. 

In addition, the agency has recently begun taking steps to strengthen 
its Salmonella standards for ground poultry. More specifically, in 
December 2012, FSIS announced that it planned to perform additional 
Salmonella sampling and testing of ground chicken and ground turkey 
products as part of an agency effort to revise the existing standards 
for those products--which currently set the expectation that no more 
than 44.6 percent of a plant's ground chicken and 49.9 percent of a 
plant's ground turkey will be contaminated with Salmonella.[Footnote 
23] FSIS officials told us the agency has begun performing additional 
Salmonella sampling and testing of these products. According to an 
April 2014 Federal Register notice, FSIS intends to announce and 
request comment in the Federal Register on the proposed revisions to 
the ground poultry standards before the end of fiscal year 2014. 
[Footnote 24] (See table 1 for details regarding FSIS Salmonella 
standards for poultry products.) 

Table 1: USDA Salmonella Standards for Poultry Products: 

Poultry product: Young chicken carcasses; 
Initial standard for maximum percentage of Salmonella 
contamination[A](Year set): 20.0%[C]; (1996); 
Maximum number of positive tests allowed during agency testing[B]: 12 
out of 51 samples tested; 
Current standard for maximum percentage of Salmonella 
contamination[A](year updated): 7.5%; (2011); 
Maximum number of positive tests allowed during agency testing[B]: 5 
out of 51 samples tested. 

Poultry product: Young turkey carcasses; 
Initial standard for maximum percentage of Salmonella 
contamination[A](Year set): 19.6%[D]; (2005); 
Maximum number of positive tests allowed during agency testing[B]: 13 
out of 56 samples tested; 
Current standard for maximum percentage of Salmonella 
contamination[A](year updated): 1.7%; (2011); 
Maximum number of positive tests allowed during agency testing[B]: 4 
out of 56 samples tested. 

Poultry product: Chicken and turkey parts (e.g., thighs, breasts, and 
wings); 
Initial standard for maximum percentage of Salmonella 
contamination[A](Year set): None exists; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable; 
Current standard for maximum percentage of Salmonella 
contamination[A](year updated): None exists[E]; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable. 

Poultry product: Ground chicken; 
Initial standard for maximum percentage of Salmonella 
contamination[A](Year set): 44.6%[C]; (1996); 
Maximum number of positive tests allowed during agency testing[B]: 26 
out of 53 samples tested; 
Current standard for maximum percentage of Salmonella 
contamination[A](year updated): 44.6%; (not updated)[F]; 
Maximum number of positive tests allowed during agency testing[B]: 26 
out of 53 samples tested. 

Poultry product: Ground turkey; 
Initial standard for maximum percentage of Salmonella 
contamination[A](Year set): 49.9%[C]; 
(1996); 
Maximum number of positive tests allowed during agency testing[B]: 29 
out of 53 samples tested; 
Current standard for maximum percentage of Salmonella 
contamination[A](year updated): 49.9%; (not updated)[F]; 
Maximum number of positive tests allowed during agency testing[B]: 29 
out of 53 samples tested. 

Poultry product: Mechanically separated chicken and turkey (used in 
products such as hot dogs and bologna)[G]; 
Initial standard for maximum percentage of Salmonella 
contamination[A](Year set): None exists; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable; 
Current standard for maximum percentage of Salmonella 
contamination[A](year updated): None exists[H]; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable. 

Sources: GAO analysis of USDA Federal Register notices and agency 
documents. GAO-14-744. 

[A] The percentage listed refers to the agency's maximum acceptable 
limit for Salmonella contamination expected at a plant. 

[B] USDA's Food Safety and Inspection Service (FSIS) has set this 
maximum at a number that, when expressed as a percentage of the sample 
set, is higher than the percentage of allowable contamination stated 
in the standards. FSIS has done so in order to minimize the chances of 
falsely finding that plants are not operating within the standards. 

[C] FSIS codified its Salmonella standards for young chicken 
carcasses, ground chicken, and ground turkey in regulations in 1996. 
When the standards for young chicken carcasses, ground chicken, and 
ground turkey were set, FSIS was still collecting data on young turkey 
carcasses. 61 Fed. Reg. 38806, 38867 (July 25, 1996). 

[D] FSIS announced its Salmonella standard for young turkey carcasses 
in February 2005 and based the standard on data collected from July 
1997 to June 1998. 

[E] There are currently no Salmonella standards for raw poultry parts. 
FSIS has begun taking steps to develop a standard for raw chicken 
parts. According to a May 2014 letter from the Secretary of USDA, FSIS 
expects to announce and request comment on the proposed Salmonella 
standard for chicken parts by the end of fiscal year 2014. 

[F] According to an April 2014 Federal Register notice, FSIS intends 
to announce and request comment in the Federal Register on the 
proposed revisions to the standards for ground chicken and ground 
turkey products before the end of fiscal year 2014. 

[G] Mechanically separated poultry is defined in FSIS regulations as 
any product resulting from the mechanical separation and removal of 
most of the bone from the attached skeletal muscle and other tissue of 
poultry carcasses, and as parts of carcasses that have a paste-like 
form and consistency, that may or may not contain skin with attached 
fat. 9 C.F.R. § 381.173(a). 

[H] There are currently no Salmonella standards for mechanically 
separated poultry products. FSIS plans to develop standards for 
mechanically separated poultry by the end of fiscal year 2014. 

[End of table] 

Developing Standards: 

Since 2012, FSIS has taken steps to develop standards for poultry 
parts and mechanically separated poultry, for which there currently 
are no standards. For example, FSIS intends to develop a Salmonella 
standard for raw chicken parts.[Footnote 25] According to FSIS 
officials, the agency is determining which chicken parts, such as 
thighs, breasts, and wings, will be subject to the new standard and 
intends to announce that information in a future Federal Register 
notice. Moreover, in June 2013, FSIS began additional Salmonella 
sampling to determine the estimated prevalence of Salmonella in 
mechanically separated poultry products. FSIS announced in April 2014 
that it would continue to test samples of mechanically separated 
poultry products for Salmonella and analyze the results as part of the 
agency's efforts to develop Salmonella standards for those 
products.[Footnote 26] According to the agency's December 2013 
Salmonella Action Plan, [Footnote 27] FSIS plans to complete a risk 
assessment,[Footnote 28] which will be used to estimate the impact of 
this action, including any potential change in human health, and 
develop new standards for poultry parts and mechanically separated 
poultry by the end of fiscal year 2014. 

Promoting Information Sharing: 

In January 2008, FSIS announced a voluntary initiative--known as the 
Salmonella Initiative Program--to improve controls for Salmonella and 
Campylobacter at young chicken and turkey slaughter plants.[Footnote 
29] Under this program, FSIS granted plants participating in the 
initiative waivers from certain regulatory requirements, such as time 
and temperature chilling requirements for when poultry carcasses are 
placed in a chiller,[Footnote 30] in order to test new procedures, 
equipment, or processing techniques intended to reduce Salmonella 
contamination. In return for such waivers, participating plants agreed 
to collect their own product samples every day during each shift, test 
the samples for common foodborne pathogens including Salmonella and 
Campylobacter, and then share these data with FSIS. In July 2011, FSIS 
announced that plants operating under regulatory waivers previously 
granted by the agency would have to join the Salmonella Initiative 
Program in order to continue operating under those waivers.[Footnote 
31] In September 2013, FSIS completed initial analyses of program data 
showing that participating plants have maintained consistent process 
controls to keep levels of Salmonella within the standards while 
operating under waivers from the time and temperature chilling 
requirements. FSIS reported that, as of June 2014, 158 of 281 (56 
percent) poultry slaughter plants were participating in the program. 

Publicizing Noncompliance on Agency Website: 

In February 2006, to encourage industry to produce safer poultry 
products, FSIS announced in a Federal Register notice that completed 
sample set results would be recorded in one of three categories based 
on the plants' ability to meet existing Salmonella standards.[Footnote 
32] In general, according to an FSIS document on inspection methods, 
those plants in Category 1 are said to be demonstrating consistent 
process controls to meet the existing Salmonella standards and that 
plants in Categories 2 and 3 are not. Specifically, FSIS defines the 
three categories as follows: 

* Category 1 plants have results from their two most recently 
completed sample sets that are at or below half of the existing 
standard, meaning, for example, that for young chicken carcasses, this 
would be 2 positive samples out of a set of 51; 

* Category 2 plants have results from their most recently completed 
sample set that are higher than half of the existing standard but do 
not exceed the standard, meaning, for example, that for young chicken 
carcasses, this would be 3 to 5 positive samples out of a set of 51; 
and: 

* Category 3 plants have results from their most recently completed 
sample set that exceed the existing standard for Salmonella, meaning, 
for example, that for young chicken carcasses, this would be 6 or more 
positive samples out of 51. 

In the February 2006 Federal Register notice, FSIS also announced that 
it would use the categories to determine the frequency of the agency's 
Salmonella verification testing. According to FSIS's sampling 
methodology, Category 1 plants are tested at least once every 2 years; 
Category 2 plants are scheduled for testing at least once a year until 
their category changes (e.g., a plant improves to Category 1); and 
Category 3 plants are scheduled for testing as close to continuously 
as possible until they produce better results and their category 
changes. 

In addition, in a January 2008 Federal Register notice, FSIS announced 
that it planned to begin publishing on the agency's website the 
results for young chicken slaughter plants that were inconsistent in 
complying with existing Salmonella standards, stating that it believed 
making such information available to the public would provide an 
incentive to industry to attain "consistent, good control for 
Salmonella."[Footnote 33] The agency began publishing this information 
on its website in March 2008. At present, FSIS publishes the names of 
young chicken slaughter plants that are in Category 3.[Footnote 34] 
For example, in August 2014, FSIS published on its website the names 
of seven young chicken slaughter plants found to be in Category 3. 
Under FSIS policy, the names of young turkey slaughter plants not in 
compliance with the agency's young turkey standard can also be 
published on FSIS's website. As of August 2014, there are no names of 
turkey plants posted on the website because no turkey plants were in 
Category 3 based on recent testing. 

Developing a Salmonella Action Plan: 

FSIS's Salmonella Action Plan, issued in December 2013, details a 
priority list of actions the agency plans to undertake as part of its 
continued efforts to address Salmonella in poultry products. According 
to the plan, FSIS intends to, among other things, (1) conduct food 
safety assessments at plants that produce ground and mechanically 
separated poultry products by the end of fiscal year 2014;[Footnote 
35] (2) consider modifying the way it publishes the category status of 
poultry plants, such as by publishing on its website the names of 
plants in Category 1 and Category 2, in addition to those in Category 
3, by the end of fiscal year 2014; (3) develop new enforcement 
strategies that take into account plants' compliance history and 
Salmonella category under the standards, among other things, which, 
according to the plan, will take over a year to accomplish; and (4) 
host a meeting with APHIS and other stakeholders to focus on poultry 
farm practices that could help decrease Salmonella contamination on 
FSIS-regulated poultry products and use the information gathered to 
inform best practice guidelines, requiring the completion of 
additional actions prior to the meeting.[Footnote 36] 

Finalizing a Rule to Modernize Poultry Slaughter Inspection: 

On August 21, 2014, FSIS published its final rule to amend the 
agency's poultry slaughter inspection process.[Footnote 37] Young 
chicken and turkey slaughter plants may choose to operate under the 
new poultry inspection system included in the rule or may continue to 
operate under the current inspection system. According to FSIS's final 
rule, modernizing poultry slaughter inspections will play a role in 
reducing Salmonella and other poultry pathogen contamination. 
Currently, FSIS inspectors conduct a variety of duties at positions on 
and off the slaughter line. For example, FSIS inspectors positioned on 
the line conduct inspections of every poultry carcass and its parts 
for defects, and inspectors working off the line move freely about the 
plant and collect samples of carcasses to test for pathogens (e.g., 
Salmonella); perform food safety checks, such as verifying that 
carcasses are free of fecal material; and ensure that carcasses comply 
with the agency's food quality standards for defects such as bruises 
on chickens, which do not affect food safety. For those poultry 
slaughter plants that choose to operate under the new poultry 
inspection system, plant employees would assume more responsibility 
for conducting the types of activities currently performed by FSIS 
inspectors on the slaughter line. For example, plant personnel would 
be responsible for identifying defects in carcasses, taking corrective 
actions if the defects can be corrected through trimming, and 
condemning unacceptable carcasses as part of on-line inspections. 
According to FSIS's final rule, these changes will allow FSIS to 
assign fewer inspectors for on-line inspections and allow inspectors 
to conduct more off-line inspections in plants operating under the new 
poultry inspection system. Moreover, according to the rule, the new 
poultry inspection system may facilitate reduction of pathogen levels 
in poultry by permitting FSIS to conduct more food-safety-related off-
line inspection activities and allowing better use of FSIS inspection 
resources, among other things. In particular, in July 2014, FSIS 
issued an updated risk assessment that estimated that there would be a 
reduction of 3,980 Salmonella illnesses attributable to young chicken 
and turkey plants combined.[Footnote 38] 

Other Actions: 

FSIS has also taken other actions that are intended to help reduce 
Salmonella contamination in poultry products. For example, in light of 
several Salmonella outbreaks associated with the consumption of ground 
turkey products, FSIS announced in a December 2012 Federal Register 
notice that it was requiring all poultry plants producing ground or 
mechanically separated chicken and turkey products to reassess their 
HACCP plans.[Footnote 39] In May 2013, FSIS instructed its inspectors 
at plants producing these types of products to verify that the plants 
had reassessed their HACCP plans. In addition, FSIS announced that it 
would expand its Salmonella verification testing program--which 
previously was limited to ground chicken and ground turkey--to include 
plants producing all forms of ground and mechanically separated 
chicken and turkey products. 

For Campylobacter, USDA Established Standards and Began Considering 
Further Actions: 

FSIS began taking actions specifically aimed at reducing Campylobacter 
contamination in poultry products in 2011. Specifically, in July 2011, 
FSIS implemented the first standards that define the amount of 
allowable contamination from Campylobacter for young chicken and 
turkey carcasses.[Footnote 40] Under the new standards, FSIS set the 
expectation that no more than 10.4 percent of a plant's young chicken 
carcasses and 0.79 percent of a plant's young turkey carcasses will be 
contaminated with Campylobacter. In addition, FSIS established maximum 
numbers of positive samples allowed during agency testing: 8 out of 51 
samples for young chicken carcasses and 3 out of 56 samples for young 
turkey carcasses (see table 2). 

Table 2: USDA Campylobacter Standards for Poultry Products: 

Poultry product: Young chicken carcasses; 
Current standard for maximum percentage of Campylobacter 
contamination[A]: (year set): 10.4%; (2011); 
Maximum number of positive tests allowed during agency testing[B]: 8 
out of 51 samples tested. 

Poultry product: Young turkey carcasses; 
Current standard for maximum percentage of Campylobacter 
contamination[A]: (year set): 0.79%; (2011); 
Maximum number of positive tests allowed during agency testing[B]: 3 
out of 56 samples tested. 

Poultry product: Chicken and turkey parts (e.g., thighs, breasts, and 
wings); 
Current standard for maximum percentage of Campylobacter 
contamination[A]: (year set): None exists[C]; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable. 

Poultry product: Ground chicken; 
Current standard for maximum percentage of Campylobacter 
contamination[A]: (year set): None exists[D]; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable. 

Poultry product: Ground turkey; 
Current standard for maximum percentage of Campylobacter 
contamination[A]: (year set): None exists[D]; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable. 

Poultry product: Mechanically separated chicken and turkey (used in 
products such as hot dogs and bologna); 
Current standard for maximum percentage of Campylobacter 
contamination[A]: (year set): None exists[D]; 
Maximum number of positive tests allowed during agency testing[B]: Not 
applicable. 

Sources: GAO analysis of USDA Federal Register notices and agency 
documents. GAO-14-744. 

[A] The percentage listed refers to the agency's maximum acceptable 
limit for Campylobacter contamination expected at a plant. 

[B] USDA's Food Safety and Inspection Service (FSIS) has set this 
maximum at a number that, when expressed as a percentage of the sample 
set, is higher than the percentage of allowable contamination stated 
in the standards. FSIS has done so in order to minimize the chances of 
falsely finding that plants are not operating within the standards. 

[C] According to a May 2014 letter from the Secretary of USDA, FSIS is 
working on developing a standard for Campylobacter for raw chicken 
parts and plans to announce the standard in the Federal Register by 
the end of fiscal year 2014. 

[D] According to a May 2014 letter from the Secretary of USDA, FSIS is 
investigating the appropriateness of Campylobacter standards for 
ground and mechanically separated poultry products. 

[End of table] 

FSIS is also considering other actions to address Campylobacter 
contamination in poultry products. According to FSIS officials, the 
agency plans to announce and request comment on a proposed standard 
for Campylobacter for raw chicken parts in the Federal Register by the 
end of fiscal year 2014. Furthermore, according to a May 2014 letter 
from the Secretary of USDA, FSIS is investigating the appropriateness 
of Campylobacter standards for ground and mechanically separated 
poultry products. 

FSIS did not take actions specifically aimed at reducing Campylobacter 
contamination before 2011, in part, because the agency believed its 
actions to reduce Salmonella would also reduce other pathogens such as 
Campylobacter. Moreover, according to documents from FSIS and the 
National Advisory Committee on Microbiological Criteria for Foods, it 
was not until 2005 that the agency began using a less time-consuming 
and more reliable sampling method for determining the presence of 
Campylobacter on poultry products. In addition, the agency was 
concerned about the observed increases in Salmonella rates, such as an 
increase in the percentage of positive Salmonella results during 
agency testing from 11.5 percent in 2002 to 16.3 percent in 2005. 

In Assessing Its Actions, USDA Does Not Have Measures for All Poultry 
Pathogen Standards to Track Progress in Protecting Human Health and 
Relies on Data with Limitations: 

To help assess the effects of the agency's actions on the incidence of 
human illnesses from Salmonella and Campylobacter contamination in 
poultry products, FSIS has developed performance measures and 
conducted research, but these efforts fall short in two ways: (1) the 
agency did not establish performance measures for certain commonly 
consumed poultry products or Campylobacter and (2) the agency has 
relied on data with limitations that affect their usefulness. 

FSIS Did Not Establish Performance Measures for Certain Commonly 
Consumed Poultry Products or Campylobacter: 

First, consistent with requirements of the Government Performance and 
Results Act of 1993 (GPRA) to measure performance toward the 
achievement of agency strategic goals, FSIS established a performance 
measure for Salmonella contamination in young chicken carcasses. We 
previously concluded that performance measures, which typically have 
numerical targets, are important management tools that help an agency 
monitor and report progress toward its goals.[Footnote 41] FSIS's 
Salmonella measure indicates whether agency actions to ensure 
compliance with the applicable standard are helping the agency meet 
its goal. The agency's goal is to maximize domestic and international 
compliance with food safety policies, which aligns with USDA's 
objective of protecting public health by ensuring food is safe. 
[Footnote 42] Specifically, the measure tracks the percentage of 
poultry slaughter plants complying with the Salmonella standard for 
young chicken carcasses; the agency then compares this percentage to a 
target level of compliance to monitor progress in meeting its goal. 
For example, FSIS set the fiscal year 2013 target at 91 percent 
compliance and reported that 90 percent of plants complied with the 
standard.[Footnote 43] According to FSIS's 2011-2016 strategic plan 
and a National Academy of Sciences report, verification testing data 
available on the exposure of the public to raw poultry products 
contaminated with Salmonella, including young chicken carcasses, 
provide a reasonable proxy for the relative risk associated with those 
products.[Footnote 44] 

However, the agency has not established similar performance measures 
and targets for other types of commonly consumed poultry products for 
which it has established Salmonella standards--that is, young turkey 
carcasses, ground chicken, and ground turkey. The agency has not 
established performance measures even though the standards for young 
turkey carcasses have been in place since 2005, with a revision in 
2011, and the standards for ground chicken and ground turkey have been 
in place for more than a decade. The majority of poultry that industry 
markets and Americans consume is ready-to-eat and further processed 
products, such as ground poultry and poultry parts, according to the 
United States International Trade Commission 2014 report. Salmonella-
contaminated ground poultry and poultry parts put consumers at greater 
risk of becoming ill than whole poultry carcasses because these 
products generally are more likely to be contaminated. As previously 
noted, FSIS is developing a Salmonella standard for raw chicken parts; 
therefore, the agency has not developed corresponding performance 
measures and targets. Moreover, FSIS has not established performance 
measures for Campylobacter contamination, even though it implemented 
standards for this pathogen in 2011. According to FSIS officials, at 
the time that the 2011 to 2016 strategic plan was written, FSIS had 
only recently implemented the first ever Campylobacter standards for 
poultry, and so the agency did not have a basis to create a 
performance measure for that pathogen. FSIS officials also stated 
that, in the absence of performance measures, the agency routinely 
collects and reviews data on individual plants' compliance with 
Salmonella and Campylobacter standards and reports quarterly the 
results of pathogen testing on its website. According to FSIS 
officials, these data are sent to FSIS leadership, and trends are 
highlighted, as appropriate. 

FSIS officials agreed that performance measures and targets should be 
developed for additional poultry products and stated that the agency 
will review its current strategic plan to determine what further 
updates are needed, namely new performance measures for ground chicken 
and ground turkey. As previously mentioned, FSIS established 
Salmonella standards for ground chicken and ground turkey in 1996, and 
the agency has yet to establish corresponding performance measures and 
targets. According to FSIS officials, it is not appropriate for the 
agency to set a performance measure or target for Salmonella 
contamination in ground chicken or ground turkey until it finishes 
revising the standards for those products, which it expects to do by 
the end of fiscal year 2014. In addition, for Campylobacter 
contamination in young chicken and turkey carcasses, FSIS officials 
told us that the agency has not developed performance measures in part 
because it is still in the process of developing plant categories, 
similar to those for Salmonella,[Footnote 45] to determine levels of 
plant compliance with Campylobacter standards. The agency will publish 
Campylobacter categories before December 31, 2014, according to FSIS's 
fiscal year 2014 annual performance plan. FSIS officials told us that 
the agency has not developed a performance measure for determining the 
agency's success in controlling Salmonella in young turkey carcasses 
because testing data have shown that young turkey plants where the 
agency has routinely collected and reviewed data are meeting the 
standard. However, FSIS stated in its fourth quarter calendar year 
2013 report on pathogen testing that 2 out of 35 (about 6 percent) of 
the turkey slaughter plants were in Category 3, meaning these plants 
did not meet the standard. In July 2014, FSIS officials told us these 
plants are no longer in Category 3 based on recent testing. However, 
without performance measures and targets for compliance with standards 
for these pathogens in commonly consumed poultry products, FSIS cannot 
quantitatively gauge its progress in assessing the effects of its 
actions related to these standards toward meeting the agency's goal of 
maximizing domestic compliance with food safety policies and, 
ultimately, protecting public health.[Footnote 46] Performance 
measures and targets are reported in the agency's strategic and annual 
plans but, in the absence of such measures and targets, performance 
information such as the trends supplied to FSIS leadership, is not 
being publicly reported. Without publicly reporting such information, 
FSIS loses the opportunity to enhance transparency by providing this 
information to the public and Congress about its progress in meeting 
this important goal, potentially limiting oversight and 
accountability.[Footnote 47] 

In addition to the performance measure for young chicken carcasses, 
FSIS established, in 2009, an "all-illness" performance measure to 
evaluate its efforts to reduce foodborne human illness resulting from 
consumption of FSIS-regulated products contaminated with Salmonella 
and two other pathogens.[Footnote 48] According to FSIS officials, the 
all-illness measure is an estimate of the number of illnesses from 
these pathogens resulting from the consumption of all FSIS-regulated 
products--including meat, poultry, and processed egg products. The 
officials said FSIS develops this estimate using CDC laboratory-
confirmed illness data, CDC outbreak data, U.S. census data, and 
foodborne illness estimation data from a CDC peer-reviewed 
publication. The all-illness measure has an associated target level of 
performance, set by the agency, for the maximum number of human 
illnesses from these pathogens attributed to FSIS-regulated products. 
[Footnote 49] For example, FSIS set the fiscal year 2013 target at 
394,770 human illnesses from the three pathogens, including 
Salmonella. USDA reported in 2013 that FSIS did not meet its target; 
there were 427,171 human illnesses. As part of the all-illness 
measure, FSIS also set individual targets for the maximum number of 
human illnesses from each of the three pathogens attributed to FSIS-
regulated products, including a Salmonella target. The annual report 
did not break out the results for each of the three pathogens, but the 
agency's fiscal year 2013 year-in-review report showed that FSIS did 
not meet the Salmonella target specifically. According to FSIS 
officials, historically the agency has not met the Salmonella target 
for the all-illness measure. However, recent actions FSIS has taken 
and plans to take are intended to address human illnesses from 
Salmonella attributed to FSIS regulated products, according to the 
agency's Salmonella Action Plan. 

FSIS Has Relied on Data with Limitations That Affect Their Usefulness: 

According to USDA's National Advisory Committee on Microbiological 
Criteria for Foods, which provides impartial, scientific advice to 
federal food safety agencies,[Footnote 50] the impact of FSIS's 
regulatory activities on the incidence of human illnesses from 
pathogens cannot be measured directly because of limitations in the 
foodborne illness attribution data the agency uses.[Footnote 51] For 
example, the outbreak data available from CDC depend on voluntary 
reporting of illnesses and do not always identify the food product 
that caused an outbreak; we have previously found limitations in these 
data, including delayed reporting and incompleteness.[Footnote 52] 
According to FSIS officials, as with most performance measures that 
seek to evaluate human health outcomes, the all-illness measure is 
subject to limitations based on the availability of data and the 
challenges in capturing accurate foodborne illness attribution. In 
commenting on this report, USDA agreed that there are limitations in 
the data but stated the agency uses the best available data. FSIS is 
considering including Campylobacter in the all-illness measure as 
well, but there is no broadly accepted estimate for the proportion of 
illnesses attributed to Campylobacter, according to FSIS officials. 

In 2011, FSIS, CDC, and FDA formed an interagency group to improve 
food safety data and coordinate analyses. Some of the group's efforts 
involve identifying links between contamination of poultry products 
and human illness, among other things.[Footnote 53] For example, FSIS, 
CDC, and FDA are working collaboratively to perform a detailed 
analysis of data on Salmonella outbreaks to better estimate the 
proportion of human illnesses caused by different food sources, 
including poultry products. According to FSIS officials, this effort 
will assist in better estimating the proportion of Salmonella 
illnesses associated with poultry products, and the agencies plan to 
present results from ongoing projects in 2015. 

In addition to performance measures, FSIS conducted research to assess 
the effects of its actions on the incidence of human illnesses from 
the consumption of poultry products, but this research also has data 
limitations. For example, in February 2012, after revising its 
Salmonella standards for young chicken carcasses, FSIS completed 
research evaluating whether reductions of Salmonella contamination 
from young chicken carcasses across the industry would offer public 
benefits, in the form of reduced human illness rates.[Footnote 54] 
However, in attempting to evaluate the effect of reductions of 
Salmonella contamination across the industry, FSIS relied on its 
verification testing data from individual plants, which the agency 
later concluded in April 2012 cannot be used to estimate prevalence 
across the industry because the agency does not randomly select plants 
for verification testing, among other things.[Footnote 55] FSIS also 
used CDC outbreak and individual illness case data in its research to 
identify the number of human illness from Salmonella contamination 
but, as we mentioned above, these data do not always distinguish 
illnesses derived from poultry products specifically. 

While FSIS's research efforts are a positive step, data limitations 
make it difficult to directly correlate agency actions to reductions 
in the rates of human illness from poultry products contaminated with 
Salmonella. To help address these limitations, FSIS has taken steps, 
such as developing a statistical model in 2012, to estimate the 
reduction in human illnesses from revised Salmonella and newly created 
Campylobacter standards.[Footnote 56] More importantly, in 2013, FSIS 
created a new testing approach for ground poultry to estimate 
prevalence of Salmonella, among other things. According to FSIS 
officials, the new verification testing program includes continuous 
weekly sampling and testing at all poultry slaughter plants producing 
raw ground poultry and increases the sensitivity of analysis so that 
lower levels of contamination can be detected. FSIS officials told us 
that the agency plans to expand this approach to other poultry 
products. According to FSIS's working group on prevalence estimates, 
testing for prevalence is necessary in order for the agency to 
effectively measure or understand how contamination rates change over 
time; set standards; develop targeted interventions; and measure the 
agency's performance toward meeting FSIS long-term strategic goals. 
The new testing approach using continuous sampling affords a more 
direct measure of prevalence across the industry, according to FSIS 
officials. 

USDA Faces Several Challenges in Reducing Salmonella and Campylobacter 
in Poultry Products across the Farm-to-Table Continuum: 

We identified several challenges that FSIS faces in reducing 
Salmonella and Campylobacter contamination in poultry products and one 
potential challenge. These include limited control outside of 
slaughter plants, pathogens not designated as hazards, limited 
enforcement authority,[Footnote 57] absence of mandatory recall 
authority, outdated or nonexistent standards, insufficient prevalence 
estimates, the complex nature of Salmonella, and limited Campylobacter 
research and testing. We identified these challenges and the potential 
challenge based on our analyses and the views of representatives of 11 
stakeholder groups, a number of academic researchers, and FSIS 
officials; the stakeholder groups representing consumers and those 
representing industry generally had differing views. 

Limited Control Outside of Plants: 

FSIS faces a challenge in reducing Salmonella and Campylobacter 
contamination in poultry products outside of slaughter plants because 
the agency does not have regulatory jurisdiction over (1) farm 
practices to reduce contamination in live poultry before they reach a 
plant or (2) some factors that may affect contamination of poultry 
products once they leave a plant. According to FSIS officials, they 
would like to address on-the-farm problems, but the agency is limited 
in its activities related to farms because its jurisdiction starts 
when products enter slaughter plants. In 2010, FSIS published updated 
compliance guidelines that detail, among other things, several on-farm 
practices to reduce Salmonella and Campylobacter in live poultry. For 
example, the guidelines recommend that farms test water to make sure 
it is free of pathogens and ensure water stations are free of leaks. 
As of July 2014, FSIS's compliance guidelines do not discuss the 
effectiveness of each recommended practice to reduce pathogens in live 
poultry. Thus, FSIS has not provided complete information to the 
poultry industry about the potential benefits of adopting certain 
practices. In contrast, when the agency developed guidelines for on-
farm practices for reducing E. coli O157: H7 in beef cattle, it 
described several practices and their effectiveness. 

In 2011, USDA's National Advisory Committee on Meat and Poultry 
Inspection recommended that (1) FSIS coordinate with APHIS, among 
other agencies, to develop best practices on the farm and (2) develop 
compliance guidelines for livestock and poultry producers, including 
information on the effectiveness of the practices in controlling 
pathogens.[Footnote 58] Representatives from two consumer groups told 
us that because the poultry industry is vertically integrated--meaning 
that individual poultry companies own or contract for all phases of 
production and processing--it is well suited to implement on-farm best 
practices to help ensure healthier birds prior to slaughter. As 
previously mentioned, according to FSIS's December 2013 Salmonella 
Action Plan, the agency will continue to work with industry to 
identify on-farm best practices, host a meeting with APHIS and other 
stakeholders to focus on on-farm practices that could help decrease 
Salmonella contamination on FSIS-regulated poultry products, and use 
the information gathered from these actions to inform future policies 
and compliance guidelines. FSIS officials told us that the agency is 
currently working with stakeholders such as FDA and APHIS to gather 
information about on-farm practices. However, even with the planned 
actions identified in the agency's Salmonella Action Plan, it remains 
unclear whether FSIS intends to incorporate information on the 
effectiveness of all practices in the guidelines as the National 
Advisory Committee on Meat and Poultry Inspection recommended. FSIS 
officials told us that the agency will publish another revision of the 
compliance guidelines by the end of calendar year 2014 but did not 
respond when asked directly about whether they would incorporate such 
information. Without providing information on the effectiveness of 
these practices in future guidelines, FSIS is not fully informing 
industry of the potential benefits of adopting them to encourage 
implementation of recommended practices. 

In addition to not having regulatory jurisdiction over reducing 
contamination in live poultry before they reach a plant, once poultry 
products leave the plant,[Footnote 59] factors beyond FSIS's control 
may affect contamination of poultry products. For example, cross-
contamination[Footnote 60] from poultry products can occur at retail 
establishments, in restaurants, and in consumers' homes, according to 
a food safety researcher we interviewed. According to FSIS officials, 
the agency has been aggressive in educating consumers on the 
importance of safe handling of raw poultry products, such as through 
an advertising campaign and changes to the safe handling label. In 
2014, FSIS proposed enhancing the safe food handling label for poultry 
products packaged for consumers to include updated information on 
proper handling. 

Pathogens Not Designated as Hazards: 

FSIS faces a challenge in reducing Salmonella and Campylobacter 
contamination in poultry products when these pathogens are not 
designated as hazards by slaughter plants. Under the HACCP approach, 
plants have discretion about whether to include Salmonella or 
Campylobacter as a hazard "reasonably likely to occur" in their HACCP 
plans and develop mitigation strategies to reduce these pathogens in 
poultry products.[Footnote 61] Representatives from all six consumer 
groups we interviewed stated that this approach is not adequate to 
keep contamination levels below standards. In contrast, 
representatives from all four industry groups we interviewed stated 
that the current voluntary designation is sufficient. Moreover, 
representatives from two industry groups we interviewed said that 
HACCP has proven itself to work well in lowering contamination in 
poultry products. In the five outbreaks we reviewed, FSIS found that 
plants linked to three of the four Salmonella outbreaks had inadequate 
HACCP plans because they had not appropriately designated Salmonella 
as a hazard reasonably likely to occur. For example, in 2011, FSIS 
investigated a company's plant linked to an outbreak from ground 
turkey and determined that the plant was in violation of federal 
regulation because it failed to identify hazards reasonably likely to 
occur, namely Salmonella. In a 2013 outbreak we reviewed, FSIS issued 
a "Notice of Intended Enforcement" action, which warns the plant 
before the initiation of a specific enforcement action, based on the 
company's inability to support why it had not designated Salmonella as 
a hazard reasonably likely to occur in its HACCP plan. FSIS officials 
we interviewed believe the agency should require plants to identify 
Salmonella and Campylobacter in their HACCP plans as hazards 
reasonably likely to occur.[Footnote 62] FSIS's final rule for 
modernizing poultry slaughter inspection requires plants to develop, 
implement, and maintain written procedures to prevent contamination by 
enteric pathogens,[Footnote 63] such as Salmonella and Campylobacter. 

Limited Enforcement Authority: 

FSIS faces a potential challenge in reducing Salmonella contamination 
in poultry products because, according to the agency and some 
stakeholder groups, its authority to enforce its Salmonella standards 
is limited for two reasons: (1) a federal court ruling and (2) FSIS 
has not classified Salmonella as an adulterant in raw poultry 
products.[Footnote 64] 

First, in 2000, a federal court ruled that FSIS could not withdraw 
inspectors, which would effectively shut down the plant, based solely 
on a plant's failure to meet Salmonella standards. A federal appeals 
court upheld the decision in 2001.[Footnote 65] Subsequently, the 
agency adopted the position that the court ruling did not affect its 
ability to use the standards as part of verifying a plant's sanitation 
and HACCP plans. For example, after a plant fails Salmonella testing 
for its first sample set, FSIS can require a reassessment of the 
plant's HACCP plan and then conduct a food safety assessment 
(evaluation of a plant's food safety system);[Footnote 66] conduct 
additional sampling; or issue a "Notice of Intended Enforcement" 
action, according to FSIS officials. FSIS can also condemn products 
that are contaminated with filth (or otherwise adulterated) or 
mislabeled, or it can condemn parts of products,[Footnote 67] and 
detain them so they cannot progress down the marketing chain. Even 
with these tools, representatives from four out of six consumer groups 
we interviewed told us that the agency does not have sufficient 
authority to ensure plants comply with FSIS's standards because FSIS 
cannot shut down plants when they fail Salmonella standards alone. 
Representatives of all four industry groups we interviewed disagreed 
and stated that FSIS has sufficient authority to ensure plants comply 
with standards because the agency has broad statutory authority and 
oversight. 

Second, because FSIS has not classified Salmonella as an adulterant in 
raw poultry products, products contaminated with this pathogen 
generally are permitted to enter commerce.[Footnote 68] However, 
according to FSIS officials, the agency can consider raw poultry 
products contaminated with Salmonella as adulterated on a case-by-case 
basis--for example, during a voluntary recall, as discussed later in 
this report. Representatives from five out of six of the consumer 
groups we interviewed said they believe that some serotypes of 
Salmonella should be declared an adulterant, such as those with 
specific antibiotic-resistance patterns. According to CDC officials, 
antibiotic resistance can be associated with a higher risk of 
hospitalization in infected individuals. For example, in two of the 
four Salmonella outbreaks we reviewed, ill persons were hospitalized 
twice as frequently as is normally seen in Salmonella outbreaks, 
according to CDC officials. Since 2013, more than a dozen consumer 
groups have supported a petition for FSIS to declare specific 
antibiotic-resistant serotypes of Salmonella as adulterants when found 
in poultry. In July 2014, FSIS denied one of the consumer group's 
petitions to have antibiotic-resistant Salmonella declared an 
adulterant.[Footnote 69] FSIS officials told us that they have found 
no conclusive scientific evidence that antibiotic-resistant 
strains[Footnote 70] of Salmonella or Campylobacter have a greater 
resistance to interventions currently used in FSIS-inspected poultry 
plants, but the agency continues to review the relevant scientific 
evidence to identify any potential challenges that these serotypes may 
present to public health. 

Representatives from a government employee stakeholder group we 
interviewed said that rather than classifying all of Salmonella an 
adulterant in raw poultry products, FSIS should consider the top 
Salmonella serotypes causing most human illnesses. For example, the 
representatives said that FSIS should consider declaring a narrow 
range of Salmonella serotypes in select raw poultry products as 
adulterants similar to the classification of E. coli O157: H7 as an 
adulterant in beef. CDC has identified three Salmonella serotypes in 
poultry associated with causing the highest number of human illnesses: 
Salmonella Enteritidis, Salmonella Typhimurium, and Salmonella 
Heidelberg. According to FSIS's risk-based inspection protocols, FSIS 
considers the top Salmonella serotypes identified through CDC data 
when ranking plants to determine the frequency of verification testing 
and inspections. Moreover, according to FSIS documents, the agency 
provides plants with data on which Salmonella serotypes were 
identified through verification testing. FSIS officials told us prior 
court cases have set a precedent that the presence of Salmonella in 
raw poultry products is not sufficient to declare the pathogen an 
adulterant because Salmonella can be killed through proper cooking. 

Representatives from all four industry groups we interviewed disagreed 
that any serotypes of Salmonella should be classified as adulterants 
for several reasons. Representatives from two industry groups told us 
rapid identification of the serotype is not available and that it can 
take several weeks for FSIS to identify specific serotypes from 
positive Salmonella verification test results. Representatives from 
another industry group we spoke with said Salmonella is already 
classified as an adulterant in fully cooked poultry products and that, 
for raw poultry products, FSIS includes instructions on proper 
handling and thorough cooking to prevent cross-contamination and 
eliminate the pathogen. In contrast with Salmonella, Campylobacter has 
received less attention from FSIS and stakeholder groups, in part 
because the pathogen is not frequently associated with outbreaks, 
making it difficult to attribute illnesses to this pathogen. 

No Mandatory Recall Authority: 

FSIS faces a challenge in reducing Salmonella and Campylobacter 
contamination in poultry products because it does not have mandatory 
food recall authority similar to that of FDA. In 2011, Congress passed 
the FDA Food Safety Modernization Act, giving FDA mandatory recall 
authority. We recommended in October 2004 that Congress should 
consider legislation to increase FSIS's authority to include mandatory 
recalls, but the agency continues not to have such enforcement 
authority.[Footnote 71] Instead, FSIS can issue public health alerts 
[Footnote 72] or request voluntary recalls,[Footnote 73] among other 
actions, to protect human health from potentially contaminated meat 
and poultry products. Before requesting a voluntary recall, FSIS must 
gather sufficient evidence through its investigation and determine 
that a product is adulterated or mislabeled, among other things. 
[Footnote 74] According to FSIS officials, the agency requests a 
voluntary recall when the agency links a product to an ill person and 
obtains specific information on the source of the product (plant), 
product type, production date, and product distribution. In the five 
outbreaks we reviewed, three of the five companies voluntarily 
recalled or stopped distributing poultry products implicated in 
outbreaks before FSIS requested a recall. For the fourth outbreak, in 
July 2014, FSIS requested that Foster Farms conduct a voluntary recall 
of select chicken parts involved in an outbreak, which had been 
ongoing since March 2013, after definitively linking the product to an 
ill person; the company recalled the products. By contrast, in the 
fifth outbreak, CDC reported that collaborative investigative efforts 
by local, state, and federal officials indicated that Foster Farms 
chicken products were the most likely source of an outbreak of 
Salmonella Heidelberg that took place from June 2012 until May 2013. 
However, FSIS did not request, and Foster Farms did not conduct, a 
voluntary recall because the agency was unable to definitively link 
the company's product to an ill person (see appendix II). FSIS 
officials told us that, rather than focusing on the lack of mandatory 
recall authority, it is more productive to work aggressively with the 
tools they have. For example, FSIS officials told us that withdrawing 
inspectors or withholding the agency's mark of inspection, thus 
preventing poultry products from entering commerce, can be as 
effective and faster tool for keeping unsafe food from the marketplace 
than FDA's recall authority.[Footnote 75] 

Outdated or Nonexistent Standards: 

Side bar: 
Poultry Production and Consumption: 

U.S. consumption of poultry products (chicken and turkey) is 
considerably higher than pork or beef, but less than total red meat 
consumption. According to USDA's Economic Research Service, the U.S. 
poultry industry is the world's largest producer and the second-
largest exporter of poultry. Before the 1970s, poultry was largely 
retailed on a "whole bird" basis. For example, chicken meat sold as 
parts, such as wings and breasts, was a small component of the 
domestic U.S. market. Chicken meat retailed as parts came about 
largely as a consequence of the inspection process at slaughter 
plants; that is, the carcass of a whole raw chicken that failed 
inspection would be cut to remove the part of the whole bird that 
caused inspection failure. The remainder of the bird was then further 
broken down and marketed as chicken parts. Market changes began to 
indicate to processors that consumers preferred particular chicken 
parts rather than whole birds. To satisfy consumers, processors began 
to break whole chickens into parts for retail sale. Trays of whole 
birds broken into constituent parts evolved into packages or bags of 
drumsticks, wings, and breasts, among other products. 

[Refer to PDF for image: illustration of constituent parts] 

Source: USDA Economic Research Service; USDA (graphic). GAO-14-744. 

[End of side bar] 

FSIS faces a challenge in reducing Salmonella and Campylobacter 
contamination in poultry products because of outdated or nonexistent 
standards. FSIS does not revise and develop standards frequently 
enough to reflect changes in industry practices and poultry 
consumption patterns. For example, it has taken FSIS nearly 2 decades 
to begin revising its Salmonella standard for ground chicken. The FSIS 
standard established in 1996 for ground chicken set an expectation 
that up to 44.6 percent of a plant's production could be contaminated 
with Salmonella without the plant being required to take corrective 
action. According to industry groups we spoke with, since the 
implementation of the standard, industry developed newer technologies 
to reduce contamination below the levels in the standard. FSIS 
officials agreed and stated that the majority of poultry slaughter 
plants perform at levels better than the standard. 

In addition, as noted earlier, FSIS has not completed development of a 
Salmonella standard for chicken parts, even though chicken parts are 
now more frequently consumed than whole chickens. As a result, plants 
are able to meet standards for young chicken carcasses but then have 
an outbreak associated with chicken parts, such as occurred with the 
2013 Foster Farms Salmonella outbreak. According to FSIS officials, 
revising standards takes time and resources, in part because the 
agency must first collect data to estimate the prevalence of pathogens 
in FSIS-regulated products, notify the public of proposed standards, 
and open a comment period, all of which can take years. As previously 
mentioned, FSIS expects to announce and request comment on the 
proposed Salmonella standard for chicken parts by the end of fiscal 
year 2014. 

Insufficient Prevalence Estimates: 

FSIS faces a challenge in reducing Salmonella and Campylobacter 
contamination in poultry products as a result of not having sufficient 
prevalence estimates. FSIS collects and analyzes data to estimate the 
prevalence of pathogens when the agency revises or creates standards 
for its regulated products. However, as discussed above, standards are 
not created or revised often, and agency officials we interviewed 
agree that this type of data collection and analysis is done 
infrequently.[Footnote 76] Moreover, the data from FSIS's verification 
testing program makes it difficult for the agency to assess 
contamination levels of poultry pathogens across the entire industry 
because the verification testing program was not designed to assess 
prevalence of pathogens industry-wide, and the agency does not 
randomly select plants for inspection. Also, as we said earlier, in 
April 2012, FSIS concluded that its verification testing data from 
individual plants cannot be used to estimate prevalence across the 
industry because the agency does not randomly select plants for 
verification testing. 

According to USDA's National Advisory Committee on Microbiological 
Criteria for Foods, estimating the prevalence of pathogens in food is 
critical to understanding and addressing the public health risk of 
foodborne illness, and these estimates provide a mechanism for 
measuring performance against public health goals, among other things. 
Similarly, the majority of representatives of consumer groups and some 
industry groups we spoke with stated that FSIS needs estimates of 
prevalence of Salmonella in poultry products to set its Salmonella 
public health goal. As previously mentioned, FSIS recently created a 
new testing approach for ground poultry to estimate prevalence. 
[Footnote 77] According to FSIS officials, the agency plans to propose 
adopting this new testing approach for all of its poultry products, 
which would allow for more frequent data collection and improve 
prevalence estimates, among other things. 

Salmonella's Complex Nature: 

FSIS faces a challenge in reducing Salmonella contamination in poultry 
products because of the complex nature of Salmonella. The majority of 
the representatives from industry and consumer groups we interviewed, 
as well as FSIS officials, agreed that Salmonella is difficult to 
control in poultry products because it is widespread in the natural 
environment. For example, according to CDC officials, there are over 
2,500 serotypes of Salmonella that have been identified--not all of 
which are equally harmful to humans. In many cases, most serotypes are 
rarely involved in human illness cases and outbreaks. In addition, a 
food safety researcher we interviewed mentioned that some serotypes of 
Salmonella are more likely to cause human illness; therefore, it is 
important to understand the genetic makeup of each serotype to 
determine which ones are more or less likely to cause human illness. 
Furthermore, the serotypes that are important in human disease and 
food contamination can differ considerably in different parts of the 
world, and different serotypes may also be associated with different 
animal hosts, habitats, and paths of transmission, according to CDC 
officials. 

Limited Campylobacter Research and Testing: 

Reducing levels of Campylobacter in poultry products poses a challenge 
for FSIS in part because less is known about Campylobacter. In 
addition, CDC officials told us that Campylobacter is less likely to 
be associated with outbreaks. Furthermore, technologies to detect 
Campylobacter might underdiagnose cases and the methods used by many 
diagnostic laboratories to isolate Campylobacter from samples are not 
standardized. Therefore, the efficacy of these tests varies 
considerably. Some countries such as New Zealand experienced greater 
success in reducing Campylobacter levels from poultry products, where 
Campylobacter cases were reduced by approximately 59 percent from 2006-
2008 after its government and industry implemented several proactive 
measures and alterations to critical control points. [Footnote 78] New 
Zealand is leading the international risk-based framework for 
Campylobacter control in poultry.[Footnote 79] Representatives from an 
industry group cautioned, however, that although the decline in 
illnesses in New Zealand is impressive, it is difficult to extrapolate 
this success to other parts of the world. FSIS officials also 
cautioned that the agency's ability to measure a reduction in 
Campylobacter illnesses will depend on its ability to attribute 
Campylobacter illnesses to poultry and other food types and said that 
an interagency analysis with CDC is under way to improve such 
attribution. 

Conclusions: 

Ensuring the safety of poultry products is critical because Americans 
consume considerably more poultry products than beef or pork. To help 
ensure the safety of poultry products, USDA's FSIS has transitioned to 
an increasingly science-based, data-driven, risk-based approach. As a 
part of this approach, USDA's FSIS has taken several actions to reduce 
Salmonella and Campylobacter contamination in poultry products to 
protect human health, including tightening existing standards for 
Salmonella contamination in young chicken and turkey carcasses, as 
well as developing a Salmonella Action Plan. FSIS has also finalized a 
rule to modernize the poultry slaughter inspection process which, 
according to the agency, will play a role in reducing Salmonella and 
other poultry pathogen contamination. To assess the effects of its 
actions on the incidence of human illness from Salmonella and 
Campylobacter, FSIS has developed performance measures and associated 
targets for young chicken carcasses to monitor whether activities to 
bring plants into compliance with the standards are meeting the 
agency's goals, which is consistent with requirements of GPRA to 
measure performance toward achievement of agency strategic goals. 
However, the agency has not developed performance measures and targets 
for certain commonly consumed poultry products, in particular ground 
chicken and ground turkey, even though the standards for Salmonella 
contamination in these products have been in place since 1996. 
Similarly, it has not developed performance measures for Campylobacter 
contamination in young chicken and turkey carcasses, or for Salmonella 
contamination in turkey carcasses. 

FSIS believes it is not appropriate to establish measures for ground 
poultry until it has revised Salmonella standards for it; similarly, 
FSIS believes it is not appropriate to establish measures for 
Campylobacter until it has established plant compliance categories for 
Campylobacter in young chicken and turkey carcasses. According to 
agency officials, revised standards will be proposed and plant 
compliance categories for Campylobacter established by the end of 
2014. FSIS officials told us that a performance measure is not 
necessary for Salmonella in young turkey carcasses because young 
turkey plants are meeting the standard, but an agency report on 
pathogen testing results from the fourth quarter of calendar year 2013 
indicated that not all turkey slaughter plants met the standard. FSIS 
officials told us that all turkey slaughter plants are meeting the 
standard as of July 2014. As previously mentioned, FSIS is developing 
a Salmonella standard for raw chicken parts; therefore, the agency has 
not developed corresponding performance measures and targets. 

In the absence of performance measures and associated targets for 
these pathogens in commonly consumed poultry products, FSIS cannot 
quantitatively assess the effects of its actions related to these 
standards in meeting the agency's goal of maximizing domestic 
compliance with food safety policies and, ultimately, protecting 
public health. Performance measures and targets are reported in the 
agency's strategic and annual plans but, in the absence of such 
measures and targets, performance information, such as data and trends 
supplied to FSIS leadership, is not being publicly reported. Without 
publicly reporting such information, FSIS loses the opportunity to 
enhance transparency by providing this information to the public and 
Congress about its progress in meeting this important goal, 
potentially limiting oversight and accountability. Expeditious 
development of these measures and targets is particularly important 
for ground poultry given that these products are a higher risk for 
contamination than whole carcasses and that the popularity of these 
products has grown over the years. 

In addition, USDA faces several challenges that could hinder its 
ability to reduce contamination in poultry products. For example, 
practices outside the slaughter plant, such as conditions on poultry 
farms, can affect contamination of poultry products. To help overcome 
this challenge, the agency has developed guidelines on practices for 
controlling Salmonella and Campylobacter on farms, but the guidelines 
do not include information on the effectiveness for each practice, as 
recommended by an internal agency committee. FSIS is working with 
industry, APHIS, and other stakeholders to collect information on on-
farm practices to inform future guidelines, but the agency has not 
confirmed that it would include information on the effectiveness of 
each on-farm practice. Without providing this information in future 
revisions of the guidelines, USDA is not fully informing industry of 
the potential benefits of adopting these practices and encouraging 
implementation of such practices. 

Recommendations for Executive Action: 

We recommend that the Secretary of Agriculture direct the 
Administrator of the Food Safety and Inspection Service (FSIS) to take 
the following four actions to help ensure that FSIS efforts protect 
human health by reducing Salmonella and Campylobacter contamination in 
FSIS-regulated poultry products: 

* Once FSIS revises its Salmonella standards for ground chicken and 
ground turkey, the agency should expeditiously develop Salmonella 
performance measures with associated targets for these products to 
monitor whether activities to bring plants into compliance with the 
standards are meeting the agency's goals. 

* Once FSIS establishes plant compliance categories for Campylobacter 
in young chicken and turkey carcasses, the agency should expeditiously 
develop Campylobacter performance measures with associated targets for 
these products to monitor whether activities to bring plants into 
compliance with the standards are meeting the agency's goals. 

* FSIS should expeditiously develop Salmonella performance measures 
with associated targets for young turkey carcasses to monitor whether 
activities to bring plants into compliance with the standards are 
meeting the agency's goals. 

* In future revisions of the compliance guidelines on controlling 
Salmonella and Campylobacter, FSIS should ensure the inclusion of 
information on the effectiveness of each recommended farm practice to 
reduce these pathogens in live poultry. 

Agency Comments and Our Evaluation: 

We provided a draft of this report for review and comment to the 
Department of Agriculture and Department of Health and Human Services. 
In written comments, USDA concurred with our four recommendations; 
USDA's written comments and our detailed response are presented in 
appendix III. According to USDA, the agency will establish appropriate 
measures and targets after collecting adequate data to determine 
whether establishments are meeting the standards for Salmonella in 
ground chicken and turkey and in young turkey carcasses, and for 
Campylobacter in young chicken and turkey carcasses. USDA also stated 
that it is committed to decreasing the number of Salmonella and 
Campylobacter illnesses associated with its regulated products, 
including poultry products, and to using performance standards and 
performance measures to achieve that reduction. Concerning the 
agency's compliance guidelines on controlling Salmonella and 
Campylobacter, USDA said that it is currently revising the guidelines 
to address the reduction of Salmonella and Campylobacter in live 
poultry and will include all available scientific information on the 
effectiveness of each recommended farm practice to reduce Salmonella 
in live poultry. USDA also provided technical comments, as did the 
Department of Health and Human Services, and we incorporated those 
comments as appropriate. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies to the 
appropriate congressional committees, the Secretary of Agriculture, 
the Secretary of Health and Human Services, and other interested 
parties. In addition, the report will be available at no charge on the 
GAO website at [hyperlink, http://www.gao.gov]. 

If you or your staff members have any questions about this report, 
please contact me at (202) 512-3841 or gomezj@gao.gov. Contact points 
for our Offices of Congressional Relations and Public Affairs may be 
found on the last page of this report. GAO staff who made key 
contributions to this report are listed in appendix IV. 

Signed by: 

J. Alfredo Gómez: 
Director, National Resources and Environment: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

This report responds to your request that we review U.S. Department of 
Agriculture's (USDA) approach to protecting human health by reducing 
Salmonella and Campylobacter contamination in poultry products. Our 
objectives for this report were to: (1) describe actions USDA has 
taken since 2006 to reduce Salmonella and Campylobacter contamination 
in poultry products that it regulates; (2) evaluate USDA's efforts to 
assess the effects of these actions on the incidence of human 
illnesses from Salmonella and Campylobacter in poultry products; and 
(3) determine what challenges, if any, USDA faces in reducing these 
pathogens in poultry products. 

To describe actions USDA has taken since 2006 to reduce Salmonella and 
Campylobacter contamination in poultry products that it regulates, we 
reviewed USDA regulations and documentation on actions taken from 
February 2006 to June 2014, including Federal Register notices, USDA's 
Food Safety and Inspection Service[Footnote 80] (FSIS) notices, 
directives, and the December 2013 Salmonella Action Plan. We chose the 
time frame because FSIS introduced several initiatives in 2006 to 
reduce Salmonella. We also reviewed FSIS budget documents from fiscal 
year 2006 to fiscal year 2015, including the agency's budget 
explanatory notes. It is not clear how much FSIS has spent or will 
continue to spend on these actions because our review of FSIS budget 
documents from fiscal year 2006 to fiscal year 2015 found that the 
agency does not provide cost data on specific poultry pathogen 
reduction actions, and the agency was unable to provide such data 
other than $2.5 million spent on poultry-related sampling programs. We 
interviewed officials at FSIS headquarters and district offices about 
recent actions taken to directly address Salmonella and Campylobacter 
contamination of poultry products including actions identified in 
Federal Register notices. 

To evaluate USDA's efforts to assess the effects of actions taken on 
the incidence of human illnesses from Salmonella and Campylobacter in 
poultry products, we reviewed USDA and FSIS strategic plans, as well 
as USDA annual performance reports to identify relevant performance 
measures, targets, and goals.[Footnote 81] We also reviewed FSIS 
quarterly progress reports on Salmonella and Campylobacter testing of 
poultry products. We evaluated the source data, statistical 
methodology, and results of FSIS research articles and a study to 
determine whether the conclusions drawn on the effects of agency 
actions were adequately supported by the evidence. We interviewed FSIS 
headquarters officials about the effects of actions taken on the 
incidence of human illnesses from these pathogens. We reviewed FSIS, 
Centers for Disease Control and Protection (CDC), and Food and Drug 
Administration (FDA) interagency group (Interagency Food Safety 
Analytics Collaboration) documentation to describe the purpose of the 
group, as well as completed and planned analytic projects. We also 
interviewed FSIS and CDC officials about the group's projects to 
improve food attribution data. 

To determine any challenges USDA faces to reduce Salmonella and 
Campylobacter in poultry products it regulates, we reviewed FSIS, USDA 
Office of Inspector General, and National Academy of Sciences reports 
on FSIS's inspections and management challenges. We also interviewed 
officials from USDA's Animal and Plant Health Inspection Service 
(APHIS),[Footnote 82] FSIS, and CDC to identify and describe any 
challenges facing USDA. Additionally, we conducted interviews with 
stakeholder groups. 

We conducted a two-stage interview process with industry, consumer, 
and government employee stakeholder groups. We selected an initial set 
of 12 stakeholder groups identified based on our previous experience 
with large national groups with food safety and slaughter inspection 
knowledge from our August 2013 report on poultry and hog inspections. 
[Footnote 83] These stakeholder groups were the American Meat 
Institute, the Center for Foodborne Illness Research and Prevention, 
the Center for Science in the Public Interest, the Consumer Federation 
of American, Food and Water Watch, the Government Accountability 
Project, the National Association of Federal Veterinarians, the 
National Chicken Council, the National Turkey Federation, the North 
American Meat Association and the Pew Charitable Trusts. We conducted 
an initial interview with the American Federation of Government 
Employees/National Joint Council of Food Inspection Locals to identify 
potential challenges, but we were unsuccessful in obtaining a 
subsequent structured interview. The sample captures a broad range of 
major stakeholder groups but, because it is a nongeneralizable sample, 
it is possible that this group did not include opinions that some 
experts on the topic may have. During the first round, we conducted 
exploratory interviews. We analyzed the results by identifying common 
challenges. During the second round, we conducted structured 
interviews using questions that covered the most common potential 
challenges cited in the exploratory interviews. 

We list the five key structured interview questions we are reporting 
on that stakeholders previously identified as challenges and potential 
challenges during exploratory interviews. For questions listed below, 
respondents had the following choices as responses: definitely yes, 
probably yes, definitely no, probably no, or don't know. There were 
several questions that would have required respondents to have in-
depth information about FSIS, such as details about the technology the 
agency uses for identifying specific serotypes or specifics about how 
the agency would implement actions from its Salmonella Action Plan. 
Lacking this information, many respondents were only able to give us 
qualified answers with caveats. Therefore, we concluded that these 
responses were not standardized to be reported, and we excluded them 
from our report. 

List of Selected Structured Interview Questions: 

* Enforcement: 

FSIS sets performance standards for Salmonella and Campylobacter. Do 
you think FSIS has sufficient authority to ensure that poultry plants 
comply with FSIS performance standards? 

FSIS has declared some pathogen strains adulterants, such as E-coli 
0157. Do you think it is necessary for FSIS to declare some strains of 
Salmonella, particularly strains more likely to cause severe illness, 
as adulterants in order to meet its Salmonella public health goal? 

Currently, poultry plants decide whether Salmonella and Campylobacter 
are hazards reasonably likely to occur in their HACCP. Do you think 
this approach is adequate to keep contamination at FSIS regulated 
plants within FSIS performance standards? 

* Verification Testing and Sampling: 

FSIS recently evaluated its pathogen sampling program and assessed 
that it could not measure prevalence of Salmonella in poultry products 
over time, in part because it does not conduct random sampling. Do you 
think FSIS needs estimates of prevalence of Salmonella in poultry 
products to set its Salmonella public health goal? 

Salmonella is known to be ubiquitous and persistent in the natural 
environment. Do you think these traits make it difficult to control 
Salmonella in poultry products FSIS regulates? 

To make sure that our results were presented in an accurate and 
balanced manner, we evaluated responses in terms of the extent of 
agreement among the three stakeholder groups. Where there was stark 
disagreement among the groups, we presented the results separately. 
Where there was general agreement, we reported overall results. We 
conducted 11 structured phone interviews from March 2014 to May 2014. 
We selected 4 of the 11 stakeholder groups and pretested the initial 
structured interview questions to ensure that the questions were 
relevant and clearly stated; based on those results, we made 
adjustments to the structured interview as necessary. Apart from the 
stakeholder interviews, we also spoke with two academic food safety 
researchers we identified based on academic work and participation in 
a 2014 food safety conference and discussed challenges related to 
controlling Salmonella and Campylobacter in poultry products. 

Additionally, we reviewed five Salmonella and Campylobacter outbreaks 
attributed to poultry products with the highest number of illnesses to 
better understand any challenges that may have contributed to the 
outbreaks; these outbreaks started during the time period from fiscal 
year 2011 through fiscal year 2013. The case study included four 
Salmonella outbreaks from poultry products and one Campylobacter 
outbreak from poultry products. These outbreaks are not generalizable 
to all outbreaks from Salmonella and Campylobacter contaminated 
poultry products. The selected outbreaks provided illustrative 
examples of challenges USDA faces. To describe the selected outbreaks, 
we interviewed state officials from departments of health or 
agriculture from California, New York, Texas, Vermont, and Washington. 
These states had the highest number of illnesses for each outbreak in 
our review. We also interviewed CDC officials with knowledge of the 
outbreaks and FSIS officials familiar with each outbreak investigation 
to learn about any challenges USDA faced and other challenges related 
to investigating outbreaks. We also reviewed states', CDC, and FSIS 
documentation on each outbreak to describe each outbreak. 

For our three objectives, we visited poultry plants in California to 
gain a better understanding of poultry plant operations and FSIS 
inspection activities. We chose California because the state has a 
number of small and large chicken and turkey plants. California also 
produces a high volume of chicken and turkey in the United States. We 
reviewed prior GAO reports on food safety, surveillance systems, and 
performance management. 

We conducted this performance audit from July 2013 to September 2014 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

[End of section] 

Appendix II: Case Study Outbreak Summary Tables: 

During the course of our review, we examined a nongeneralizable sample 
of recent Salmonella and Campylobacter outbreaks linked to poultry 
products since 2011. We limited the scope of our review to four of the 
six most recent Salmonella outbreaks that had the highest number of 
confirmed illnesses that the Centers for Disease Control and 
Prevention (CDC) and the U.S. Department of Agriculture's Food Safety 
and Inspection Service (FSIS) investigated since 2009. In addition, we 
also identified one Campylobacter outbreak that occurred within the 
time period established for those Salmonella outbreaks included in our 
review. As this was a review of a nongeneralizable sample of 
outbreaks, the information is not generalizable to all outbreaks but 
provides important illustrative information. 

In addition to the information provided for each of the five outbreaks 
in the summary tables below (see tables 3-7), other findings from our 
review included the following: 

* Each plant linked to the four Salmonella outbreaks included in our 
review was a Category 1 plant at the time the outbreaks occurred, 
meaning that the plants were considered by FSIS to be demonstrating 
consistent process controls to meet the agency's existing Salmonella 
standards. 

* During the five outbreaks, FSIS requested that one company conduct a 
voluntary recall. FSIS made the request after the agency and CDC 
linked a person sickened by an outbreak strain of Salmonella 
Heidelberg to chicken produced by that company and collected from the 
ill person's home. According to FSIS officials, FSIS did not request 
voluntary recalls during the other outbreaks for several reasons. 
First, for two of the Salmonella outbreaks, the companies initiated 
voluntary recalls before FSIS could make a formal request. Second, for 
the remaining Salmonella outbreak, FSIS did not identify the level of 
information necessary to request a voluntary recall. Finally, for the 
Campylobacter outbreak, the company voluntarily elected to cease 
harvesting and selling of raw chicken livers. 

* During the five outbreaks, FSIS issued two public health alerts. 
Specifically, FSIS issued an alert for the outbreak of Salmonella 
linked to ground turkey products and another for the outbreak of 
Salmonella linked to specific brands of chicken products.[Footnote 84] 

* FSIS found that each of the plants linked to three of the four 
Salmonella outbreaks had inadequate Hazard Analysis and Critical 
Control Point (HACCP) plans. 

* FSIS found that each of the plants linked to three of the four 
Salmonella outbreaks failed to maintain sanitary conditions and comply 
with the agency's regulatory requirements for sanitation. 

* FSIS determined that each of the plants linked to three of the four 
Salmonella outbreaks had not adequately supported parts of their 
hazard analyses because they did not identify Salmonella as a food 
safety hazard reasonably likely to occur during certain production 
processes. 

* In response to the Campylobacter outbreak, FSIS conducted a food 
safety assessment and found that the plant failed to correctly 
identify Campylobacter as a pathogen of concern in its hazard 
analyses, a noncompliance with FSIS regulations. 

Table 3: 2013 - 2014 Multistate Outbreak of Salmonella Linked to 
Chicken Products Produced in California: 

Outbreak pathogen: Salmonella Heidelberg; 
States affected by outbreak: 29, plus Puerto Rico; 
Number of reported illnesses[A]: 634; 
(77% of cases reported from California); 
Number of hospitalizations: 200 of 528 (38%) based on available 
information; 
Product recalled? Yes; 
Antibiotic resistance identified? Yes. 

Summary: 
According to CDC's outbreak investigation report, an outbreak linked 
to Foster Farms brand chicken products sickened a reported 634 people 
between March 2013 and July 2014. CDC first reported a number of 
illnesses linked to a strain of Salmonella Heidelberg to FSIS in June 
2013. Additional investigation identified illness onset dates linked 
to this outbreak that began in March 2013. Epidemiologic, laboratory, 
and traceback investigations conducted by officials in local, state, 
and federal public health, agriculture, and regulatory agencies 
identified Foster Farms brand chicken as the likely source of this 
outbreak. For example, FSIS inspectors identified one or more of the 
outbreak strains of Salmonella in the company's chicken products while 
conducting intensified Salmonella testing at three Foster Farms plants 
in California linked to the outbreak. In addition, health 
investigators linked seven strains of Salmonella Heidelberg to this 
outbreak, including four strains rarely reported to CDC. Moreover, 15 
percent of those sickened developed blood infections as a result of 
their illness. Typically, approximately 5 percent of persons ill with 
Salmonella infections develop blood infections. CDC also found that 
the outbreak strains were resistant to antibiotics, with some being 
multidrug resistant; none of the reported antibiotics that the strains 
were found to be resistant to were those commonly used to treat 
serious Salmonella infections in humans (e.g., fluoroquinolones and 
third generation cephalosporins). 

On October 7, 2013, FSIS issued a public health alert due to concerns 
that illnesses caused by strains of Salmonella Heidelberg were 
associated with raw chicken products produced at three Foster Farms 
plants in California. On October 8, the Kroger Company, which owns a 
number of grocery stores, announced that it was removing all raw 
chicken products produced at the three Foster Farms plants from its 
stores and warehouses. In addition, a Costco location in California 
later conducted two voluntary recalls--the first on October 12 and the 
second on October 17--of its rotisserie chicken products (which used 
Foster Farms brand chickens) as those products may have been 
contaminated with Salmonella Heidelberg as a result of cross-
contamination. On July 3, 2014, Foster Farms announced it was 
conducting a voluntary recall of an undetermined amount of its chicken 
products that may be contaminated with one of the outbreak strains of 
Salmonella Heidelberg. According to CDC's outbreak investigation 
report, the recall resulted from FSIS identifying one of the outbreak 
strains of Salmonella Heidelberg in leftover Foster Farms brand 
chicken with labeling information collected from the home of a person 
infected with the same strain in California. Prior to the July 2014 
recall, FSIS officials said that the agency had not requested that 
Foster Farms conduct a voluntary recall of any of its chicken products 
because the agency had not made a definitive connection between a 
sickened patient and a specific Foster Farms chicken product that 
provided information pertaining to production lot numbers and 
production dates, among other things. On July 31, 2014, CDC announced 
that it believed that the outbreak was over after it determined that 
the number of reported infections had returned to levels the agency 
normally expects to see. 

Sources: GAO summary of FSIS and CDC documents. GAO-14-744. 

[A] According to CDC's food safety progress report for 2013, among all 
Salmonella cases reported to CDC, for each reported case there are 29 
cases not diagnosed or reported. Therefore, the number of people 
sickened as a result of this outbreak is likely substantially higher 
than the official case count. 

[End of table] 

Table 4: 2012 -2013 Multistate Outbreak of Salmonella Linked to 
Chicken Products: 

Outbreak pathogen: Salmonella Heidelberg; 
States affected by outbreak: 13; 
Number of reported illnesses[A]: 134; 
(about 43% of cases reported from Washington); 
Number of hospitalizations: 33 of 105 (31%) based on available 
information; 
Product recalled? No; 
Antibiotic resistance identified? Yes. 

Summary: 
According to CDC's outbreak investigation report, an outbreak linked 
to Foster Farms chicken products sickened a reported 134 people 
between June 2012 and May 2013. The Oregon Health Authority and the 
Washington State Department of Health first identified the outbreak in 
June 2012 after identifying a number of illnesses caused by a strain 
of Salmonella Heidelberg. In August, CDC detected additional 
Salmonella illnesses matching the previously identified outbreak 
pattern. Working together, local, state, and federal agencies 
conducted a traceback investigation and determined the likely source 
was chicken products produced by Foster Farms. Specifically, 
Washington State Public Health Laboratories conducted tests on four 
samples of chicken collected from three ill persons' homes in 
Washington and identified the outbreak strain in those samples, which 
were then traced back to two Foster Farms slaughter plants. In 
addition, the National Antimicrobial Resistance Monitoring System--
Enteric Bacteria, a collaborative surveillance program between state 
and local public health departments, CDC, and the U.S. Food and Drug 
Administration, among others, isolated the outbreak strain of 
Salmonella Heidelberg from Oregon and Washington Foster Farms retail 
chicken samples. 

In February 2013, the states of Washington and Oregon issued public 
health alerts warning the public about the outbreak of Salmonella and 
that it had been linked to chicken produced in Foster Farms plants in 
Washington and California. FSIS, however, did not issue a public 
health alert in response to this outbreak. According to FSIS 
officials, the agency was unable to make a definitive connection 
between a sickened patient and a specific Foster Farms chicken product 
that provided information pertaining to production lot numbers and 
production dates, among other things. As a result, FSIS did not 
request that Foster Farms conduct a voluntary recall of any of its 
chicken products, and Foster Farms did not act independently to 
initiate a voluntary recall during this outbreak. On July 10, 2013, 
CDC announced that it believed that the outbreak was over after it 
determined that the number of reported infections had returned to 
levels the agency normally expects to see. 

Sources: GAO summary of CDC and Washington State Department of Health 
documents. GAO-14-744. 

[A] According to CDC's food safety progress report for 2013, among all 
Salmonella cases reported to CDC, for each reported case there are 29 
cases not diagnosed or reported. Therefore, the number of people 
sickened as a result of this outbreak is likely substantially higher 
than the official case count. 

[End of table] 

Table 5: 2012 Campylobacter Outbreak Linked to Chicken Livers in 
Vermont: 

Outbreak pathogen: Campylobacter jejuni; 
States affected by outbreak: 3; 
Number of reported illnesses[A]: 6; 
(67% of cases reported from Vermont); 
Number of hospitalizations: 2 (33%); 
based on available information; 
Product recalled? No[B]; 
Antibiotic resistance identified? Yes. 

Summary: 
According to CDC's Morbidity and Mortality Weekly Report, an outbreak 
linked to chicken livers from an establishment in Vermont sickened a 
reported six people from April 2012 to September 2012. Vermont's 
Department of Health identified the outbreak in October 2012 after 
identifying three cases of Campylobacter infection in Vermont 
residents. The department, in collaboration with its Health 
Laboratory, carried out the investigation, interviewed patients, 
conducted human and food laboratory testing, and initiated the 
traceback actions to identify the source of the outbreak. As a result 
of its investigative work, the Vermont Department of Health determined 
that all six of those sickened had either been exposed to or consumed 
raw or undercooked chicken livers that had been produced at the same 
poultry establishment. It was also able to trace the chicken livers 
back to a single source, Establishment A, located in Vermont. 

In response to the outbreak, FSIS conducted a food safety assessment 
at the establishment. No major violations were identified, but the 
FSIS inspector determined that the plant had failed to correctly 
identify Campylobacter as a pathogen of concern in its hazard 
analyses, a noncompliance with FSIS regulations. In addition, the 
establishment responded to the outbreak by electing to cease 
harvesting and selling chicken livers on November 9, 2012. According 
to the Vermont Department of Health, this outbreak is of note because 
it was the first-ever documented multistate outbreak of Campylobacter 
due to chicken liver consumption in the United States and the first 
Campylobacter outbreak in which the outbreak strain was found in the 
chicken livers. 

Sources: GAO summary of FSIS, CDC, and Vermont Department of Health 
documents. GAO-14-744. 

[A] According to CDC's food safety progress report for 2013, among all 
Campylobacter cases reported to CDC, for each reported case there are 
30 cases not diagnosed or reported. Therefore, the number of people 
sickened as a result of this outbreak is likely substantially higher 
than the official case count. 

[B] No recall occurred during this outbreak. However, in December 
2012, the establishment formally notified FSIS of its decision to stop 
harvesting and selling chicken livers as a result of the outbreak. 

[End of table] 

Table 6: 2011 Multistate Outbreak of Salmonella Linked to Broiled 
Chicken Liver Products: 

Outbreak pathogen: Salmonella Heidelberg; 
States affected by outbreak: 6; 
Number of reported illnesses[A]: 190; 
(57% of cases reported from New York); 
Number of hospitalizations: 30 of 154 (19%) based on available 
information; 
Product recalled? Yes; 
Antibiotic resistance identified? Not Known[B]. 

Summary; 
According to CDC's outbreak investigation report, an outbreak linked 
to broiled chicken livers produced by Schreiber Processing Corporation 
sickened a reported 190 people from April 1 to November 17, 2011. CDC 
first identified the outbreak in August 2011 when the agency observed 
a sustained increase in the number of Salmonella Heidelberg illnesses 
reported from New York and New Jersey. Collaborative investigative 
efforts of local, state, and federal public health and regulatory 
agencies linked the outbreak to eating broiled chicken livers from 
Schreiber Processing Corporation. On November 8, 2011, FSIS announced 
that Schreiber Processing Corporation was conducting a voluntary 
recall of an undetermined amount of broiled chicken liver products. On 
November 9, the New York State Health Department and the New York City 
Department of Health and Mental Hygiene issued public health warnings 
as an additional measure to inform consumers about the recall and the 
link between broiled chicken livers produced by Schreiber Processing 
and the Salmonella outbreak. 

Following the recall, Schreiber Processing elected to stop producing 
these particular broiled chicken liver products. FSIS also conducted a 
food safety assessment at the company's plant from November 30, 2011 
to January 18, 2012. Based on its findings, FSIS determined that the 
manner in which the broiled chicken liver products were labeled played 
a role in the outbreak. Specifically, the products appeared to be 
ready-to-eat, but were in fact partially cooked, and therefore needed 
to be fully cooked before eating. As a result, consumers may have 
incorrectly thought the use of "broiled" meant the chicken livers were 
cooked and ready-to-eat. In response to these findings, FSIS is 
considering revising its safe food handling label. Prior to the 
conclusion of FSIS's food safety assessment, CDC announced on January 
11, 2012, that the outbreak was over. 

Sources: GAO summary of FSIS, CDC, New York State Department of 
Health, and New York City Department of Health and Mental Hygiene 
documents. GAO-14-744. 

[A] According to CDC's food safety progress report for 2013, among all 
Salmonella cases reported to CDC, for each reported case there are 29 
cases not diagnosed or reported. Therefore, the number of people 
sickened as a result of this outbreak is likely substantially higher 
than the official case count. 

[B] During our review, FSIS and CDC officials told us that they were 
unable to determine whether the outbreak strain showed any antibiotic 
resistance. 

[End of table] 

Table 7: 2011 Multistate Outbreak of Salmonella Linked to Ground 
Turkey: 

Outbreak pathogen: Salmonella Heidelberg; 
States affected by outbreak: 34; 
Number of reported illnesses[A]: 136; 
(13% of cases reported from Texas); 
Number of hospitalizations: 37 of 94 (39%) based on available 
information; 
Product recalled? Yes; 
Antibiotic resistance identified? Yes. 

Summary: 
According to CDC's outbreak investigation report, an outbreak linked 
to ground turkey produced by Cargill Meat Solutions Corporation was 
associated with one death and sickened a reported 136 people from 
February to September 2011. CDC identified the outbreak in May 2011 
when it observed an increase in Salmonella Heidelberg illnesses in 
multiple states. Collaborative investigative efforts of state, local, 
and federal public health and regulatory agencies led to the 
identification of matching strains of Salmonella obtained from ground 
turkey products purchased at different retail locations. Preliminary 
information indicated that these products originated from a common 
source: a Cargill plant in Springdale, Arkansas. FSIS began its 
initial traceback investigation in July 2011. On July 29, 2011, FSIS 
released a public health alert warning the public that frozen and 
fresh ground turkey products had been linked to an outbreak of 
Salmonella. That same day, FSIS informed Cargill that the results of 
its investigative work indicated that the company's Springdale, 
Arkansas plant was the likely, but not definitive, source of the 
outbreak. 

On August 3, 2011, Cargill voluntarily recalled approximately 36 
million pounds of ground turkey products produced at the Springdale 
plant. Cargill also suspended its ground turkey operations at the 
plant in order to identify the source of the Salmonella Heidelberg and 
to implement corrective actions. Production resumed 2 weeks later. 
However, Cargill conducted a second voluntary recall of approximately 
185,000 pounds of ground turkey products on September 11, 2011. The 
second recall occurred after product samples at the plant tested 
positive for the outbreak strain of Salmonella Heidelberg based on 
sample results from an intensive in-plant investigation performed by 
FSIS. Cargill again suspended operations and sought to implement 
further corrective actions. According to CDC's outbreak investigation 
report, no further illnesses were reported after September 13, 2011, 
and on November 10, 2011, it announced that the outbreak appeared to 
be over. In December 2011, following USDA's approval of its enhanced 
food safety plan, Cargill's Springdale plant resumed its ground turkey 
production. 

Sources: GAO summary of FSIS and CDC documents and Cargill press 
releases. GAO-14-744. 

[A] According to CDC's food safety progress report for 2013, among all 
Salmonella cases reported to CDC, for each reported case there are 29 
cases not diagnosed or reported. Therefore, the number of people 
sickened as a result of this outbreak is likely substantially higher 
than the official case count. 

[End of table] 

[End of section] 

Appendix III: Comments from the U.S. Department of Agriculture: 

Note: GAO comments supplementing those in the report text appear at 
the end of this appendix. 

Department of Agriculture: 
Office of The Secretary: 
Washington, D.C. 20250: 

September 15, 2014: 

J. Alfredo Gomez:
Director:
Natural Resources and Environment: 
United States Government Accountability Office: 
441 G. Street, N.W. 
Washington, D.C., 20508: 

Dear Mr. Gomez: 

The United States Department of Agriculture (USDA) appreciates the 
opportunity to review the U.S. Government Accountability Office's 
(GAO) draft report entitled Food Safety: USDA Needs to Strengthen Its 
Approach to Protecting Human Health from Pathogens in Poultry Products 
(GAO-14-744). USDA appreciates GAO's work in planning, conducting and 
issuing this report. There are a few general comments we would like to 
make on the report. 

The GAO report discusses the Agency's Salmonella and Campylobacter 
performance standards and the performance measures and targets that 
the Agency has set to track its progress. We would like to highlight 
that one of the key performance measures that FSIS has set a target 
for and tracks is the All-Illness Measure. An important component of 
that measure is the estimated number of Salmonella illnesses 
attributed to FSIS-regulated products. As part of monitoring the All-
Illness Measure, FSIS tracks the estimated number of Salmonella 
illnesses from FSIS-regulated products on a quarterly basis. Poultry 
is the source of a large percentage of those Salmonella illnesses, 
making this a very important performance measure related to poultry 
pathogens. FSIS uses its evaluation of progress on that measure to 
inform its policy decisions related to poultry pathogens. For example, 
examinations and discussions of the Agency's progress toward its All-
Illness Measure targets were part of the impetus for FSIS' Strategic 
Performance Working Group to develop the Agency's Salmonella Action 
Plan. 

As GAO notes, the Agency has not established formal performance 
measures for all poultry product and pathogen combinations. We would 
like to emphasize, however, that FSIS has established or is in the 
process of establishing performance standards for a number of pathogen—
product pairs for poultry. The development of performance standards is 
a formal, rigorous process that often includes the conduct of baseline 
studies or pathogen sampling, risk assessments, and numerous analyses, 
all of which often takes a couple of years to accomplish and therefore 
reflects a substantial effort by the Agency. Once formal performance 
standards are developed and implemented, then FSIS can establish 
performance measures to track progress related to those standards. 
Although the Agency has highlighted specific targets through the 
establishment of some formal performance measures and targets, FSIS 
does not limit its ability to set goals and track its progress and the 
progress of the industries it regulates to those specific measures. 
FSIS tracks, and in some cases makes public, that progress regardless 
of whether it has formalized performance measures in place. For 
example, FSIS does not have a specific performance measure and target 
for the percentage of establishments in different categories of 
Campylobacter performance, but the Agency reports publicly on the 
turkey establishments in the different categories and percentage of 
positives for Campylobacter. That information allows for routine 
assessment of trends by those within and outside the Agency, and 
provides the opportunity for FSIS and the regulated industry the 
ability to adjust food safety activities on the basis of those trends. 

We have also noted a few more technical aspects in the report that we 
would like to clarify. Throughout this report, GAO uses the terms 
"mechanically separated" and "ground" when referring to poultry 
products. FSIS prefers the word 'comminuted' and its standard 
definition. [See comment 1] 

FSIS would also like to clarify and correct a statement made on page 2 
of the report. The data on the incidence of Salmonella and 
Campylobacter in humans is misleading and provides an inaccurate 
assessment of the burden of Salmonella illnesses specifically 
attributed to FSIS-regulated products. The CDC data being cited in 
this section of the report comes from the CDC-led FoodNet program and 
represents all Salmonella illnesses captured through the program, not 
just those attributed to FSIS-regulated products. Salmonella 
attribution to FSIS-regulated products has actually decreased over the 
past 5 years. If additional language is needed, please consider the 
following discussion of FSIS attribution: [See comment 2] 

FY2014 data: 

* Thus far in FY2014, FSIS met its targets for Salmonella illnesses 
reductions for all quarters (Q1-Q3, FY2014). 

From FSIS 2013 Year-End Report: 

* The overall number of illnesses in the U.S. population caused by 
Salmonella declined by 7.6% over the course of FY 2013, which 
positively affected the Agency's All-Illness measure. 

* Of the number of Salmonella outbreaks associated with FSIS-regulated 
products—-or FSIS' attribution fraction for Salmonella—-there was a 
decrease from 33.1% in the 2008-2010 period to 31.7% in the 2009-2011 
period. 

* The Agency closed the gap on the All-Illness measure from FY 2012 to 
FY 2013. (In FY 2012, the Agency missed the All-Illness measure by 
18%. In FY 2013, the Agency only missed the All-Illness measure by 8%). 

We would also like to note the work of the Interagency Food Safety 
Analytics Collaboration (IFSAC). IFSAC is a major effort for the 
Agency, as well as the US Centers for Disease Control and Prevention 
(CDC) and the US Food and Drug Administration (FDA). The current focus 
of IFSAC's activities is foodborne illness source attribution, defined 
as the process of estimating the most common food sources responsible 
for specific foodborne illnesses. Since 2009, IFSAC has completed four 
projects and 2 external collaborations. IFSAC also held a public 
meeting in 2012, broadcast two public webinars to share results from 
completed projects, launched a new web resource, and is in the process 
of planning a public meeting for 2015 to share results from completed 
and on-going projects. A number of these interagency efforts involve 
work to better estimate attribution for Salmonella and Campylobacter, 
including evaluating a pathogen subtype model to better estimate the 
number of Salmonella illnesses associated with different food sources 
and the development of a new, robust method to produce harmonized 
attribution estimates for Salmonella, Campylobacter, E. call 0157:H7 
and Listeria monocytogenes. This project will assist FSIS in better 
estimating the proportion of Salmonella illnesses associated with 
poultry. On page 20 of its report, GAO discusses the limitations of 
the data on which FSIS relies for measuring illnesses. FSIS agrees 
that there are limitations in the data, but would like to point out 
that it uses the best data that are available. The limitations in the 
data are why, in part, IFSAC was formed. [See comment 3] 

USDA recommends consulting with CDC to address some inaccuracies in 
the text regarding outbreak investigations, specifically the text and 
figure on page 6. FSIS typically does not identify outbreaks through 
the CDC outbreak data, it is the State and local health departments 
that do so. [See comment 4] 

On page 26 of the report where there is discussion of limited 
enforcement authority, the report states that "... after a plant fails 
Salmonella testing for its first sample set, FSIS can require a 
reassessment of the plant's HACCP plan and then conduct a food safety 
assessment; conduct additional sampling or issue a notice of intended
enforcement action,". This is not quite accurate. The final rule 
requires that establishments address enteric pathogens such as 
Salmonella and Campylobacter in a written plan. This plan can be in 
their HACCP, sanitation SOP, or other prerequisite program and 
therefore the establishment's food safety system. [See comment 5] 

On page 31 the report states that FSIS' prevalence estimates are 
insufficient. FSIS acknowledges this issue and plans to remedy it by 
implementing routine sampling throughout the year for Salmonella. FSIS 
has announced in two Federal Register Notices (FRN) that for 
Salmonella verification sampling it is considering an approach that 
involves routine sampling throughout the year, overlaid with a moving 
window to determine an establishment's compliance with the performance 
standard. One option that FSIS is considering is a moving window 
sampling plan in which FSIS would evaluate a set number of sequential 
results from a single establishment to assess process control (78 FR 
53019 and 79 FR 32437). FSIS would put the moving window approach into 
effect when the revised performance standard is implemented. 

FSIS appreciates and concurs with the recommendations from GAO. The 
Agency's response to each recommendation is included below. 

Recommendation 1: 

Once FSIS revises its Salmonella standards for ground chicken and 
ground turkey, the Agency should expeditiously develop a Salmonella 
performance measure with associated targets for these products to 
monitor whether activities to bring plants into compliance with the 
standards are meeting the Agency's goals. 

USDA Response: 

FSIS concurs with this recommendation. FSIS is committed to decreasing 
the number of Salmonella illnesses associated with its regulated 
products, including poultry products, and to using performance 
standards and performance measures to achieve that reduction. FSIS has 
already begun additional Salmonella sampling and testing of comminuted 
poultry as a part of an Agency effort to revise the existing 
standards. As GAO notes in its report, FSIS is also in the process of 
developing performance standards for Salmonella and Campylobacter in 
raw chicken parts and comminuted poultry (chicken and turkey). The 
Agency anticipates the release of the FRN with those standards within 
this calendar year. Those risk assessment-based standards are designed 
to integrate performance metrics into the standards, and link 
establishment compliance levels with targeted decreases in Salmonella 
illnesses, which are aligned to the Agency's All-Illness Measure. The 
FRN also establishes a framework for monitoring the standards through 
categorizing establishments and posting establishment categories on 
FSIS' website. After the standards have been implemented and adequate 
data are gathered to determine the percentage of industry complying 
with the standards, FSIS will set additional measures and targets, as 
necessary, to meet its public health goals. 

Recommendation 2: 

Once FSIS establishes plant compliance categories for Campylobacter in 
young chicken and turkey carcasses, the Agency should expeditiously 
develop Campylobacter performance measures with associated targets for 
these products to monitor whether activities to bring plants into 
compliance with the standards are meeting the Agency's goals. 

USDA Response: 

FSIS concurs with this recommendation. FSIS is committed to decreasing 
the number of Campylobacter illnesses from its regulated products, 
including poultry products, and to using performance standards and 
performance measures to achieve that reduction. As is stated in the 
previously mentioned FRN that FSIS anticipates publishing this 
calendar year, FSIS also intends to develop categories for 
establishments with Campylobacter sampling results for young chicken 
and turkeys, on the basis of the 2011 performance standards for 
Campylobacter in those products. Once FSIS has collected enough data 
to determine whether establishments are meeting the existing 
performance standards, the Agency will establish appropriate measures 
and targets and also appropriate follow up actions if establishments 
are not meeting the standards. 

Recommendation 3: 

FSIS should expeditiously develop Salmonella performance measures with 
associated targets for young turkey carcasses to monitor whether 
activities to bring plants into compliance with the standards are 
meeting the Agency's goals. 

USDA Response: 

FSIS concurs with this recommendation. FSIS is committed to decreasing 
the number of Salmonella illnesses from its regulated products, 
including poultry products, and to using performance standards and 
performance measures to achieve that reduction. FSIS developed 
performance standards for Salmonella in turkey establishments in 2011. 
Similar to the Campylobacter-related activities outlined in the 
Agency's response to Recommendation 2, FSIS informed turkey 
establishments of their Salmonella category status at the end of 2013. 
The Agency also posts sampling data on its website on a quarterly 
basis. Once FSIS has collected adequate data to determine whether 
establishments are meeting the performance standards, the Agency will 
establish appropriate measures and targets and also appropriate follow 
up actions if the establishments are not meeting the standards. 

Recommendation 4: 

In future revisions of the compliance guidelines on controlling 
Salmonella and Campylobacter, FSIS should ensure the inclusion of 
information on the effectiveness of each recommended farm practice to 
reduce these pathogens in live poultry. 

USDA Response: 

FSIS concurs with this recommendation. FSIS does not have authority 
over pre-harvest activities, but it does incorporate pre-harvest-
related research results from the published literature into its 
compliance guidelines. FSIS is currently revising compliance 
guidelines to address the reduction of Salmonella and Campylobacter in 
live poultry. The Agency will include all available scientific 
information on the effectiveness of each recommended farm practice to 
reduce Salmonella in live poultry in those guidelines. Although FSIS 
is not a research organization and is not able to conduct its own 
research on farm practices, it does recognize the importance of such 
research to carry out its mission. The Agency has highlighted the need 
to "[e]valuate the potential effectiveness of pre-harvest 
interventions on finished products" as a top research priority. FSIS 
has posted that particular priority and more specific research 
objectives on its website [hyperlink, 
http://wwvv.fsis.usda.gov/wps/portal/fsishopics/science/food-safety-
researchpriorities], and collaborates with its research partners—
including government, academic and industry researchers—encouraging 
them to conduct that research. 

Again, thank you for the opportunity to review and comment on this 
draft report. We look forward to working with you on future USDA 
engagements. 

Sincerely, 

Signed by: 

Brian Ronholm: 
Deputy Under Secretary: 
Food Safety: 

GAO Comments: 

The following are GAO's comments on the U.S. Department of 
Agriculture's (USDA) letter dated September 15, 2014. 

1. USDA commented that, throughout our report, we use the terms 
"mechanically separated" and "ground" when referring to poultry 
products and that the agency prefers the term "comminuted" and its 
standard definition. USDA defines comminuted poultry products as 
"poultry (chicken or turkey) products that have been ground, 
mechanically separated, or hand-or mechanically-deboned and further 
chopped, flaked, minced, or otherwise processed to reduce particle 
size." For our report, it is important to distinguish among these 
products because there are standards for ground poultry products but 
not for mechanically separated or other products that are included 
under the umbrella of "comminuted" poultry. 

2. USDA commented that our use of data from the Centers for Disease 
Control and Prevention (CDC) for describing the incidence of 
Salmonella and Campylobacter is misleading and provides an inaccurate 
assessment of the burden of Salmonella illnesses specifically 
attributed to products regulated by USDA's Food Safety and Inspection 
Service (FSIS). The purpose of these data in the introduction is to 
provide general context for our review, not to provide information on 
illnesses specifically attributed to FSIS-regulated products, which 
are discussed later in the report. We modified our report to note that 
CDC's data for describing the incidence of Salmonella and 
Campylobacter include illnesses attributed to other sources in 
addition to FSIS-regulated products. 

3. USDA commented that it agrees that there are limitations in the 
data that it relies upon for measuring illness but stated that the 
agency uses the best data that are available. We modified our report 
to state that, according to USDA, the agency is using the best data 
available. 

4. We provided a draft of this report to the Department of Health and 
Human Services, which includes CDC. We incorporated the agency's 
technical comments as appropriate. 

5. USDA commented that our statement about an example of limited 
enforcement authority did not include a new requirement under the 
agency's final rule to modernize poultry slaughter inspection. We 
modified our report to note this new requirement. 

[End of section] 

Appendix IV: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

J. Alfredo Gómez, (202) 512-3841 or gomezj@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, Mary Denigan-Macauley 
(Assistant Director), Carl Barden, Kevin Bray, Mark Braza, Nkenge 
Gibson, Cynthia Norris, Josephine Ostrander, and Kevin Remondini made 
key contributions to this report. 

[End of section] 

Footnotes: 

[1] We did not independently verify the data underlying CDC's 
statement. 

[2] Pathogens are disease-causing organisms including bacteria such as 
Salmonella and Campylobacter. 

[3] John A. Painter et. al., "Attribution of Foodborne Illnesses, 
Hospitalizations, and Deaths to Food Commodities by using Outbreak 
Data, United States, 1998-2008," Emerging Infectious Diseases 19, no. 
3 (2013), 407-415. 

[4] 21 U.S.C. §§ 601-683 and 21 U.S.C. §§ 451-472. The Federal Meat 
Inspection Act was originally enacted in 1907 as part of the USDA 
appropriations act, and the Poultry Products Inspection Act was 
enacted in 1957. Both pieces of legislation have been amended a number 
of times throughout the years. 

[5] FSIS standards apply to poultry products that are raw or not fully 
cooked to eliminate Salmonella or Campylobacter, and the agency has a 
zero tolerance for these pathogens in fully cooked, also known as 
ready-to-eat, poultry products. 

[6] In CDC's food safety progress report for 2013, CDC used 
surveillance data that include Salmonella and Campylobacter illnesses 
attributed to other sources in addition to FSIS-regulated products. 

[7] The Government Performance and Results Act of 1993 as amended by 
the GPRA Modernization Act of 2010 (GPRA) requires executive agencies 
to complete strategic plans in which they define their missions, 
establish results-oriented goals, and identify the strategies that 
will be needed to achieve those goals. GPRA also requires agencies to 
complete annual performance plans that establish performance goals--
which contribute to the strategic goals--and measure performance 
toward achieving performance goals. Performance measures are important 
management tools that help agencies monitor and report progress toward 
their goals. Numerical targets are a key attribute of successful 
performance measures because they allow managers to compare planned 
performance with actual results. 

[8] GAO, Food Safety: More Disclosure and Data Needed to Clarify 
Impact of Changes to Poultry and Hog Inspections, [hyperlink, 
http://www.gao.gov/products/GAO-13-775] (Washington, D.C.: Aug. 22, 
2013). 

[9] One of the initial 12 stakeholder groups was not available for a 
structured interview. 

[10] An outbreak consists of two or more cases of a similar illness 
that come from consumption of a common food, according to CDC. 

[11] Serotypes are groups within a single species of microorganisms, 
such as bacteria, that share distinctive surface structures. 

[12] Species are the largest group of microorganisms capable of 
reproduction. 

[13] CDC, National Enteric Disease Surveillance: Salmonella Annual 
Report, 2011 (Atlanta, Ga: 2013). 

[14] Reactive arthritis is a form of inflammatory arthritis that 
develops in reaction to an infection by bacteria and can last for 
years. 

[15] 61 Fed. Reg. 38806 (July 25, 1996). 

[16] The farm-to-table continuum for food safety includes all facets 
of the food production process: on the farm, animal slaughter in FSIS-
regulated poultry plants, food processing within regulated plants, 
retail or market establishments (e.g., grocery stores), and home 
environments. 

[17] Chemical interventions are processing aides, such as chlorine 
rinses, used to eliminate or reduce pathogens. 

[18] Prevalence is the proportion of poultry that would test positive 
for pathogens if the entire population of poultry products were 
sampled and analyzed during a specific period of time. 

[19] 76 Fed. Reg. 15282 (Mar. 21, 2011). FSIS's revised Salmonella 
standards for young chicken and turkey carcasses took effect in July 
2011. FSIS finalized these standards based on data it evaluated for 
young chickens from July 2007 through June 2008 and for young turkeys 
from August 2008 through July 2009. In this report, we use "young 
chickens" to refer to chickens raised for consumption that are less 
than 10 weeks of age at the time of slaughter and "young turkeys" to 
refer to turkeys raised for consumption that are less than 8 months of 
age at the time of slaughter. 

[20] In 1996, FSIS established the first Salmonella standard for young 
chicken carcasses, which set the expectation that no more than 20 
percent of a plant's young chicken carcasses will be contaminated with 
Salmonella. 61 Fed. Reg. 38806, 38867 (July 25, 1996) (codified at 9 
C.F.R. § 381.94(b)(1), rescinded by 79 Fed. Reg. 49566, 49636) (Aug. 
21, 2014). FSIS established this standard, as well as Salmonella 
standards for ground chicken and ground turkey, as part of a larger 
agency effort designed to reduce both the presence of poultry 
pathogens and the incidence of foodborne illnesses associated with the 
consumption of poultry products, among other things. In 2005, FSIS 
established its initial Salmonella standard for young turkey 
carcasses, which set the expectation that no more than 19.6 percent of 
a plant's young turkey carcasses will be contaminated with Salmonella. 
70 Fed. Reg. 8058 (Feb. 17, 2005). 

[21] The agency has set this maximum at a number that, when expressed 
as a percentage of the sample set, is higher than the percentage of 
allowable contamination stated in the standard.Therefore, if 5 out of 
the 51 (about 9.8 percent) samples collected by the agency actually 
test positive for Salmonella, the plant would still be in compliance 
even though it exceeds the agency's standard of 7.5 percent. According 
to FSIS, the number of samples tested and the number of positive 
results allowed from those samples were calculated to provide a plant 
with about an 80 percent chance of passing agency testing when the 
plant is operating just within the standards. Because the 51 samples 
are randomly selected, and the sample size is relatively small 
compared with total production, FSIS adopted this approach to minimize 
the chances of a false finding of noncompliance for plants that are 
operating within the agency's standard for young chicken carcasses. 

[22] FSIS, Compliance Guideline for Controlling Salmonella and 
Campylobacter in Poultry, 3RD ed. (May 2010). FSIS issued the updated 
guideline after initially announcing the agency's new Salmonella and 
Campylobacter standards in a May 2010 Federal Register notice. The 
guideline includes recommendations rather than regulatory requirements. 

[23] 77 Fed. Reg. 72686 (Dec. 6, 2012). According to FSIS research 
analysis documents, ground products may have a higher incidence of 
Salmonella than whole raw carcasses because the grinding process 
increases the surface area and distributes pathogen microorganisms, 
originally present only on the surface, throughout the product where 
the environment is favorable for their growth. 

[24] 79 Fed. Reg. 22084 (Apr. 21, 2014). 

[25] FSIS conducted a baseline survey of poultry plants producing raw 
chicken parts from January 2012 to August 2012 and used data collected 
from the samples to estimate prevalence of Salmonella in raw chicken 
parts. 

[26] 79 Fed. Reg. 22084 (Apr. 21, 2014). 

[27] The plan details a priority list of actions the agency plans to 
undertake as part of its continued efforts to address Salmonella in 
poultry products. 

[28] A risk assessment is a systematic way of preparing, organizing, 
and analyzing information to help make regulatory decisions, establish 
programs, and prioritize research and development efforts. 

[29] 71 Fed. Reg. 9772 (Feb. 27, 2006). 

[30] The chiller is the location where eviscerated carcasses--
carcasses that have had internal organs and any processing defects 
removed--are chilled in order to inhibit microbial growth and meet the 
regulatory requirements of 9 C.F.R. § 381.66(b)(1). 

[31] 76 Fed. Reg. 41186 (July 13, 2011). Examples of the types of 
waivers previously granted by FSIS include those made for on-line 
reprocessing, the HACCP-Based Inspection Models Project, or other 
waivers that impact slaughter processes. Previously, FSIS established 
additional program criteria to which plants participating in the 
Salmonella Initiative Program must adhere, including that the number 
of samples collected by FSIS or by plant personnel that tests positive 
for Salmonella needs to be maintained at or below half the acceptable 
number of positive samples under the agency's existing standard. 73 
Fed. Reg. 4767 (Jan. 28, 2008). FSIS based the program criteria on 
criteria initially developed in August 2007 for a prototype of the 
Salmonella Initiative Program. 

[32] 71 Fed. Reg. 9772 (Feb. 27, 2006). 

[33] 73 Fed. Reg. 4767, 4770 (Jan. 28, 2008). 

[34] 76 Fed. Reg. 15282 (Mar. 21, 2011). As noted above, FSIS started 
publishing monthly the names of young chicken slaughter plants that 
fell into Category 2 or Category 3 on the agency's website in March 
2008. However, starting in July 2011, following implementation of 
FSIS's revised Salmonella standards for young chicken and turkey 
carcasses, the agency ceased publishing on its website the names of 
plants in Category 2. 

[35] A food safety assessment is performed to assess the design and 
validity of food safety systems in a plant. Moreover, food safety 
assessments are conducted routinely and periodically and also "for 
cause" when prompted by sample results testing positive for pathogens, 
potential production and shipment of adulterated products, or any 
other high-priority food-safety-related incident. 

[36] The additional actions to be completed prior to the meeting 
include developing a document that, among other things, summarizes 
lessons learned from previous meetings discussing poultry and beef 
farm practices and continuing to work with members of industry on 
specific outbreaks to identify best practices they have developed for 
on-farm practices, according to FSIS's Salmonella Action Plan. 

[37] 79 Fed. Reg. 49566 (Aug. 21, 2014). The final rule will take 
effect on October 20, 2014. 

[38] We did not analyze the updated risk assessment as part of this 
report. For our August 2013 report, we were unable to determine if the 
results of the previous risk assessment conducted by FSIS to estimate 
the public health impact of the proposed rule accurately stated the 
public health benefits because the risk assessment did not include 
sufficient detail about its methodology. See [hyperlink, 
http://www.gao.gov/products/GAO-13-775]. 

[39] 77 Fed. Reg. 72686 (Dec. 6, 2012). The agency's HACCP regulations 
require that every plant reassess the adequacy of its HACCP plan at 
least annually and whenever any changes occur that could affect the 
hazard analysis or alter the HACCP plan. In the agency's view, the 
occurrence of these recent Salmonella outbreaks represented a change 
in the sanitary conditions involved in the manufacture of these 
products and therefore a change that could affect the hazard analysis 
or alter the HACCP plans for ground and mechanically separated chicken 
and turkey products. 

[40] 76 Fed. Reg. 15282 (Mar. 21, 2011). FSIS's new Campylobacter 
standards for young chicken and turkey carcasses took effect in July 
2011. 

[41] GAO, Homeland Security: An Overall Strategy Is Needed to 
Strengthen Disease Surveillance in Livestock and Poultry, [hyperlink, 
http://www.gao.gov/products/GAO-13-424] (Washington, D.C.: May 21, 
2013) and Performance Management and Evaluation: Definitions and 
Relationships, [hyperlink, http://www.gao.gov/products/GAO-11-646SP] 
(Washington, D.C.: May 2, 2011). 

[42] USDA, Food Safety and Inspection Service Strategic Plan Fiscal 
Year 2011-2016. 

[43] USDA, Fiscal Year 2013 Annual Performance Report. According to 
USDA, the target was set as an attainable target based on the agency's 
assessment of industry performance when it revised the standard in 
2011, and past agency experience with industry's response to 
Salmonella policies. 

[44] Institute of Medicine of the National Academies, Review of the 
Use of Process Control Indicators in the FSIS Public Health Risk-Based 
Inspection System: A Letter Report (Washington, D.C.: The National 
Academies Press, 2009). 

[45] According to FSIS documents, those plants in Category 1 are said 
to be demonstrating consistent process controls to meet the existing 
Salmonella standards while plants in Categories 2 and 3 are not 
maintaining consistent process controls. 

[46] As we have previously found, without performance measures, 
agencies cannot determine whether the activities they are carrying out 
are accomplishing the goals they intend to achieve. See GAO, Results 
Oriented Government: GPRA Has Established a Solid Foundation for 
Achieving Greater Results, [hyperlink, 
http://www.gao.gov/products/GAO-04-38] (Washington, D.C.: Mar. 10, 
2004). 

[47] As we have previously reported, having pertinent and reliable 
performance information available is necessary for Congress to 
adequately assess agencies' progress in making performance and 
management improvements and ensure accountability for results. GAO, 
Managing for Results: A Guide for Using the GPRA Modernization Act to 
Help Inform Congressional Decision Making, [hyperlink, 
http://www.gao.gov/products/GAO-12-621SP] (Washington, D.C.: June 15, 
2012). 

[48] According to USDA, Salmonella is the largest component of the 
measure, representing 92 percent of foodborne human illnesses from the 
three pathogens. The two other pathogens are Listeria and Escherichia 
coli O157: H7 (E. coli). 

[49] FSIS links the target level of performance to the Healthy People 
2020 goals. Healthy People 2020 provides a set of goals and objectives 
with 10-year targets, designed to guide national health promotion and 
disease prevention efforts, improving the health of all people in the 
United States. FSIS collaborated with the Department of Health and 
Human Services when the department developed human health goals for 
Salmonella and Campylobacter. The Healthy People 2020 goal for 
Salmonella illnesses is 11.4 cases per 100,000 people, an increase 
from 6.8 cases per 100,000 people--the Healthy People 2010 goal. For 
Campylobacter, the 2020 goal is 8.5 cases per 100,000 people. 

[50] The National Advisory Committee on Microbiological Criteria for 
Foods was established in 1988 to advise the Secretaries of Agriculture 
and Health and Human Services. The committee covers public health 
issues relative to the safety and wholesomeness of the U.S. food 
supply, including development of microbiological criteria, as well as 
evaluation of data and methodologies for assessing microbiological 
hazards in foods. The committee's work also assists the CDC and the 
Departments of Commerce and Defense. The committee is the outcome of 
recommendations from the National Academy of Sciences and U.S. House 
of Representatives Committee on Appropriations. 

[51] National Advisory Committee on Microbiological Criteria for 
Foods. "Response to Questions Posed by the Food Safety and Inspection 
Service Regarding Determination of the Most Appropriate Technologies 
for the Food Safety and Inspection Service to Adopt in Performing 
Routine and Baseline Microbiological Analyses," Journal of Food 
Protection 73, no. 6 (2010). 

[52] GAO, Food Safety: CDC Is Working to Address Limitations in 
Several of Its Foodborne Disease Surveillance Systems, [hyperlink, 
http://www.gao.gov/products/GAO-01-973] (Washington, D.C.: Sept. 7, 
2001). FSIS officials acknowledged that the all-illness performance 
measure depends on data from CDC, and these data have several 
limitations when used for foodborne illness attribution. For example, 
the data do not always identify the food product that caused the 
outbreak, data are not complete because CDC relies on states 
voluntarily reporting outbreak data, and outbreak data are not readily 
available due to delayed reporting. 

[53] In 2011, FSIS, CDC, and FDA formed the chartered Interagency Food 
Safety Analytics Collaboration which is focused on analytic 
collaboration on foodborne illness attribution. The group operates 
with the understanding that data improvements and development of 
multiple analytic methods are needed to generate good estimates across 
the broad range of food commodities and along all points in the food 
supply chain. 

[54] FSIS, A comparison of Salmonella serotype incidence in FSIS-
regulated products and salmonellosis cases, February 2012. 

[55] FSIS, Use of FSIS Regulatory Verification Sampling to Generate 
Prevalence Estimates, April 2012. 

[56] Ebel, et.al. "Simplified framework for predicting changes in 
public health from performance standards applied in slaughter 
establishment," Food Control 28 (2012), 250-257. 

[57] We identified limited enforcement authority as a potential 
challenge, rather than a challenge, because FSIS officials told us 
that the agency has tools to overcome enforcement authority 
limitations. 

[58] Established in 1971, the National Advisory Committee on Meat and 
Poultry Inspection advises the Secretary of Agriculture on matters 
affecting federal and state inspection program activities. National 
Advisory Committee on Meat and Poultry Inspection, Final Report - 
NACMPI Recommendations for FSIS on Pre-Harvest Salmonella (Washington, 
D.C.: September 2011). 

[59] However, after poultry products leave a plant, FSIS has authority 
to ensure that poultry products are correctly labeled and packaged. 

[60] Cross-contamination occurs when bacteria spread from a food to a 
surface, from a surface to another food, or from one food to another. 

[61] Federal regulations require plants to conduct a hazard analysis 
to determine food safety hazards "reasonably likely to occur" in the 
production process and identify the preventive measures the plant can 
apply to control those hazards. According to FSIS directive 5000.6, 
Rev.-1, all plants must conduct a hazard analysis that is an 
evaluation by a plant of its operations to determine the food safety 
hazards specific to the plant's operations that, if not controlled, 
are reasonably likely to occur and to cause injury or illness. Some 
plants may not have a HACCP plan because they can support that there 
is not a food safety hazard that is reasonably likely to occur; 
instead, the plants would maintain a record of their hazard analysis 
for inspection purposes. 

[62] In addition, CDC officials told us its data show that, in 2002, 6 
years after the HACCP/Pathogen Reduction Rule had been introduced, 
they found no change in the incidence of E. coli O157: H7 and that CDC 
was investigating many E. coli O157: H7 outbreaks due to contaminated 
ground beef. In that year, FSIS noticed that HACCP plans from meat 
grinding operations rarely included E. coli O157: H7 as an explicitly 
named hazard. Later that year, FSIS directed all U.S. meat grinders to 
declare E. coli O157: H7 explicitly in their HACCP plans. According to 
CDC officials, in 2003 there was a major drop in E. coli O157: H7 
infections and FSIS noted a significant drop in frequency of 
contamination of ground beef with E. coli O157: H7. 

[63] 79 Fed. Reg. 49566 (Aug. 21, 2014). Enteric pathogens are 
bacteria that reside normally in the guts of many animals, including 
humans, and some are pathogenic, causing disease in certain animal 
species. 

[64] FSIS defines an adulterant in meat and poultry products to 
include "any substance that contains any poisonous or deleterious 
substance which may render it injurious to health." 

[65] Supreme Beef Processors v. U.S. Dep't of Agric., 113 F. Supp. 2d 
1048 (N.D. Tex. 2000). A federal appeals court upheld this decision in 
2001. The appeals court held that USDA's Salmonella standards conflict 
with the agency's authority to enforce the prohibition against meat or 
meat products prepared, packed, or held in insanitary conditions and 
the agency could not use Salmonella verification tests results on 
products alone to determine whether a plant was in compliance with 
sanitary requirements. Supreme Beef Processors v. U.S. Dep't of Agric. 
275 F. 3d 432 (5th Cir. 2001). On a day-to-day basis, if an FSIS 
inspector finds plant conditions or procedures insanitary, the 
inspector can refuse to perform inspection and temporarily suspend the 
plant's operation until the problem is corrected. 

[66] As previously noted, FSIS's final rule for modernizing poultry 
slaughter inspection requires plants to develop, implement, and 
maintain written procedures to prevent contamination by Salmonella and 
Campylobacter. 

[67] The carcass of a whole raw chicken or turkey that failed 
inspection would undergo a cutting process for removal of the part of 
the whole bird that caused inspection failure, and the part is 
condemned as contaminated. 

[68] In contrast, raw beef products contaminated with any level of E. 
coli O157: H7 are not permitted to enter commerce since this pathogen 
is classified as an adulterant. FSIS declared this serotype of E. coli 
an adulterant in beef--the first time a foodborne pathogen on raw 
product was declared an adulterant under meat inspection law--after 
the 1992-1993 Jack-in-the-Box incident, which sickened more than 500 
people and killed four children. The symptoms for E. coli vary for 
each person but often include severe stomach cramps, diarrhea, and 
vomiting. 

[69] On July 31, 2014, FSIS announced that the agency denied the 
Center for Science in the Public Interest's petition to declare 
antibiotic-resistant Salmonella as an adulterant. On May 28, 2014, the 
Center for Science in the Public Interest filed a lawsuit to compel 
FSIS to respond to a petition it submitted on May 25, 2011, that 
requested FSIS declare specific antibiotic-resistant patterns of 
Salmonella as adulterants when found in ground poultry. The four 
strains of antibiotic-resistant Salmonella subject to the petition are 
Salmonella Hadar, Salmonella Heidelberg, Salmonella Newport, and 
Salmonella Typhimurium. The Center for Science in the Public Interest 
is a nonprofit, nongovernmental consumer advocacy organization focused 
primarily on nutrition, health, and food safety issues. 

[70] A strain refers to a specific biologic version of a 
microorganism, such as bacteria, within a specific serotype defined by 
its genetic makeup identified through DNA fingerprinting. 

[71] GAO, Food Safety: USDA and FDA Need to Better Ensure Prompt and 
Complete Recalls of Potentially Unsafe Food, [hyperlink, 
http://www.gao.gov/products/GAO-05-51] (Washington, D.C.: Oct. 7, 
2004). 

[72] If FSIS personnel have reason to believe that a meat or poultry 
product may be associated with human illnesses, but they cannot 
identify a specific product that FSIS could recommend be recalled, 
then FSIS can issue a public health alert to notify the public 
including actions to take to avoid illnesses. 

[73] Such recalls are voluntary actions taken by plants under the 
direction of their company. When companies discover that they may have 
distributed food that is contaminated with disease-causing bacteria, 
or that contains allergens that can cause serious illness or death, 
they will usually conduct a voluntary recall. That is, they will 
contact their customers and instruct them to contact the wholesalers, 
retailers, and others in the food's distribution chain and ask them to 
return or destroy the potentially unsafe food. If a plant conducts a 
recall, FSIS provides assistance and monitors the recall. If a plant 
does not conduct a recall that FSIS has requested, FSIS is limited to 
its authority to detain and seize the products in question. 

[74] FSIS's recall committee determines the reason a particular 
product may need to be removed from commerce and whether there is a 
statutory basis to recommend a voluntary recall. 

[75] FSIS applies the official mark of inspection to all meat and 
poultry products produced under federal inspection. The presence of 
the official mark of inspection indicates that the product was 
produced in compliance with federal regulations and under federal 
oversight. 

[76] This data collection and analyses is known as a baseline study. 
The most recent baseline study is on chicken parts: FSIS, "The 
Nationwide Microbiological Baseline Data Collection Program: Raw 
Chicken Parts Survey" (Washington, D.C.: January 2012-August 2012). 

[77] The agency is using the results from the new testing approach to 
revise its performance standards for raw ground poultry products and 
create new standards for mechanically separated poultry. 

[78] A critical control point is a point, step, or procedure in a food 
process at which control can be applied and, as a result, a food 
safety hazard can be prevented, eliminated, or reduced to acceptable 
levels. 

[79] The Codex Alimentarius Commission is regarded as a key body for 
international food related standard setting activities. Codex was set 
up jointly by the Food and Agriculture Organization of the United 
Nations and the World Health Organization in 1963. 

[80] FSIS is the agency responsible for inspections at all meat and 
poultry slaughter processing plants and for ensuring plants' 
compliance with regulatory requirements. 

[81] The Government Performance and Results Act of 1993 as amended by 
the GPRA Modernization Act of 2010 (GPRA) requires executive agencies 
to complete strategic plans in which they define their missions, 
establish results-oriented goals, and identify the strategies that 
will be needed to achieve those goals. GPRA also requires agencies to 
complete annual performance plans that establish performance goals--
which contribute to the strategic goals--and measure performance 
toward achieving performance goals. Performance measures are important 
management tools that help agencies monitor and report progress toward 
their goals. Numerical targets are a key attribute of successful 
performance measures because they allow managers to compare planned 
performance with actual results. 

[82] APHIS oversees voluntary programs to evaluate and certify that 
poultry are free of certain diseases. 

[83] GAO, Food Safety: More Disclosure and Data Needed to Clarify 
Impact of Changes to Poultry and Hog Inspections, [hyperlink, 
http://www.gao.gov/products/GAO-13-775] (Washington, D.C.: Aug. 22, 
2013). 

[84] In two of the other Salmonella outbreaks, state departments of 
health investigating the outbreaks issued public health notices 
independent of FSIS. No health alerts were issued in response to the 
Campylobacter outbreak. 

[End of section] 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation, and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the 
performance and accountability of the federal government for the 
American people. GAO examines the use of public funds; evaluates 
federal programs and policies; and provides analyses, recommendations, 
and other assistance to help Congress make informed oversight, policy, 
and funding decisions. GAO's commitment to good government is 
reflected in its core values of accountability, integrity, and 
reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's website [hyperlink, http://www.gao.gov]. Each 
weekday afternoon, GAO posts on its website newly released reports, 
testimony, and correspondence. To have GAO e-mail you a list of newly 
posted products, go to [hyperlink, http://www.gao.gov] and select 
"E-mail Updates." 

Order by Phone: 

The price of each GAO publication reflects GAO's actual cost of 
production and distribution and depends on the number of pages in the 
publication and whether the publication is printed in color or black 
and white. Pricing and ordering information is posted on GAO's 
website, [hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or 
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card, 
MasterCard, Visa, check, or money order. Call for additional 
information. 

Connect with GAO: 

Connect with GAO on facebook, flickr, twitter, and YouTube.
Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts.
Visit GAO on the web at [hyperlink, http://www.gao.gov]. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 
Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; 
E-mail: fraudnet@gao.gov; 
Automated answering system: (800) 424-5454 or (202) 512-7470. 

Congressional Relations: 

Katherine Siggerud, Managing Director, siggerudk@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, DC 20548. 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, DC 20548. 

[End of document]