Food Safety: Risk-Based Inspections and Microbial Monitoring Needed for
Meat and Poultry (Letter Report, 05/19/94, GAO/RCED-94-110).
The federal inspection system falls short in protecting the public from
the most serious health risks caused by microbial contamination.
Resources that could be more effectively used in a risk-based system are
drained away by labor-intensive inspection procedures and inflexible
inspection schedules. Under current law, federal inspectors must examine
each carcass slaughtered--nearly 7 billion birds and livestock
annually--and visit each of the 5,900 processing plants at least once
during each operating shift. During these inspections, inspectors rely
on their sense of smell, touch, and feel to make judgments about disease
conditions, contaminations, and contamination. However, these methods
cannot spot microbial contamination. The government must move to a
modern, scientific, risk-based inspection system that targets resources
toward higher-risk meat and poultry products. Meat processing plants
that have started microbial testing programs have used the test results
to identify problems and made changes to improve the safety of their
products. However, the government has not supported this effort by
designing generic programs or disseminating information gained from
individual testing programs. A Hazard Analysis and Critical Control
Point system is generally considered the best available approach for
ensuring safe foods because it seeks to prevent contamination in the
first place. The government plans to require meat and poultry plants to
use such a system but has yet to determine (1) whether microbial testing
will be required or (2) who should do it--the plants or government
inspectors.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: RCED-94-110
TITLE: Food Safety: Risk-Based Inspections and Microbial
Monitoring Needed for Meat and Poultry
DATE: 05/19/94
SUBJECT: Food inspection
Meat inspection
Safety regulation
Poultry inspection
Safety standards
Contaminated foods
Meat packing industry
Human resources utilization
Testing
Consumer protection
IDENTIFIER: FSIS Hazard Analysis and Critical Control Point System
**************************************************************************
* This file contains an ASCII representation of the text of a GAO *
* report. Delineations within the text indicating chapter titles, *
* headings, and bullets are preserved. Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are *
* identified by double and single lines. The numbers on the right end *
* of these lines indicate the position of each of the subsections in the *
* document outline. These numbers do NOT correspond with the page *
* numbers of the printed product. *
* *
* No attempt has been made to display graphic images, although figure *
* captions are reproduced. Tables are included, but may not resemble *
* those in the printed version. *
* *
* A printed copy of this report may be obtained from the GAO Document *
* Distribution Facility by calling (202) 512-6000, by faxing your *
* request to (301) 258-4066, or by writing to P.O. Box 6015, *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time. *
**************************************************************************
Cover
================================================================ COVER
Report to Congressional Requesters
May 1994
FOOD SAFETY - RISK-BASED
INSPECTIONS AND MICROBIAL
MONITORING NEEDED FOR MEAT AND
POULTRY
GAO/RCED-94-110
Meat and Poultry Inspection
Abbreviations
=============================================================== ABBREV
CDC - Centers for Disease Control
FSIS - Food Safety and Inspection Service
GAO - General Accounting Office
HACCP - Hazard Analysis and Critical Control Point
USDA - U.S. Department of Agriculture
Letter
=============================================================== LETTER
B-256276
May 19, 1994
The Honorable Charles W. Stenholm
Chairman, Subcommittee on Department
Operations and Nutrition
Committee on Agriculture
House of Representatives
The Honorable Harold L. Volkmer
Chairman
The Honorable Steve Gunderson
Ranking Minority Member
Subcommittee on Livestock
Committee on Agriculture
House of Representatives
Harmful microorganisms in food, such as some types of E. coli and
salmonella, cause millions of illnesses each year and cost
government, industry, and consumers billions of dollars. Microbial
contamination is widely recognized as today's most serious public
health risk associated with meat and poultry. The Food Safety and
Inspection Service (FSIS), within the U.S. Department of Agriculture
(USDA), spent about $558 million and 10,750 staff years in fiscal
year 1993 to inspect meat and poultry, relying primarily on visual
methods that may ensure a clean-looking plant environment and meat
and poultry products free of visible adulteration but cannot detect
microbial contamination. FSIS has recognized the need to modernize
its inspection system since at least 1985, when the National Academy
of Sciences recommended that FSIS intensify its efforts to control
and eliminate contamination from microorganisms that cause disease in
humans.\1
Concerned about the effectiveness of FSIS' meat and poultry
inspection system, you asked us to evaluate whether (1) the system
makes the most effective use of its resources to ensure food safety,
(2) meat and poultry plants have programs to test for microorganisms,
and (3) a quality control concept known as Hazard Analysis and
Critical Control Point is an effective approach for ensuring food
safety. As appendix I shows, since 1969 we and other organizations
have issued numerous reports and testimony describing the limitations
of the current inspection system. In particular, during the last few
years, we have emphasized the need to build a scientific, risk-based
meat and poultry inspection system to better protect the public from
foodborne illnesses. This report includes detailed information, not
previously available, that we developed on FSIS' use of inspection
resources and the extent of microbial testing in selected meat and
poultry plants.
--------------------
\1 Meat and Poultry Inspection: The Scientific Basis of the Nation's
Program, National Research Council, National Academy of Sciences
(1985).
RESULTS IN BRIEF
------------------------------------------------------------ Letter :1
FSIS' meat and poultry inspection system does not efficiently and
effectively use its resources to protect the public from the most
serious health risks associated with meat and poultry--microbial
contamination. The system--originally designed around the turn of
the century to protect against health threats from diseased
animals--is hampered by inflexible legal requirements and relies on
outdated, labor-intensive inspection methods. Under current law,
federal inspectors must examine each carcass slaughtered--nearly 7
billion birds and livestock annually--and visit each of the
approximately 5,900 processing plants at least once during each
operating shift. During these inspections, FSIS inspectors rely on
their senses (smell, touch, and feel) to make judgments about disease
conditions, contamination, and sanitation. However, these
inspections, which consumed about two-thirds of FSIS' 10,750 staff
year budget in fiscal year 1993, cannot detect microbial
contamination.
While FSIS does not routinely test for microbial contaminants or
require industry to conduct such tests, some plants do so to ensure
the safety and quality of their products. Of 157 plants we
contacted, 76 conducted periodic microbial testing to monitor the
level of microorganisms on equipment and products and in the plant
environment.\2 On the basis of the results of their monitoring, 74 of
the 76 plants have made changes to their facilities and/or operations
to improve product safety and quality. However, because each plant
developed its own testing program, with little or no assistance from
FSIS, the programs varied considerably in their sampling methodology
and criteria for evaluating test results. As a result, the plants'
microbial monitoring programs vary in their effectiveness because
some plants could be testing the wrong things or using criteria for
evaluating microbial test results that are too lenient. Furthermore,
while most of the larger plants we contacted have testing programs,
only a few of the smaller ones do because the programs are considered
costly, and FSIS does not provide assistance or guidance on how to
develop and implement such programs.
A Hazard Analysis and Critical Control Point (HACCP) system is
generally recognized as the best currently available approach for
ensuring safe foods because it focuses on preventing contamination
rather than detecting contamination once it has occurred. FSIS plans
to issue a proposed regulation in 1994 requiring that each meat and
poultry plant develop and implement a Hazard Analysis and Critical
Control Point system, but it has not yet determined (1) whether
microbial testing will be required as part of such a system or (2)
who should do it--the plants or FSIS inspectors. If such testing is
done, regardless of who does it, FSIS will need to develop guidelines
for evaluating the results of such tests.
--------------------
\2 Although the 157 plants were selected to obtain a cross section of
all federally inspected meat and poultry plants, they cannot be used
to make statements about the entire meat and poultry industry because
they were judgmentally selected.
BACKGROUND
------------------------------------------------------------ Letter :2
At the turn of the century, Upton Sinclair's The Jungle raised a
public outcry about contagious animal diseases, unsanitary
conditions, deceptive practices, and lax government inspection at
meat packing plants. The Congress responded to this outcry by
passing the Federal Meat Inspection Act in 1907. This act and a
subsequent poultry act require federal inspection of meat and poultry
products to ensure that they are safe, wholesome, and correctly
labeled and packaged.
To achieve these objectives, the acts require that during slaughter
operations, each individual animal carcass be examined by an on-line
FSIS inspector. During this "post mortem" inspection, largely
unchanged for 87 years, inspectors make judgments about disease
conditions, abnormalities, and contamination in carcasses on the
basis of what they see, feel, and smell--a process known as
organoleptic inspection. (See app. II for a detailed description of
these inspection procedures.)
Meat and poultry from government-inspected carcasses can be inspected
again during further processing. (Processing operations can include
simple cutting and grinding operations, complex canning procedures,
or preparation of ready-to-eat products.) Under the meat and poultry
inspection acts, FSIS is to inspect all processed products in order
to ensure that they are wholesome, not adulterated, and properly
labeled. FSIS has long implemented these statutory responsibilities
through daily inspections, under which all meat and poultry
processing plants are visited by an inspector at least once during
each operating shift. That is, plants with one shift are visited at
least once daily by an FSIS inspector, and plants with two shifts are
visited at least twice daily. Also, plants that operate overtime
generally receive an additional separate inspection visit. During
each plant visit, a processing inspector may spend from 15 minutes to
several hours performing various inspection duties, based primarily
on organoleptic methods.
Nevertheless, the safety of meat and poultry remains a concern.
Because many cases of foodborne illness go undiagnosed or unreported,
the actual extent of the problem is unknown with estimates varying
widely from 6.5 million to more than 80 million cases annually.
Moreover, according to the Centers for Disease Control, meat and
poultry products have been recognized as an important source of
foodborne disease. In economic terms, USDA estimates that the annual
cost of foodborne illness in the United States ranges from $5.2
billion to $6.1 billion, with more than one-half of the costs, or
$3.9 billion to $4.3 billion, of the illnesses attributable to meat
and poultry products.
FSIS IS UNABLE TO USE ITS
RESOURCES EFFECTIVELY
------------------------------------------------------------ Letter :3
Because of inflexible statutory inspection requirements and
labor-intensive inspection procedures that are of limited value in
detecting microbial pathogens, FSIS is not able to target its
resources on the principal health risk associated with meat and
poultry--microbial contamination.\3 Moreover, the usefulness of FSIS'
current approach is likely to diminish further because FSIS' current
resources cannot keep pace with industry growth. We estimate that
FSIS allocated about two-thirds of its 10,750 staff year budget in
fiscal year 1993 to comply with statutory requirements--about 47
percent to examine every carcass and about 20 percent to inspect each
processing plant at least once daily. (See app. III for details on
FSIS' resource use.)
--------------------
\3 In Food Safety: Building a Scientific, Risk-Based Meat and
Poultry Inspection System (GAO/T-RCED-93-22, Mar. 16, 1993), we
testified that statutory mandates restrict FSIS' flexibility to
respond to changes in risk, organoleptic inspections are not capable
of detecting microbial pathogens, and labor-intensive inspection
procedures drain resources from the development of a risk-based
system.
MANDATED SLAUGHTER
INSPECTIONS USE SUBSTANTIAL
RESOURCES BUT PROVIDE
QUESTIONABLE BENEFITS
---------------------------------------------------------- Letter :3.1
FSIS annually allocates over 5,000 staff years, or 47 percent of its
total staff year budget, to meet the legal requirement that it
examine every carcass. In 1992, FSIS inspectors, using organoleptic
methods, examined 126 million cattle, swine, sheep, horse, and lamb
carcasses and 6.8 billion poultry carcasses.
The impact on FSIS' resources of inspecting every carcass is best
illustrated by the number of on-line inspectors needed to inspect the
6.8 billion birds slaughtered in 1992. At the fastest line speeds,
an inspector has about 2 seconds to visually examine the inside and
outside surfaces of each bird and feel the eviscerated internal
organs. About 2,100 full-time inspectors are needed to carry out
these inspections--1,830 regularly assigned inspectors and 260 relief
inspectors.\4
Inspection needs are likely to increase. While meat production has
been relatively constant since 1981, poultry production has increased
by 4 percent a year. This rate of growth is expected to continue for
at least the next several years, increasing poultry production by 1.5
billion birds to 8.3 billion birds in 1997. Another 460 inspectors
will be required by 1997 to keep pace with the increased production.
But it is uncertain where these additional resources will come from.
In line with federal initiatives to control spending, FSIS' staff
resources have remained relatively constant since 1981, and FSIS has
said that it currently needs about 300 more inspectors to meet even
today's requirements.
While FSIS could reduce its resource requirements by giving
inspectors less time to examine each carcass, some experts already
have questioned the effectiveness of an inspector who examines 12,000
or more birds a day under current line speeds. On the other hand,
the inspectors' union and consumer groups believe that inspectors
would be more effective if slaughter lines were slowed and inspectors
were given more time to inspect each carcass. However, such an
approach could significantly increase FSIS' resource requirements.
For example, a modest increase of 1 second in the inspection time for
poultry, from 2 seconds to 3 seconds per carcass, would increase the
number of inspectors needed by 50 percent, requiring FSIS to hire
another 1,030 inspectors just to meet current production levels.
Moreover, experts have increasingly questioned the public health
benefits of the organoleptic inspection that FSIS relies on.
According to a 1985 National Academy of Sciences report, while
organoleptic inspection serves its original purpose of protecting
consumers from grossly visible lesions or diseases, it cannot
identify microbial pathogens--today's principal health risk.
Similarly, an October 1993 conference of the World Congress on Meat
and Poultry Inspection, an international association of government
regulators from meat trading countries, concluded that post-mortem
organoleptic inspection must be changed because (1) it wastes
resources and cannot detect microbial pathogens, (2) the animal
diseases for which it was originally designed have been eradicated in
many countries, and (3) it results in unnecessary cross-contamination
because the hands-on inspection techniques used virtually ensure that
contamination is spread from one carcass to another.
In our March 1993 testimony, we concluded that while careful
organoleptic examination of some animals, such as old dairy cows, may
still be needed, the benefit of such inspections for the young,
market animals that account for the vast majority of slaughtered
animals is less certain.
--------------------
\4 These figures are GAO's calculations based on FSIS' data. Relief
inspectors are needed because inspectors must be present at each
inspection station during slaughter operations. Therefore, FSIS uses
relief inspectors to cover the inspection stations when regularly
assigned inspectors are on break or leave or are away for training.
PROCESSING PLANT INSPECTIONS
ARE INEFFICIENT AND NOT
BASED ON RISK
---------------------------------------------------------- Letter :3.2
FSIS annually allocates about 2,200 staff years, or 20 percent of its
total staff years, to inspect about 5,900 meat and poultry processing
plants at least once during each operating shift and again during
overtime operations, under current law. On average, more than 1,300
of these 5,900 plants require a second or third FSIS inspection each
day because the plants run second shifts and/or overtime operations.
This inflexible daily inspection requirement is labor-intensive and
costly because FSIS inspectors must visit thousands of plants daily
regardless of the potential health risk involved. Because most
processing plants are small, producing less than 1 million pounds of
product a year, and do not require a full-time inspector, a "patrol"
assignment is established in which one inspector is responsible for
several plants. Most of FSIS' 1,400 patrol assignments comprise
three to six plants and thus require considerable travel time and
transportation cost. FSIS estimates that each patrol inspector
spends, on average, about 80 minutes per day traveling between
plants. This equals about 240 staff years, or 11 percent of the
total processing staff years, and costs FSIS about $8.1 million
annually based on an average yearly salary of $33,800 for a
processing inspector. In addition, about $11 million is spent
annually to reimburse inspectors for use of their cars and for other
travel expenses associated with patrol assignments.
To redirect FSIS' inspections toward firms and food processes that
pose the greatest risk, the Congress passed the Processed Products
Inspection Improvement Act of 1986, which amended the requirements on
inspection frequency for meat processing plants. For a 6-year
period, the act authorized FSIS to use its own discretion to
determine the frequency of inspection. However, except for three
limited pilot tests, FSIS did not implement its discretionary
authority, which lapsed on November 10, 1992.
In 1977 and again in 1992, we reported that inspection resources
could be used more efficiently and effectively if FSIS tailored the
frequency and intensity of inspection to the potential risks
associated with individual processing plants.\5 We concluded that
discretionary inspection could lead to safer products and help reduce
costs because scarce federal inspection resources would be redirected
from low-risk operations to areas that may need greater coverage
because they present a higher risk potential. In our 1992 report, we
asked the Congress to consider extending FSIS' discretionary
authority for processing plants.
--------------------
\5 Food Safety and Quality: Uniform, Risk-Based Inspection System
Needed to Ensure Safe Food Supply (GAO/RCED-92-152, June 26, 1992)
and A Better Way for the Department of Agriculture to Inspect Meat
and Poultry Processing Plants (GAO/RCED-78-11, Dec. 9, 1977).
SOME PLANTS INDEPENDENTLY TEST
FOR MICROBIAL CONTAMINANTS
------------------------------------------------------------ Letter :4
FSIS does not routinely test for microbial contamination nor does it
require industry to do so. Consequently, FSIS does not keep a list
of meat and poultry plants with microbial testing programs.
Therefore, to obtain information on industry's microbial testing
programs (also referred to as "microtesting" in this report), we
contacted 157 meat and poultry plants judgmentally selected from
various regions of the country. We worked with FSIS staff to select
a group of plants representative of all federally inspected meat and
poultry plants. These included large and small slaughter and
processing plants. When possible, we obtained documentation of the
plants' testing programs or reviewed the plants' documents when such
information was considered proprietary by plant officials and
therefore not releasable. Nevertheless, since the plants were
judgmentally selected, the results of our work are limited to the
information collected from the 157 plants that we contacted and
cannot be used to make statements about any of the other federally
inspected plants or group of plants.
Of the 157 meat and poultry plants we contacted, 76 have implemented
testing programs to monitor the level of microorganisms on equipment
and products and in the plant environment, including 74 that have
used the test results to make changes to their processing procedures
or facilities. Plant officials said that these changes were aimed at
improving the safety and quality of their meat and poultry products.
For example, one plant found through microbial testing that it had
listeria--a pathogenic bacteria--in its product. With the help of a
commercial laboratory, the source of the problem was traced to a
slicer that was contaminating the product. The plant replaced the
slicer and changed cleaning procedures, eliminating the listeria
problem. (App. IV describes the plants' microbial testing programs
and provides examples of the changes the plants have made because of
these programs.)
Larger plants--those producing more than 1 million pounds of meat and
poultry per year--were more likely to have testing programs than
smaller plants. Of the 97 larger plants that we contacted, 61 had
testing programs, compared with 15 of the 60 smaller plants that we
contacted. Plants without testing programs generally cited cost as
the main obstacle to adopting such programs. In addition to program
development costs, the operating costs for plants with microtesting
programs ranged from a low of $600 per year to a high of $750,000 per
year.
Plants encounter these costs, in part, because they cannot turn to
FSIS for assistance in program design and operations. Although FSIS
has a Division of Microbiology located in Washington, D.C., FSIS'
assistance to plants is generally limited to guidelines for
processing plants seeking authorization to substitute microbial
testing programs for cleaning the plant between shifts (mid-shift
cleanup), which is the current requirement. For example, FSIS
regional officials told us that they have received many calls from
plants inquiring about microbial testing programs. These officials
said that they refer these plants to industry associations because
FSIS is not set up to provide such assistance. While industry
associations can provide valuable assistance to plants establishing
or operating microbial testing programs, the extent of such
assistance varies among associations and not all meat and poultry
plants belong to an industry association.
To fill this void, plants seek assistance from commercial
laboratories or design their programs in-house. Therefore, the
sampling methodologies, type of tests performed, and test evaluation
criteria vary from plant to plant. For example, plant microbial
testing programs range from weekly sampling to determine the general
bacteria levels on equipment surfaces to daily sampling of equipment
surfaces and products to determine general bacteria levels as well as
to identify the presence of specific pathogens, such as listeria and
salmonella.
Plants also used different standards to evaluate the results of
microbial tests. For example, the strictest standards for general
bacteria levels allowed on equipment surfaces before operations begin
ranged from 0 to 500 colonies per square inch. Similarly, the level
of staphylococcus bacteria allowed on raw meat and poultry ranged
from 0 to 500 colonies per gram. While plant officials were
reluctant to endorse specific standards, they believed that guidance
from FSIS would be beneficial. Officials in 48 of the 76 plants with
microbial testing programs said that FSIS should provide guidelines
for evaluating test results. While FSIS has a general guide of no
more than 100 colonies per square inch for evaluating preoperational
test results from equipment, such information is not disseminated to
all meat and poultry plants. FSIS generally just provides this
information to plants seeking to substitute microbial testing
programs for mid-shift cleanup because FSIS is reluctant to
promulgate an industrywide guide until further research is conducted.
Furthermore, FSIS does not build upon the information developed by
the plants through microbial tests so that it can be informed and be
able to provide assistance when called upon. For example, FSIS did
not collect, analyze, or disseminate microbial testing information
developed by the 18 plants in which it had inspection tasks requiring
periodic review of the plants' testing programs. For the other 58
plants with testing programs, FSIS did not officially monitor the
programs in 43 plants, and information was not available to determine
if FSIS monitored the microbial testing programs in 15 plants.
Our findings on the wide variation in plant microtesting programs and
the lack of FSIS' assistance are consistent with those reported in
June 1992 by a Science Review Panel, established by the Secretary of
Agriculture to evaluate beef slaughter inspection methods.\6 Among
its major findings, the panel, which included veterinarians,
microbiologists, and other scientists, reported that in the plants it
visited, it found a great diversity in the microbiological sampling
and testing methodologies being used. The panel concluded that FSIS
should undertake the leadership role in the development of more
uniform methodologies and programs that will permit proper
comparisons of data and provide feedback for corrective actions.
More specifically, the panel reported that FSIS, in concert with
industry, should evaluate the effectiveness of proposed programs,
develop new and standardized methodologies when needed, design and
develop scientific data bases, and monitor the success of new
programs and technologies.
Without FSIS' support, plants are less likely to learn from each
other's experiences. As a result, plants spend time and resources
identifying and correcting problems already resolved by others. For
example, four plants we contacted found independently, through
microbial testing that one type of conveyor belt could not be
sanitized and therefore was likely to harbor microorganisms. Each
plant went through a laborious process of determining the source of
its high microbial counts--experimenting with different sanitizers
and evaluating employee's hygiene and work habits--before determining
that it needed to switch to a different type of conveyor belt.
Moreover, other plants that use this type of conveyor belt had not
been informed of the problems it presents.
--------------------
\6 Report of Comparative Review of USDA Streamlined Inspection System
for Cattle and Traditional Inspection Methods, Andrulis Research
Corporation (June 1992).
HACCP IS A GENERALLY ACCEPTED
APPROACH FOR ENSURING FOOD
SAFETY, BUT ROLE OF MICROBIAL
TESTING IS UNCERTAIN
------------------------------------------------------------ Letter :5
A Hazard Analysis and Critical Control Point system is generally
considered to be the best approach currently available for ensuring
safe foods because it focuses on preventing contamination rather than
detecting contamination once it has occurred. To strengthen
regulation of the industry and help ensure safer meat and poultry,
the Secretary of Agriculture announced in May 1993 that each meat and
poultry plant would be required to develop and implement a HACCP
system. While FSIS plans to publish its proposed HACCP requirements
in 1994, its plans to date do not specifically require microbial
testing to monitor plants' HACCP systems to ensure that they are
working effectively. Furthermore, FSIS has no plans to develop
guidelines for evaluating the results of microbial tests and
determining when remedial actions are needed.
HACCP IS CONSIDERED THE BEST
APPROACH FOR ENSURING THE
SAFETY OF MEAT AND POULTRY
---------------------------------------------------------- Letter :5.1
To prevent food safety problems before they occur, the HACCP approach
focuses on (1) identifying hazards and assessing risks associated
with each phase of food production,\7 (2) determining the critical
points where the identified hazards can be controlled, and (3)
establishing procedures to monitor these critical control points.
For example, during the slaughtering of cattle, hide removal is
considered a critical control point. If the hide, which is not
cleaned before slaughter, is not properly removed, it could
contaminate the carcass. Under a HACCP system, a plant quality
control employee could observe the hide removal process on a regular
basis, such as every hour, to determine that it was performed in
accordance with established procedures. If the proper procedures
were not being followed, the line would be immediately stopped and
corrective actions would be implemented. The determinations made
during the observations would be documented and retained for future
review by FSIS.
The scientific community has endorsed HACCP as an effective approach
for ensuring food safety. For example, during the 1980s, several
scientific panels convened by the National Academy of Sciences
recommended wider use of HACCP in food regulation, particularly for
the meat and poultry inspection program.\8 In addition, the National
Advisory Committee on Microbiological Criteria for Food has endorsed
HACCP as an effective and rational approach for ensuring food
safety.\9
--------------------
\7 Hazards include any biological, chemical, or physical property
that may cause an unacceptable consumer health risk.
\8 Meat and Poultry Inspection: The Scientific Basis of the Nation's
Program (1985) and Poultry Inspection: The Basis for a
Risk-Assessment Approach (1987), National Academy of Sciences.
\9 This committee, whose members include food safety and public
health experts from government, industry, and academia, was
established in 1987 to provide the U.S. Department of Agriculture
and the Departments of Health and Human Services, Commerce, and
Defense with advice and recommendations on the development of
microbiological criteria that could be used to assess the safety and
wholesomeness of food.
ROLE OF MICROBIAL TESTING IN
FSIS' PROPOSED HACCP
APPROACH IS UNCERTAIN
---------------------------------------------------------- Letter :5.2
Although the HACCP concept has wide support, consumer groups, the
inspectors' union, and other parties have raised various concerns
about how FSIS intends to implement HACCP in the meat and poultry
industry. These concerns include whether FSIS would (1) relinquish
too much responsibility for food safety to industry, (2) not require
FSIS' approval of plants' HACCP systems, (3) lack adequate authority
to access relevant plant records, (4) provide adequate HACCP training
to its inspectors, and (5) lack the authority to impose sanctions and
penalties on plants failing to comply with HACCP requirements.
Another concern that we have is whether a mandatory HACCP system will
include adequate microbial testing to ensure that microbial hazards
are controlled. FSIS' documents on HACCP provide no details on
whether such testing will be required.
The HACCP concept uses a two-step process to ensure its
effectiveness--evaluation of the individual critical control points
and an overall evaluation (called "verification") of the entire
system. Individual critical control points must be evaluated on a
real-time basis; that is, evaluation results must be immediately
available so that corrective action can be taken as soon as possible.
Real-time evaluation tools include physical observation and testing
for chemical residues. Microbial testing does not provide real-time
results because under today's technology, results are not available
for at least 24 hours, although FSIS has been encouraging the
development of quicker testing methodologies.\10
The HACCP concept also requires verification that a plant's overall
processing system is working, not just the individual control points.
This verification need not be done on a real-time basis but can rely
on, among other methods, testing samples of products taken at various
times throughout production. Microbial testing can provide this
overall verification, allowing judgments to be made on product safety
and alerting the plants to deficiencies in processing, distribution,
storage, or marketing.
FSIS recognizes the benefits of microbial testing but has not yet
determined if such testing will be required as an integral part of
plants' HACCP systems. FSIS officials said that they are continuing
to evaluate the need for microbial testing, including who should do
it--the plant or FSIS inspectors.
As would be expected, plants with microtesting programs that were
included in our survey were more in favor of making microtesting
mandatory than plants without such programs. For example, of the 76
plants that had microtesting programs, 53 said that such testing
should be mandatory, 16 said that it should remain voluntary, and 7
expressed no opinion. Of the 81 plants that did not do microtesting,
54 said that such testing should remain voluntary, 15 said that it
should be mandatory but generally believed the government should do
it, and 12 expressed no opinion.
If microbial testing is required, regardless of who does it,
guidelines will have to be developed to help plants or FSIS
inspectors determine when microbial test results should require
remedial action. Without guidelines, plants or FSIS inspectors will
have to rely on their own judgments on when to take action, which
would vary widely, as indicated above.
--------------------
\10 USDA's Agricultural Research Service is evaluating a rapid test
that determines general bacterial levels on meat within minutes.
CONCLUSIONS
------------------------------------------------------------ Letter :6
The federal inspection system is neither efficient nor effective in
protecting the public from the most serious health risks caused by
microbial contamination. Resources that could be more effectively
used in a risk-based system are drained away by labor-intensive
inspection procedures and inflexible inspection frequencies. For
example, FSIS continues to rely primarily on organoleptic inspection
procedures that are not capable of detecting harmful bacteria. In
addition, under current law, federal inspectors must examine each
carcass slaughtered and visit each processing plant at least daily.
To better protect the public from foodborne illnesses, FSIS must move
to a modern, scientific, risk-based inspection system. Such a system
would allow FSIS to target its resources towards the higher-risk meat
and poultry products by increasing inspection of such products,
developing methods or tools that would help inspectors detect
microbial contamination, and/or increasing the testing of such
products.
The plants that have initiated microbial testing programs have used
the test results to identify problem areas and made numerous changes
that were designed to improve the safety of their products. However,
FSIS, the federal agency responsible for overseeing the meat and
poultry industry and ensuring the safety of meat and poultry, has not
supported this effort by designing generic programs and/or
disseminating information gained from individual testing programs.
As a result, the investment required of plants interested in
improving their processes by implementing microbial testing programs
could be significant and thereby discourage such testing,
particularly in small plants.
The HACCP approach is generally considered the most effective
approach currently available for preventing microbial contamination.
FSIS, however, has not yet determined whether microbial testing will
be an essential component of HACCP requirements. Without specifying
testing requirements and criteria, FSIS cannot ensure that each
plant's HACCP system will effectively monitor microbial
contamination.
RECOMMENDATION
------------------------------------------------------------ Letter :7
To improve the safety of meat and poultry, we recommend that the
Secretary of Agriculture direct the Administrator, FSIS, to develop a
mandatory HACCP system that includes specific requirements for
microbial testing and guidelines for determining when microbial test
results warrant action by the plant. As part of this effort, the
Administrator should assist meat and poultry plants in the
development of their microbial testing programs by, among other
things, disseminating information on the programs already in
operation.
RECOMMENDATION TO THE CONGRESS
------------------------------------------------------------ Letter :8
We recommend that the Congress revise the meat and poultry acts to
provide FSIS with the flexibility and discretion to target its
inspection resources to the most serious food safety risks.
AGENCY COMMENTS AND OUR
EVALUATION
------------------------------------------------------------ Letter :9
We received written comments from the U.S. Department of Agriculture
(USDA) on a draft of this report. (See app. V.) USDA agreed that
(1) microbiological hazards are the major threat to public health
identified with the consumption of meat and poultry, (2) the current
inspection system must be shifted to a system based on science and
risk, (3) greater flexibility in adjusting resources to target the
most serious food safety risks may be important, and (4) HACCP holds
promise for preventing contamination. USDA also provided a
description of various initiatives undertaken during the past year to
improve the existing inspection system, including a proposed animal
identification and traceback program; nationwide, microbial baseline
studies to determine the presence and levels of pathogens on meat and
poultry; a final rule mandating labels on raw meat and poultry
describing safe handling techniques; and strengthened oversight and
regulatory enforcement.
In February 1994 testimony before a Senate agriculture subcommittee,
we provided our assessment of these same initiatives.\10 We concluded
that although USDA's efforts had produced some constructive changes,
USDA had not dealt with the inspection system's inherent weaknesses
nor fundamentally changed the system's reliance on sensory inspection
methods. USDA still has not mandated routine microbial testing by
industry or government inspectors nor sought legislative changes to
allow it to target its resources to the most serious food safety
risks.
--------------------
\10 Meat Safety: Inspection System's Ability to Detect Harmful
Bacteria Remains Limited (GAO/T-RCED-94-123, Feb. 10, 1994).
---------------------------------------------------------- Letter :9.1
We performed our review between April 1993 and January 1994, in
accordance with generally accepted government auditing standards. We
conducted our work primarily at FSIS headquarters in Washington,
D.C., and its Western Regional Office in Alameda, California. We
also contacted 157 meat and poultry plants to determine if they had
microtesting programs. The plants contacted were judgmentally
selected to provide the broadest coverage (that is, large and small
slaughter and processing plants in various sections of the country)
with the resources available. Because the selection process was
judgmental, the information obtained only applies to the 157 plants
that we contacted and cannot be used to make statements about the
universe of meat and poultry plants. (Further details on our
objectives, scope, and methodology are provided in app. VI.)
As arranged with your offices, unless you publicly announce its
contents earlier, we plan no further distribution of this report
until 30 days after the date of this letter. At that time, we will
send copies of this report to interested congressional committees and
the Secretary of Agriculture. We will also make copies available to
others upon request.
This work was performed under the direction of John W. Harman,
Director, Food and Agriculture Issues, who can be reached on (202)
512-5138. Major contributors to this report are listed in appendix
VII.
Keith O. Fultz
Assistant Comptroller General
GENERAL ACCOUNTING OFFICE AND
OTHER REPORTS SINCE 1969 ON THE
FEDERAL MEAT AND POULTRY
INSPECTION SYSTEM
=========================================================== Appendix I
GAO REPORTS
--------------------------------------------------------- Appendix I:1
Meat Safety: Inspection System's Ability to Detect Harmful Bacteria
Remains Limited (GAO/T-RCED-94-123, Feb. 10, 1994).
Food Safety: A Unified, Risk-Based System Needed to Enhance Food
Safety (GAO/T-RCED-94-71, Nov. 4, 1993).
Food Safety: Building a Scientific, Risk-Based Meat and Poultry
Inspection System (GAO/T-RCED-93-22, Mar. 16, 1993).
Food Safety: Inspection of Domestic and Imported Meat Should Be
Risk-Based (GAO/T-RCED-93-10, Feb. 18. 1993).
Food Safety and Quality: Uniform, Risk-based Inspection System
Needed to Ensure Safe Food Supply (GAO/RCED-92-152, June 26, 1992).
Food Safety and Quality: Salmonella Control Efforts Show Need for
More Coordination (GAO/RCED-92-69, Apr. 21, 1992).
Food Safety and Quality: FDA Needs Stronger Controls Over the
Approval Process for New Animal Drugs (GAO/RCED-92-63, Jan. 17,
1992).
U.S. Department of Agriculture: Improving Management of
Cross-Cutting Agricultural Issues (GAO/RCED-91-41, Mar. 12, 1991).
Food Safety and Quality: Who Does What in the Federal Government
(GAO/RCED-91-19A&B, Dec. 21, 1990).
Food Safety and Inspection Service's Performance-Based Inspection
System (GAO/T-RCED-89-53, July 31, 1989).
Internal Controls: Program to Address Problem Meat and Poultry
Plants Needs Improvement (GAO/RCED-89-55, Mar. 31, 1989).
Imported Meat and Livestock: Chemical Residue Detection and the
Issue of Labeling (GAO/RCED-87-142, Sept. 30, 1987).
Inspection Activities of the Food Safety and Inspection Service
(GAO/T-GGD-87-15, May 15, 1987).
Compendium of GAO's Views on the Cost Saving Proposals of the Grace
Commission, Vol. II--Individual Issue Analyses (GAO/OCG-85-1, Feb.
19, 1985).
Monitoring and Enforcing Food Safety--An Overview of Past Studies
(GAO/RCED-83-153, Sept. 9, 1983).
Improved Management of Import Meat Inspection Program Needed
(GAO/RCED-83-81, June 15, 1983).
Improving Sanitation and Federal Inspection at Slaughter Plants: How
to Get Better Results for the Inspection Dollar (CED-81-118, July 30,
1981).
A Better Way for the Department of Agriculture to Inspect Meat and
Poultry Processing Plants (CED-78-11, Dec. 9, 1977).
Selected Aspects of the Administration of the Meat and Poultry
Inspection Program (CED-76-140, Aug. 25, 1976).
Consumer Protection Would Be Increased by Improving the
Administration of Intrastate Meat Plant Inspection Programs
(B-163450, Nov. 2, 1973).
Consumer and Marketing Service's Enforcement of Federal Sanitation
Standards at Poultry Plants Continues to Be Weak (B-163450, Nov. 16,
1971).
Need to Reassess Food Inspection Roles of Federal Organizations
(B-168966, June 30, 1970).
Weak Enforcement of Federal Sanitation Standards at Meat Plants by
the Consumer and Marketing Service (B-163450, June 24, 1970).
Enforcement of Sanitary, Facility, and Moisture Requirements at
Federally Inspected Poultry Plants (B-163450, Sept. 10, 1969).
USDA OFFICE OF INSPECTOR
GENERAL REPORTS
--------------------------------------------------------- Appendix I:2
Food Safety and Inspection Service: Quality Control Programs (Audit
Report No. 24600-1-Ch, Nov. 18, 1993).
Food Safety and Inspection Service: Evaluation of Regulation of
Cornhusker Packing Company, Omaha, Nebraska (Audit Report No.
24800-1-KC, Aug. 1993).
Food Safety and Inspection Service: Monitoring of Drug Residues
(Audit Report No. 24600-1-At, Sept. 30, 1991).
Food Safety and Inspection Service: Labeling Policies and Approvals
(Audit Report No. 24099-5-At, June 1990).
Food Safety and Inspection Service: Follow-Up Audit of the Imported
Meat Process (Audit Report No. 38002-4-Hy, Mar. 29, 1989).
Food Safety and Inspection Service: Audit of the Imported Meat
Process (Audit Report No. 38002-2-Hy, Jan. 14, 1987).
Food Safety and Inspection Service: Meat and Poultry Inspection
Program (Audit Report No. 38607-1-At, Sept. 26, 1986).
STUDIES BY THE CONGRESS,
SCIENTIFIC ORGANIZATIONS, AND
OTHERS
--------------------------------------------------------- Appendix I:3
Meat and Poultry Inspection: Background and Current Issues,
Congressional Research Service, (Report No. 93-574 ENR, June 9,
1993).
Report of Comparative Review of USDA Streamlined Inspection System
for Cattle and Traditional Inspection Methods, Andrulis Research
Corp., (Contract No. 53-3A94-0-07, June 5, 1992).
Setting the Food Safety and Inspection Service on a Path to Renewal,
report of USDA's Management Evaluation Team, (Nov. 1991).
Cattle Inspection, Food and Nutrition Board, Institute of Medicine,
National Academy of Sciences, (1990).
Federal Poultry Inspection: A Briefing, Congressional Research
Service, (Report No. 87-432 ENR, May 8, 1987).
Food Safety Policy: Scientific and Regulatory Issues, Congressional
Research Service, (Order Code IB83158, Feb. 13, 1987).
Poultry Inspection: The Basis for a Risk-Assessment Approach,
National Research Council, National Academy of Sciences, (1987).
Meat and Poultry Inspection--The Scientific Basis of the Nation's
Program, National Research Council, National Academy of Sciences,
(1985).
Food Safety Policy Issues, Congressional Research Service, (Report
No. 81-155 SPR, June 1981).
Study on Federal Regulation, Regulatory Organization, Committee on
Governmental Affairs, U.S. Senate, vol. V, (Dec. 1977).
Study of the Federal Meat and Poultry Inspection System, Booz, Allen,
and Hamilton, Inc., (June 1977).
CARCASS-BY-CARCASS INSPECTION
========================================================== Appendix II
This appendix describes the Food Safety and Inspection Service's
(FSIS) inspection procedures conducted prior to (ante mortem) and
following slaughter (post mortem) of animals. The federal meat and
poultry inspection acts mandate that FSIS inspectors examine each
carcass as it moves through the plant. (The requirement to inspect
each individual carcass is often referred to as carcass-by-carcass or
bird-by-bird inspection.) Current inspection procedures, largely
unchanged for 87 years, rely on organoleptic techniques by which
inspectors use their sight, smell, and touch to determine whether
animals and carcasses are diseased or contaminated and unfit for
human consumption. Figures II.1 through II.8 illustrate cattle and
poultry slaughter processes and corresponding points at which FSIS
conducts inspection. We observed these slaughter processes and
corresponding inspection points on our visits to meat and poultry
plants. In addition, we had FSIS inspection operation officials
review this appendix for accuracy, and their suggestions have been
included, where appropriate.
Figure II.1: The Cattle
Slaughter Process and
Corresponding FSIS Inspection
Points
(See figure in printed
edition.)
CATTLE ANTE MORTEM INSPECTION
-------------------------------------------------------- Appendix II:1
Ante mortem inspection is intended to prevent animals that are
obviously unfit for human consumption because of abnormalities or
diseases, such as central nervous system disorders, from entering the
slaughter process. FSIS' ante mortem inspection for cattle includes
(1) observing the animals at rest and in motion and (2) examining
animals identified as abnormal or diseased to determine their
disposition. Figure II.2 shows an inspector viewing cattle in a
holding pen.
Figure II.2: Ante Mortem
Inspection of Cattle
(See figure in printed
edition.)
Source: USDA.
(See figure in printed
edition.)
CATTLE POST MORTEM INSPECTION
-------------------------------------------------------- Appendix II:2
The post mortem inspection is designed to identify any carcasses or
parts that are diseased or abnormal, thus rendering them unwholesome
or adulterated. For example, visual pathological conditions and
contamination, such as stomach or intestinal abnormalities, would be
reason to remove a carcass or portion thereof from further
production. FSIS has established three distinct points during the
slaughter process for inspection of cattle carcasses--head, viscera,
and final eviscerated carcass. Figures II.3 through II.5 illustrate
each inspection point.
Figure II.3: Head Examination
During Post Mortem Inspection
(See figure in printed
edition.)
Source: USDA.
(See figure in printed
edition.)
In head inspections, the inspector examines and cuts the eight lymph
nodes located in the head and the cheek muscles and observes and
palpates, that is, feels, the tongue.
Figure II.4: Viscera
Examination of the Heart,
Liver, and Digestive Tract
(See figure in printed
edition.)
Source: USDA.
(See figure in printed
edition.)
The inspector observes the carcass for contamination and the viscera
(internal organs) for obvious abnormalities. The inspector then cuts
and observes the heart and the bronchial and hepatic lymph nodes and
observes the cranial and caudal mesenteric lymph nodes. The
inspector palpates the surfaces of the lungs and liver and the
junction between two of the four stomachs.
Figure II.5: Carcass
Examination During Post Mortem
Inspection
(See figure in printed
edition.)
Source: USDA.
(See figure in printed
edition.)
The inspector examines the entire eviscerated carcass. This
inspection occurs after the carcass is split in half but before the
carcass halves are washed. The inspector observes the carcass
exterior, interior, and cut surfaces of the muscles and bones and
observes and palpates the kidneys and the diaphragm.
Figure II.6: The Poultry
Slaughter Process and
Corresponding FSIS Inspection
Points
(See figure in printed
edition.)
POULTRY ANTE MORTEM INSPECTION
-------------------------------------------------------- Appendix II:3
The purpose of ante mortem inspection for poultry is the same as that
for red meat animals--to prevent animals that are obviously unfit for
human consumption because of abnormalities or diseases from entering
the slaughter process. However, for poultry FSIS inspectors
generally perform ante mortem inspection on a lot-basis while the
birds remain caged, as shown in figure II.7. As plant employees
remove the birds from the cages, they will also identify and withhold
from slaughter any dead and abnormal appearing poultry. Later, a
veterinarian examines any abnormal or diseased birds. Dead birds are
condemned and kept separate from poultry intended for human
consumption.
Figure II.7: Ante Mortem
Inspection of Poultry
(See figure in printed
edition.)
Source: USDA.
(See figure in printed
edition.)
POULTRY POST MORTEM INSPECTION
-------------------------------------------------------- Appendix II:4
The plant is responsible for proper presentation of the carcass to
the inspector. The viscera must be removed from the body cavity with
care to avoid fecal or other contamination. The inspector then
observes the inside of the body cavity, including internal organs,
tissues, and body walls. The inspector will next examine the
suspended visceral organs, including the spleen, heart, and liver.
The inspector observes the carcass for skin tumors, bruises, or other
process defects as the bird approaches and again as the bird departs
from the inspector's view. Figure II.8 shows an inspector examining
the internal organs of a bird.
Figure II.8: Post Mortem
Inspection of Poultry
(See figure in printed
edition.)
Source: USDA.
(See figure in printed
edition.)
FSIS INSPECTION RESOURCES
========================================================= Appendix III
This appendix provides information on how FSIS allocates and uses its
resources to operate the meat and poultry inspection system mandated
by legislation. Overall, our analysis of FSIS' budget, personnel,
and staffing data showed the following highlights.
Since 1981, FSIS' budget, adjusted for inflation, as well as its
staff resources have remained relatively constant.
Two-thirds of FSIS' 10,750 total agency staff years in 1993 are
dedicated to meeting inspection requirements based on
legislation. (About 47 percent is allocated for
carcass-by-carcass inspection and another 20 percent is
allocated to daily inspection of processing plants.)
Absent hiring more inspectors or increasing inspection line speeds,
the growth in poultry production will continue to stress FSIS'
resource use.
FSIS BUDGET REMAINS CONSTANT
WHEN ADJUSTED FOR INFLATION
------------------------------------------------------- Appendix III:1
Since fiscal year 1981, FSIS' budget has grown from a total of $331
million (federal and nonfederal funds) to an estimated $558 million
for fiscal year 1993. (See fig. III.1.) However, when adjusted for
inflation, the budget has fluctuated only slightly between $519
million in fiscal year 1981 to $558 million for fiscal year 1993.
(See fig. III.2.) We begin our analysis in 1981 because (1) FSIS was
created in June 1981 and (2) according to FSIS officials, the agency
functions have remained essentially unchanged since then.
Figure III.1: FSIS Budget 1981
through 1993
(See figure in printed
edition.)
Nonfederal funds include reimbursements from meat and poultry plants
to FSIS for overtime, holiday, and voluntary inspections.
The fiscal year 1993 budget includes a supplemental appropriation of
$4 million, which was granted to FSIS for the last 6 months of fiscal
year 1993 to hire an additional 160 inspectors.
Source: 1983 through 1994 USDA Budgets. The 1994 budget contains
data on 1993 appropriations.
Figure III.2: FSIS 1981
through 1993 Budget in Constant
1993 Dollars
(See figure in printed
edition.)
Source: 1983 through 1994 USDA Budgets. The 1994 budget contains
data on 1993 appropriations.
MOST OF FSIS' BUDGET ALLOCATED
TO STAFFING COSTS
------------------------------------------------------- Appendix III:2
For fiscal year 1993, FSIS officials estimated that staffing costs
account for 82 percent of the agency's total budget and travel costs
account for 4 percent. (See fig. III.3.) This allocation shows that
FSIS is a labor-intensive agency. FSIS budget increases primarily
cover annual inflationary and salary increases. FSIS officials told
us that because the agency is so labor-intensive, it has few
discretionary funds. For example, the agency's travel budget is
primarily for inspectors' local travel between plants to perform
their inspection tasks. FSIS estimated that one-half of the
remaining 14 percent of the agency's federal funds would be allocated
to items such as supplies, equipment, printing, communications, and
utilities. FSIS is required to direct the remaining 7 percent of the
agency's total budget to required cost-sharing--50 percent--for those
states that have their own inspection programs. FSIS officials said
that these funds are not available for use in FSIS' federal programs.
FSIS provides oversight and annual reviews to state inspection
programs to ensure that these programs maintain equality with the
federal inspection program.
Figure III.3: Estimated
Percent of FSIS Budget
Allocated to Salaries and Other
Expenses--Fiscal Year 1993
(See figure in printed
edition.)
Note: Other Expenses include items such as supplies, equipment,
printing, communications, and utilities.
Source: 1994 USDA Budget, containing data on 1993 appropriations.
FSIS TOTAL AGENCY AND INSPECTOR
STAFF YEARS REMAIN CONSTANT
------------------------------------------------------- Appendix III:3
As reflected in FSIS' budget, total agency staff years have remained
relatively constant, fluctuating slightly between 10,705 staff years
in fiscal year 1981 to an estimated 10,754 staff years in 1993.
Staff years for in-plant slaughter and processing inspectors have
also fluctuated only slightly during this period from 8,857 inspector
staff years in 1981 to 8,734 staff years in 1993. Between 1981 and
1993, minor increases in slaughter inspection staff years have
accompanied minor decreases in processing inspection staff years.
That is, according to agency officials, as the poultry industry grew
and FSIS had to examine growing numbers of slaughtered poultry with
limited resources, the agency shifted processing inspectors to
slaughter inspection activities. Despite the shift in resources, the
proportion of slaughter inspectors to processing inspectors has
remained at approximately a 3:1 ratio, respectively, throughout the
13-year span. In addition, according to FSIS officials, fluctuations
in FSIS headquarters and field management staff have occurred in
conjunction with implementation of food safety initiatives, including
a processing management information system and residue monitoring
programs. (See fig. III.4.)
Figure III.4: FSIS Total
Agency Staff Years 1981 through
1993
(See figure in printed
edition.)
Notes: Inspection Force Management Staff includes 5 regional
offices; 26 area offices; and 185 circuit supervisors, where 15 to 35
plants comprise a "circuit," several circuits comprise an "area," and
several areas comprise a "region."
Other FSIS program and staff offices conduct management and program
support activities, including personnel and budget functions,
laboratory analysis, compliance reviews, recalls, and investigations
of contaminated products.
Source: GAO calculations based on data from FSIS' Office of
Inspection Operations and 1983 through 1993 USDA budgets.
ALMOST ONE-HALF OF FSIS STAFF
YEARS ALLOCATED TO
CARCASS-BY-CARCASS INSPECTION
------------------------------------------------------- Appendix III:4
Figure III.5 shows how FSIS allocated its staff years in fiscal year
1993. We estimate, based on discussions with agency officials, that
FSIS' Inspection Operations Program Office, which houses the agency's
domestic inspection resources, consumes 88 percent of the total
agency staff years. (See shaded area in fig. 3.5.) Inspection
resources include slaughter inspectors conducting carcass-by-carcass
inspection, slaughter inspectors conducting other inspection duties,
processing inspectors, and the inspection force management staff.
Resources dedicated to carcass-by-carcass inspection account for 47
percent of all FSIS staff years. Carcass-by-carcass inspection
requirements dictate that an inspector must always be present at each
inspection point on the slaughter line. Therefore, when slaughter
inspectors assigned to carcass-by-carcass inspection take breaks,
annual or sick leave, or training, FSIS must fill the vacancies that
result by using relief inspectors, transferring processing
inspectors, or reassigning inspectors conducting other inspection
duties in the slaughter plant. Although FSIS has more flexibility in
filling vacant processing inspection assignments, it still requires
daily inspection of each processing plant. That is, FSIS simply
assigns existing processing inspectors additional daily plant visits,
in effect, reducing the average amount of time an inspector spends in
a processing plant.
We calculated that FSIS dedicates 12 percent of its total staff years
to other program and staff offices. These offices are responsible
for conducting much of FSIS' management and program support
activities, including personnel and budget functions, laboratory
analysis, compliance reviews, recalls, and investigations of
contaminated products.
Figure III.5: 1993 Estimated
FSIS Total Agency Staff Year
Allocation, Including
Carcass-by-Carcass Inspection
(See figure in printed
edition.)
Notes: One staff year is the equivalent of one full-time employee.
Figure III.5 represents permanent full-time, part-time, and temporary
staff, as well as any overtime worked.
Carcass-by-carcass inspection is strictly slaughter postmortem
inspection tasks.
Other Inspection Duties include ante mortem, supervision, finished
product standards, sample collection, residue testing, sanitation,
monitoring of condemned products, and other activities.
Source: FSIS Office of Personnel and Office of Inspection
Operations.
POULTRY INDUSTRY GROWTH
INCREASES DEMAND FOR FSIS
INSPECTION RESOURCES UNDER
CURRENT REQUIREMENTS
------------------------------------------------------- Appendix III:5
Although the meat industry has experienced relatively little growth
between 1981 and 1992, the poultry industry has grown 54 percent
during this period from 4.4 billion slaughtered and federally
inspected birds in 1981 to 6.8 billion in 1992. (See fig. III.6.)
Because each carcass must be examined, this significant increase in
the number of poultry slaughtered taxes FSIS' inspection workforce.
For example, under the fastest line speeds currently employed, an
inspector has about 2 seconds to visually examine the inside and
outside surfaces of each bird and feel the eviscerated internal
organs. This equates to 1,800 birds per hour. We calculate that
approximately 2,100 full-time inspectors, at a minimum, were needed
to examine the 6.8 billion birds slaughtered in fiscal year
1992--about 1,830 regularly assigned inspectors plus another 260
relief inspectors. (FSIS' Resource Management Staff uses a ratio of
one relief inspector for every seven regularly assigned inspectors.)
Economists at USDA expect the poultry industry to continue to grow an
average of 4 percent per year for the next 5 to 10 years. To
maintain inspection time of 2 seconds per bird and handle such
growth, FSIS would need to hire an additional 460 inspectors by
fiscal year 1997 to inspect the projected 8.3 billion birds
slaughtered that year. (About 305 inspectors are needed to examine
each 1 billion birds slaughtered. See fig. III.7.) FSIS officials
were doubtful that lowering inspection time below 2 seconds per bird
would be an effective measure for handling the poultry growth.
Figure III.6: Actual Poultry
Federally Inspected 1981
through 1992 and Projections to
1997
(See figure in printed
edition.)
_____ Actual figures.
���� Projected growth based on information provided by the Economic
Research Service, USDA, and the National Broiler Council.
Source: 1981 through 1992 FSIS Annual Reports and USDA and National
Broiler Council officials.
Figure III.7: Minimum Staff
Year Requirements for
Carcass-by-Carcass Poultry
Inspection at the Fastest Line
Speeds
(See figure in printed
edition.)
Source: GAO calculations based on FSIS inspection requirements.
DAILY INSPECTION OF NUMEROUS
PLANTS LIMITS FSIS' ABILITY TO
ALLOCATE RESOURCES
------------------------------------------------------- Appendix III:6
Daily inspection of processing plants limits FSIS' ability to adjust
inspection frequencies to changing public health risks. FSIS'
inspection policy requires FSIS to inspect all processing plants
daily, or once per shift if a plant operates more than one 8-hour
shift. Therefore, all plants, regardless of public health risk,
history of compliance with regulatory requirements, or type of
processing controls, receive the same frequency and intensity of
inspection. FSIS officials told us that 5,906 meat and poultry
plants were conducting processing activities as of March 1993. Of
these plants, 56 percent, or 3,321, are small or "low" volume
operations that produce an estimated 1 million pounds or less of
product annually. (See fig. III.8.)
Figure III.8: Size of Meat and
Poultry Processing Plants
(See figure in printed
edition.)
Notes: Low volume: Processes 1 million pounds per year or less.
Medium volume: Processes between 1 and 10 million pounds per year.
High volume: Processes more than 10 million pounds per year.
Source: PBIS National Database, FSIS, (Jan.-Mar. 1993).
Because all processing plants must be inspected at least daily and
small plants do not provide a full days work, many processing
inspectors, who operate under "patrol assignments," must spend time
traveling between several plants daily. Most of FSIS' 1,400 patrol
assignments (both first and second shifts) comprise three to six
plants and thus require considerable travel time and transportation
cost. (See fig. III.9.)
Processing inspectors conduct various inspection tasks daily. On
average, an estimated 11 percent of processing inspectors' time is
spent traveling. (See fig. III.10.) This equals about 240 staff
years and costs FSIS about $8.1 million annually based on an average
yearly salary of $33,800 for a processing inspector. In addition,
about $11 million is spent annually to reimburse inspectors for use
of their cars and other travel expenses associated with patrol
assignments.
FSIS believes that in order for an inspector to make a judgment that
a processing plant's overall sanitation and operations are adequate
to ensure a wholesome product, the inspector must maintain a certain
presence in the plant over a period of time. FSIS has determined
that a minimal presence requires an inspector to visit a plant not
only once a day, but during every shift and additionally during
overtime hours. FSIS' rationale includes the argument that because
employees and product mix are likely to change between shifts, it
would be inappropriate to assume that the overall sanitation and
operations measured during one shift represented a plant's
performance during another shift. On average, about 1,300 of the
5,900 processing plants require a second FSIS inspection each day
because the plants run a second shift. In addition, on average, at
least 258 plants receive a separate inspection because they conduct
overtime operations. (See fig. III.11.)
Figure III.9: Number of
Processing Patrol Assignments
by Number of Plants Visited
(See figure in printed
edition.)
Source: FSIS, Office of Inspection Operations.
Figure III.10: Percent of Time
Spent on Processing Inspection
Tasks Between April 1992 and
March 1993
(See figure in printed
edition.)
Note: Administrative Tasks include paperwork, follow-up on problem
areas, and personal time.
Source: Percentages estimated by GAO based on data provided by the
Processed Products Inspection Division, FSIS.
Figure III.11: Frequency of
Daily Inspection Visits to Meat
and Poultry Processing Plants
(See figure in printed
edition.)
\a Although FSIS could not determine actual number of visits
conducted during overtime hours, we estimated that, on average, a
minimum of 258 visits would occur each day.
Source: FSIS, Office of Inspection Operations.
SOME PLANTS INDEPENDENTLY TEST FOR
MICROBIAL CONTAMINANTS
========================================================== Appendix IV
While FSIS has no microtesting requirements, many of the plants that
we contacted conducted tests for microbial contaminants. This
appendix provides detailed information on the microtesting conducted
by the plants we contacted.
NUMBER OF PLANTS THAT CONDUCT
MICROTESTING
-------------------------------------------------------- Appendix IV:1
Almost one-half of the plants that we contacted conducted
microtesting. Processing plants were more likely to conduct
microtesting than slaughter plants. For example, 56 percent, or 46,
of the 82 processing plants conducted microtesting, compared with 21
percent, or 3, of the 14 slaughter plants. Of the 27 combination
plants--plants that performed both slaughter and processing
operations--24 conducted microtesting in both their slaughter and
processing facilities, the other 3 combination plants conducted tests
only in their processing facilities.
Table IV.1
Plants With Microtesting Programs
Number of Number of Percent of
plants plants plants
Plants contacted microtesting microtesting
---------------- ---------- -------------- --------------
Slaughter 14 3 21
Processing 82 46 56
Combination 61 27 44
============================================================
Total 157 76 48
------------------------------------------------------------
Source: GAO meat and poultry plant survey.
PLANT TESTING PRIMARILY AIMED
AT IMPROVING PRODUCT SAFETY
-------------------------------------------------------- Appendix IV:2
Product safety was the most common reason plants' cited for
microtesting. Product quality was a close second. Figure IV.1 shows
plants' responses to the question, "Why do you microtest?" Plant
officials were not restricted in their responses and often cited more
than one reason.
Figure IV.1: Plants' Reasons
for Microtesting
(See figure in printed
edition.)
Note: Limit liability means that some plants used microtesting as
one way to verify the safety and quality of their products should
there be a legal dispute with a customer or consumer.
Source: GAO meat and poultry plant survey.
MICROTESTING LEADS TO
CONSTRUCTIVE CHANGES
-------------------------------------------------------- Appendix IV:3
Based on high bacterial counts or the presence of pathogens
identified through microtesting, 74 of the 76 plants made
constructive changes to improve their product safety and quality.
These changes can be placed into four general categories:
equipment/facility design; sanitation practices; employee awareness;
and supplier/product management. The following are specific examples
of plant changes made to reduce high microbial counts.
EQUIPMENT/FACILITY DESIGN
------------------------------------------------------ Appendix IV:3.1
Thirty of the plants discovered through their microtesting programs
that certain pieces of equipment or the facility itself tended to
harbor bacteria. Examples of actual changes made in the design of
the equipment or facility include:
Cutting holes in machines to provide for better access for
cleaning.
Cutting grooves in floor drains to facilitate flow and reduce drain
water backup.
Removing scrapers designed to dislodge excess fat from the conveyor
belt because they spread bacteria across the entire conveyor
belt.
SANITATION PRACTICES
------------------------------------------------------ Appendix IV:3.2
Sixty-four of the plants identified through microtesting more
effective sanitation practices. Examples of improved sanitation
practices include:
Changing cleaning solutions.
Replacing rather than trying to clean conveyor belts.
Removing conveyor belts for cleaning rather than cleaning them in
place.
Soaking carcass shrouds in salt and iodine solutions to reduce
bacterial growth.
EMPLOYEE AWARENESS
------------------------------------------------------ Appendix IV:3.3
Forty plants said that microtesting enabled them to increase employee
awareness of and appreciation for good hygiene practices and proper
sanitation procedures. Examples of actions taken to improve employee
awareness include:
Focusing employee training on problem areas identified through
microtesting.
Posting additional sanitation signs near identified problem areas.
Providing disposable sleeves and requiring employees to use them.
Having the plant microbiologist spend time looking for violations
of plant sanitation and employee hygiene practices.
SUPPLIER/PRODUCT MANAGEMENT
------------------------------------------------------ Appendix IV:3.4
Forty-four plants used microtesting to improve product safety and
quality by better managing their suppliers or production processes.
Examples of actions taken to manage suppliers and production
processes include:
Advising suppliers of high bacterial counts and requesting that the
counts be lowered.
Assisting suppliers identify and eliminate sources of high bacteria
counts.
Dropping suppliers that continually provided products with
excessive bacterial counts.
Tying-off the intestinal tract and clipping the hide prior to their
removal.
Washing, after each use, the clips used to hang carcass.
Recalling or placing potentially unsafe products on hold.
Recooking products that exceeded plant microbial standards.
Dropping product lines that consistently exceeded plant microbial
standards.
LARGER PLANTS MORE LIKELY TO
CONDUCT MICROTESTING
-------------------------------------------------------- Appendix IV:4
The more a plant produced, the more likely it was to conduct
microtesting. Figure IV.2 shows the number of low-, medium-, and
high-volume plants that conducted microtesting.
Figure IV.2: Larger Plants Are
More Likely to Conduct
Microbial Tests
(See figure in printed
edition.)
Notes: Low volume: Processes 1 million pounds per year or less.
Medium volume: Processes between 1 and 10 million pounds per year.
High volume: Processes more than 10 million pounds per year.
Source: GAO meat and poultry plant survey.
PLANT PROGRAMS VARY
-------------------------------------------------------- Appendix IV:5
Plants' microtesting programs varied according to when and where they
tested, how often they tested, and the types of tests they conducted.
TESTING FREQUENCIES VARIED
------------------------------------------------------ Appendix IV:5.1
Plants varied in how often they tested. Some plant programs called
for a few equipment or product tests supplemented by additional tests
when a bacterial or pathogen problem was suspected, such as when they
found a drop in shelf life. Other plants, however, tested much more
frequently because microtesting was a key part of their quality
control program. Table IV:2 shows the number of tests plants
performed throughout the week.
Table IV.2
Frequency of Testing
Number of plants Number of plants
Tests per week testing equipment testing product
-------------------- ------------------ ------------------
1-10 6 7
11-100 40 23
101-500 10 6
Over 500 2 2
Do not test 1 8
Not specified 17 30
============================================================
Total 76 76
------------------------------------------------------------
Source: GAO meat and poultry plant survey.
TYPES OF TESTS CONDUCTED
VARIED
------------------------------------------------------ Appendix IV:5.2
Another variation in program design was what was tested and the types
of tests conducted. There are different types of microbial tests,
each designed to indicate the presence of microbial organisms. Many
plants chose to test for the total bacteria level, which does not
identify specific bacteria but shows the cleanliness of equipment
surfaces or the general product condition. Other plants tested for
specific types of bacteria, particularly bacteria that cause
foodborne illnesses. Figure IV.3 shows the number of plants that
tested equipment surfaces before and during operations and raw and
ready-to-eat product samples. Figures IV.4 through IV.7 show by test
type, the number of plants testing equipment surfaces before and
during operations and the number of plants testing raw and
ready-to-eat products. The data shown in figures IV.4 through IV.7
may differ from the data shown on figure IV.3 because some plants
test for general bacterial levels only, some plants test for specific
pathogens and not for general bacteria levels, and others test for
both general bacteria levels and specific pathogens.
Figure IV.3: Number of Plants
Testing Equipment Surfaces and
Products
(See figure in printed
edition.)
Source: GAO meat and poultry plant survey.
Figure IV.4: Types of Tests
Conducted on Equipment Prior to
Operations
(See figure in printed
edition.)
Note: Examples of pathogen tests include E. coli, salmonella,
listeria, and staphylococcus. Examples of other tests include yeast
and mold.
Source: GAO meat and poultry plant survey.
Figure IV.5: Types of Tests
Conducted on Equipment During
Operations
(See figure in printed
edition.)
Note: Examples of pathogen tests include E. coli, salmonella,
listeria, and staphylococcus. Examples of other tests include yeast
and mold.
Source: GAO meat and poultry plant survey.
Figure: IV.6: Types of Tests
Conducted on Raw Products
(See figure in printed
edition.)
Note: Examples of pathogen tests include E. coli, salmonella,
listeria, and staphylococcus. Examples of other tests include yeast
and mold.
Source: GAO meat and poultry plant survey.
Figure IV.7: Types of Tests
Conducted on Ready-to-Eat
Products
(See figure in printed
edition.)
Note: Examples of pathogen tests include E. coli, salmonella,
listeria, and staphylococcus. Examples of other tests include yeast
and mold.
Source: GAO meat and poultry plant survey.
EVALUATION STANDARDS VARY
------------------------------------------------------ Appendix IV:5.3
Plants use different standards to evaluate their test results. What
was an acceptable level of bacteria for one plant, was unacceptable
at another plant. To illustrate this point, figures IV.8 through
IV.11 show the variation in plants' test evaluation standards.
Figure IV.8: Standard Plate
Count Criteria Applied to
Equipment Tests Made Prior to
Operations
(See figure in printed
edition.)
Notes: Strict Standard: 0-500 colonies per square inch.
Moderate Standard: 501-100,000 colonies per square inch.
Minimal/No Standard: Over 100,000 colonies per square inch.
Not testing: Plants not testing for bacteria during operations.
Source: GAO meat and poultry plant survey.
Figure IV.9: Standard Plate
Count Criteria Applied to
Equipment Tests Made During
Operations
(See figure in printed
edition.)
Notes: Strict Standard: 0-500 colonies per square inch.
Moderate Standard: 501-100,000 colonies per square inch.
Minimal/No Standard: Over 100,000 colonies per square inch.
Not testing: Plants not testing for bacteria during operations.
Source: GAO meat and poultry plant survey.
Figure IV.10: Standard Plate
Count Criteria Applied to Tests
of Raw Products
(See figure in printed
edition.)
Notes: Strict Standard: 0-500 colonies per square inch.
Moderate Standard: 501-100,000 colonies per square inch.
Minimal/No Standard: Over 100,000 colonies per square inch.
Not testing: Plants not testing for bacteria during operations.
Source: GAO meat and poultry plant survey.
Figure IV.11: Standard Plate
Count Criteria Applied to Tests
of Ready-to-Eat Products
(See figure in printed
edition.)
Notes: Strict Standard: 0-500 colonies per square inch.
Moderate Standard: 501-100,000 colonies per square inch.
Minimal/No Standard: Over 100,000 colonies per square inch.
Not testing: Plants not testing for bacteria during operations.
Source: GAO meat and poultry plant survey.
PROGRAM COSTS VARY
-------------------------------------------------------- Appendix IV:6
Plants' microtesting programs varied in their annual costs. Figure
IV.12 shows a breakout of plants according to how much they spent
each year.
Figure IV.12: Approximate
Annual Costs of Plants'
Microtesting Programs
(See figure in printed
edition.)
Note: Not reported: Program costs were either not readily available
or considered proprietary.
Source: GAO meat and poultry plant survey.
PLANTS SEEK OUTSIDE ASSISTANCE
-------------------------------------------------------- Appendix IV:7
FSIS provided plants little assistance in developing and/or operating
microtesting programs so plants sought such assistance elsewhere or
relied on in-house staff expertise. Figure IV.13 shows the
percentage of plants that sought outside assistance, relied on
in-house expertise, or used both outside assistance and in-house
expertise.
Figure IV.13: Sources Used for
the Design and Operation of
Plant Microtesting Programs
(See figure in printed
edition.)
Source: GAO meat and poultry plant survey.
NEED FOR FSIS ASSISTANCE
-------------------------------------------------------- Appendix IV:8
Almost two-thirds of the plants with microtesting programs believed
that FSIS should provide more assistance. Figure IV.14 shows plants'
views on whether FSIS should provide more guidance to plants
operating microtesting programs and assistance to plants wanting to
establish microtesting programs.
Figure IV.14: Should FSIS
Provide Plants Guidelines and
Assistance on Microtesting?
(See figure in printed
edition.)
Note: These percentages represent the views of the 76 plants GAO
contacted that had microtesting programs.
Source: GAO meat and poultry plant survey.
FSIS NOT AWARE OF PLANTS'
MICROBIAL TESTING RESULTS
-------------------------------------------------------- Appendix IV:9
FSIS does not routinely monitor plant microtesting programs and thus
is not in the position to disseminate information on problems
identified and/or improvements made because of microbial testing.
Figure IV.15 shows that FSIS routinely monitored about one-fourth of
the plants microtesting procedures or test results.
Figure IV.15: Plant
Microtesting Programs Monitored
by FSIS
(See figure in printed
edition.)
Source: GAO analysis of FSIS records.
(See figure in printed edition.)Appendix V
COMMENTS FROM THE U.S. DEPARTMENT
OF AGRICULTURE
========================================================== Appendix IV
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
OBJECTIVES, SCOPE, AND METHODOLOGY
========================================================== Appendix VI
Concerned about the effectiveness of FSIS' meat and poultry
inspection system, the Chairman, Subcommittee on Department
Operations and Nutrition, and the Ranking Minority Member,
Subcommittee on Livestock, House Committee on Agriculture, asked us
to evaluate whether (1) the system makes the most effective use of
its resources to ensure food safety, (2) meat and poultry plants have
programs to test for microorganisms, and (3) a quality control
concept known as Hazard Analysis and Critical Control Point (HACCP)
is an effective approach for ensuring food safety.
To assess whether FSIS makes the most effective use of its resources
to ensure that the food supply is safe, we interviewed FSIS officials
and obtained data on FSIS resource allocation and slaughter volume
between fiscal years 1981, the year FSIS was created, and 1993. We
(1) reviewed scientific publications, including reports by the
National Academy of Sciences and Centers for Disease Control (CDC),
which explained the principal public health risks associated with
meat and poultry products and (2) evaluated how FSIS's resources are
allocated to various inspection tasks intended to protect the public
from these risks. In addition, we interviewed CDC officials to
determine the extent and causes of foodborne illnesses associated
with meat and poultry products.
We also reviewed legislation pertaining to meat and poultry
inspection, USDA and FSIS budgets, and FSIS documents to determine
federal inspection requirements in domestic meat and poultry plants
and how FSIS has allocated personnel and funds to fulfill these
requirements. We obtained USDA and private sector officials' views
on the future growth of the meat and poultry industry and how such
growth would impact FSIS' ability to meet its mandated inspection
requirements. Where complete data on use of resources were
unavailable, we computed estimates by using methodologies suggested
by FSIS officials.
To determine the extent of industry microbial testing programs, we
interviewed officials at FSIS headquarters in Washington, D.C., and
FSIS North Central, Northeastern, Southwestern, and Western Regional
Offices in Des Moines, Iowa; Philadelphia, Pennsylvania; Dallas,
Texas; and Alameda, California, respectively. We also reviewed FSIS
inspection plans and reports. In addition, we contacted 157
judgmentally selected meat and poultry plants. The plants were
selected to both ensure geographic coverage and coverage of plant
operations--that is, slaughter and processing operations. FSIS
officials assisted us in identifying the FSIS regions and circuits (a
group of plants under an individual FSIS supervisor) from which we
selected the plants to contact. From each circuit we selected all
the slaughter and combination slaughter/processing plants and
randomly selected from the remainder of the circuit the number of
large and small processing plants that could be contacted in the time
allowed for field visits. However, because the plants contacted were
judgmentally selected, the results obtained cannot be used to make
statements about the total universe of meat and poultry plants. The
information obtained about plant microtesting programs only applies
to the 157 plants we contacted in California, Kansas, Maryland,
Pennsylvania, Virginia, and Wisconsin.
We obtained explanatory documentation of plant programs, test
results, and changes made, when possible. Although some plants
viewed this information as proprietary and thus were not willing to
provide copies of such information, they generally showed us program
documents that supported the general scope and results of their
microtesting programs.
To determine whether HACCP is an effective quality control system for
ensuring food safety, we interviewed USDA, Food and Drug
Administration, and industry officials and reviewed documentation on
past and present uses of this system. We also interviewed various
industry, federal, and advocacy organization members to obtain their
views on FSIS' current inspection system and whether the need for a
new system existed.
We performed our review between April 1993 and January 1994, in
accordance with generally accepted government auditing standards.
MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII
RESOURCES, COMMUNITY, AND
ECONOMIC DEVELOPMENT DIVISION,
WASHINGTON, D.C.
------------------------------------------------------- Appendix VII:1
Edward M. Zadjura, Assistant Director
Louis J. Schuster, Assignment Manager
John M. Nicholson, Jr., Evaluator-in-Charge
Marci D. Kramer, Staff Evaluator
SAN FRANCISCO REGIONAL OFFICE
------------------------------------------------------- Appendix VII:2
Keith W. Oleson, Issue Area Manager
Nathan L. Brown, Staff Evaluator
Christine L. McIntyre, Staff Evaluator