Modernization of Poultry Slaughter Inspection
Federal RegisterAug 21, 2014
Ask Donna
What actually matters in this document.
Text
DEPARTMENT OF AGRICULTURE
Food Safety and Inspection Service
9 CFR Parts 381 and 500
[Docket No. FSIS-2011-0012]
RIN 0583-AD32
Modernization of Poultry Slaughter Inspection
AGENCY:
Food Safety and Inspection Service, USDA.
ACTION:
Final rule.
SUMMARY:
The Food Safety and Inspection Service (FSIS) is amending the poultry products inspection regulations to establish a new inspection system for young chicken and all turkey slaughter establishments. Young chicken and turkey slaughter establishments that do not choose to operate under the new poultry inspection system may continue to operate under their current inspection system. The Agency is also making several changes to the regulations that will affect all establishments that slaughter poultry other than ratites. This final rule is a result of the Agency's 2011 regulatory review efforts conducted under Executive Order 13563 on Improving Regulation and Regulatory Review.
DATES:
Effective Date:
October 20, 2014.
Notification Date:
All young chicken and turkey slaughter establishments will initially have until February 23, 2015, to notify their District Office in writing of their intent to operate under the New Poultry Inspection System (NPIS). Establishments that do not notify their District Office of their intent by February 23, 2015, will be deemed to have chosen the inspection system that they are currently operating under. Young chicken and turkey slaughter establishments that decide that they would like to convert to NPIS after the initial notification date may notify FSIS of their intent at any time after that date. The Agency will implement the NPIS in the additional establishments that intend to convert on a schedule consistent with Agency resources and readiness. The Agency intends to implement the NPIS in all young chicken and turkey establishments that choose to operate under the NPIS, regardless of when the establishment notifies FSIS of its intent to transition to the NPIS. However, the initial implementation wave will only include those establishments that submitted their notifications within the initial notification period.
After October 20, 2014, FSIS will begin selecting from those establishments that have notified FSIS of their intent to switch to the NPIS. The Agency will use a computerized ranking system to determine the schedule of establishments for implementation of the NPIS. This ranking system will take into consideration several factors, such as FSIS staffing needs, past performance of the establishment, the location of the establishment with respect to other federally-inspected establishments, and establishment readiness to transition to the NPIS. FSIS will implement the NPIS in phases by clusters of establishments in close geographic proximity to one another. The initial implementation wave will only include those establishments that notified FSIS of their intent to switch to the NPIS during the initial six-month notification period. FSIS expects that in subsequent years many more establishments will choose to transition to the new system. The Agency's implementation strategy for the NPIS is described in more detail in the preamble to this final rule.
Applicability Dates:
The regulations that prescribe procedures for controlling visible fecal contamination in 9 CFR 381.65(f), the regulations that prescribe procedures for controlling contamination throughout the slaughter and dressing process in 9 CFR 381.65(g), and the regulations that prescribe recordkeeping requirements in 9 CFR 381.65(h) will be applicable as follows:
• In large establishments, defined as all establishments with 500 or more employees, on November 19, 2014;
• In small establishments, defined as all establishments with 10 or more employees but fewer than 500, on December 19, 2014;
• In very small establishments, defined as all establishments with fewer than 10 employees or annual sales of less than $2.5 million February 17, 2015.
FOR FURTHER INFORMATION CONTACT:
Daniel Engeljohn, Assistant Administrator, Office of Policy and Program Development, FSIS, U.S. Department of Agriculture, 1400 Independence Avenue SW., Washington, DC 20250-3700, (202) 205-0495.
SUPPLEMENTARY INFORMATION:
Executive Summary
In January 2011, President Obama issued Executive Order (E.O.) 13563 on Improving Regulation and Regulatory Review. As part of this E.O., agencies were asked to review existing rules that may be outmoded, ineffective, insufficient, or excessively burdensome, and to modify, streamline, expand, or repeal them accordingly. As a result of FSIS's regulatory review efforts conducted under E.O. 13563, on January 27, 2012, the Agency published a proposed rule to modernize poultry slaughter inspection (“
Modernization of Poultry Slaughter Inspection,
” 77 FR 13512). This final rule adopts, with modifications, the provisions in the January 2012 proposal. FSIS is issuing this rule to facilitate pathogen reduction in poultry products, improve the effectiveness of poultry slaughter inspection, make better use of the Agency's resources, and remove unnecessary regulatory obstacles to innovation.
This final rule will establish a New Poultry Inspection System (NPIS) for young chicken and all turkey slaughter establishments. The NPIS will not replace, as was proposed, the current Streamlined Inspection System (SIS), the New Line Speed Inspection System (NELS), or the New Turkey Inspection System (NTIS). As such, young chicken and turkey slaughter establishments may choose to operate under the NPIS or may continue to operate under their current inspection system, i.e., SIS, NELS, NTIS, or Traditional Inspection, as modified by this final rule. Establishments that slaughter poultry other than young chickens or turkeys are not eligible to operate under the NPIS unless they obtain a waiver under the Salmonella Initiative Program. The Agency is not limiting the number of online inspectors in Traditional Inspection to two, as was proposed. FSIS will continue to staff all establishments that do not choose to operate under the NPIS with their current number of online inspectors.
The NPIS is designed to facilitate pathogen reduction in poultry products by shifting Agency resources to allow FSIS inspectors to perform more offline inspection activities that are more effective in ensuring food safety, while providing for a more efficient and effective online carcass-by-carcass inspection. Data from the Agency's Hazard Analysis and Critical Control Point Systems (HACCP)-Based Inspection Models Project (HIMP) pilot study,
1
which was used to inform the NPIS, show that an inspection system that provides for increased offline inspection activities that are more directly related to food safety results in greater compliance with sanitation and HACCP regulations, carcasses with
lower levels of visible fecal contamination, and carcasses with equivalent or lower levels of
Salmonella
contamination.
1
See “Evaluation of HACCP Inspection Models Project (HIMP), August 2011 (available on the FSIS Web site at:
http://www.fsis.usda.gov/wps/wcm/connect/fcd9ca3e-3f08-421f-84a7-936bc410627c/Evaluation_HACCP_HIMP.pdf?MOD=AJPERES
).
Key elements of the NPIS include: (1) Requiring that establishment personnel sort carcasses and remove unacceptable carcasses and parts before the birds are presented to the FSIS carcass inspector; (2) shifting Agency resources to conduct more offline inspection activities that are more effective in ensuring food safety, which will allow for one offline verification inspector per line per shift and will reduce the number of online inspectors to one; (3) replacing the Finished Product Standards (FPS), which will apply to establishments that continue operating under SIS, NELS, and NTIS, with a requirement that establishments that operate under the NPIS maintain records to document that the products resulting from their slaughter operations meet the definition of ready-to-cook (RTC) poultry; and (4) authorizing young chicken slaughter establishments to operate at a maximum line speed of 140 birds per minute (bpm), provided that they maintain process control.
Under all of the current inspection systems, online inspectors visually inspect every carcass, with its corresponding viscera, at fixed locations on the evisceration line immediately after separation of the viscera from the interior of the carcasses. The online inspectors are responsible for identifying unacceptable carcasses and parts, examining carcasses for visual defects, and directing establishment employees to take appropriate corrective actions if the defects can be corrected through trimming or reprocessing. The maximum line speeds authorized under the existing inspection systems reflect the time it takes for an inspector to effectively perform the online carcass inspection procedures required under these systems.
Under the NPIS, there will be one online carcass inspector (CI) and one offline verification inspector (VI) assigned to each evisceration line. As under the HIMP inspection system, VIs and CIs under the NPIS will have different but complementary roles in ensuring that poultry products leaving the slaughter line are safe and wholesome. Under the NPIS, CIs will conduct a continuous online inspection of each carcass at a fixed location immediately before the chiller to determine whether each carcass is not adulterated. CIs under the NPIS will be able to conduct a more efficient and effective online carcass inspection than online inspectors do under the current inspection systems because the CIs are presented with carcasses that have been sorted, washed, and trimmed by establishment employees, and are thus much more likely to pass inspection.
The VIs under the NPIS will conduct offline food safety-related inspection activities and will monitor and evaluate establishment process controls. The VIs will conduct carcass verification checks on carcass samples collected before the CI station to ensure that the establishment is effectively sorting carcasses and that it is producing products that comply with the Agency's zero visible fecal tolerance and other performance standards. The VI and CI will work with the inspector-in-charge (IIC) to ensure that the carcasses presented to the CI are not affected with food safety defects or other conditions at levels that may impair the CI's ability to effectively inspect each carcass. VIs will also perform offline activities in addition to carcass verification checks, such as verifying compliance with sanitation standard operating procedures (SOPs), sanitation performance standards (SPS), and HACCP regulatory requirements, and ensuring that the establishment is meeting all regulatory requirements and is effectively preventing contamination by enteric pathogens and fecal material throughout the entire slaughter and dressing process.
The fastest maximum line speed authorized under the current inspection systems is 140 bpm under the SIS for young chickens. To determine line speeds for SIS, FSIS conducted field and work measurement studies of online inspectors to determine the time needed for an inspector to perform the SIS inspection procedure. The studies showed that online inspectors can perform the SIS inspection procedure at line speeds of up to 140 bpm if each inspector is presented with up to 35 bpm. Thus, under SIS, establishments with automated evisceration equipment may operate at 140 bpm with four FSIS online inspectors assigned to the line. The maximum line speeds authorized under the other inspection systems are 91 bpm with three online inspectors for NELS, and 51 bpm for light turkeys with two online inspectors and 45 bpm for heavy turkeys with two online inspectors for NTIS. As noted in the proposed rule, Traditional Inspection is typically employed at smaller lower production volume establishments that eviscerate carcasses by hand (77 FR 4410). Thus, the maximum line speeds authorized under Traditional Inspection are slower than those under SIS, NELS, and NTIS. The maximum line speed for young chickens under Traditional Inspection is 64 bpm with four online inspectors. The maximum line speed for turkeys under Traditional Inspection is 39 bpm with three online inspectors.
As discussed in more detail later in this document, since 2007, HIMP young chicken establishments have been authorized to operate at line speeds of up to 175 bpm, depending on their ability to demonstrate consistent process control. Experience from the HIMP pilot shows that HIMP establishments operate with an average line speed of 131 bpm, and, although they are authorized to do so, most of the young chicken HIMP establishments do not operate line speeds at 175 bpm. Establishments determine their line speeds based on their equipment and facilities, bird size and flock conditions, and their ability to maintain process control when operating at a given line speed. In addition, line speeds under HIMP depend on the number of employees that the establishments hire and train to perform sorting activities. Although the maximum line speed under the NPIS is 140 bpm and not 175 bpm as authorized under HIMP, FSIS believes that establishments choosing to operate under the NPIS will determine their line speeds based on the same factors that establishments considered when setting line speeds under HIMP for the past 15 years.
Regardless of line speed, because HIMP and NPIS do not require that establishments configure their evisceration lines to accommodate more than one online carcass inspector, establishments operating under the NPIS will have greater control over their lines and greater flexibility over their production process. For example, as under HIMP, establishments operating under the NPIS will have the flexibility to reconfigure and consolidate lines if they determine that they need more space to conduct other activities in their facilities. In addition, because only one online inspector is required at the end of the line, establishments operating under the NPIS will not need to adjust their production based on the availability of FSIS inspection personnel to be stationed online. Establishment employees will staff the lines to perform the online sorting activities. Establishments that operate under NPIS will also have greater flexibility to increase production to respond to customer demands.
As under HIMP, in addition to having more control over their production process, establishments operating under the NPIS will also have more opportunities for innovation and greater flexibility to develop and implement certain types of new technologies. Currently, if an establishment operating
under the existing inspection systems wants to use new technologies for evisceration or for sorting, the establishment must work directly with the Agency to accommodate FSIS`s online slaughter inspection methodologies. Doing so takes time and can become an obstacle to innovation. Under the NPIS, establishments will have direct control of the sorting process within their facilities and therefore will have the flexibility to implement and assess the technologies they think are beneficial to their operations.
In addition to the NPIS for young chickens and turkeys, this final rule includes changes to the regulations that will apply to all establishments that slaughter poultry other than ratites. Under this final rule, all poultry slaughter establishments must develop, implement, and maintain written procedures to ensure that carcasses contaminated with visible fecal material do not enter the chiller, and they must incorporate these procedures into their HACCP plans, or sanitation SOPs, or other prerequisite programs (also referred to collectively as “the HACCP system” in this document). This final rule also requires that all poultry slaughter establishments develop, implement, and maintain written procedures to prevent contamination of carcasses and parts by enteric pathogens and fecal material throughout the entire slaughter and dressing operation, and that they incorporate their procedures into their HACCP systems. At a minimum, these procedures must include sampling and analysis for microbial organisms at the pre- and post-chill points in the process to monitor process control for enteric pathogens, with some exceptions for very small and very low volume establishments. Establishments will be required to maintain daily records sufficient to document the implementation and monitoring of these procedures. These new requirements will ensure that all poultry slaughter establishments implement appropriate measures to prevent contamination of carcasses by enteric pathogens and visible fecal material and that both FSIS and establishments have the documentation they need to verify the effectiveness of these measures on an ongoing basis.
FSIS is also rescinding the regulation that requires that poultry establishments test carcasses for generic
E. coli
to monitor for process control. The generic E. coli regulations will be replaced by the new testing requirements described above. The new testing requirements will allow establishments to develop sampling plans that are more tailored, thus more effective in monitoring their specific process control than the current generic E. coli criteria. The Agency has concluded that the use of generic
E. coli
as an indicator for process control may not be as useful in broiler operations as originally thought. The Agency is taking this action to allow establishments to use other more relevant indicators of process control. The Agency established new performance standards for Salmonella and Campylobacter in 2011 to more effectively manage these pathogens (76 FR 15282). Therefore, FSIS is removing the codified
Salmonella
pathogen reduction performance standards for poultry.
Finally, FSIS is removing the prescriptive time and temperature parameters from the chilling requirements for RTC poultry and instead is requiring that poultry establishments incorporate procedures for chilling poultry into their HACCP systems. The Agency is also amending the regulations to permit poultry slaughter establishments to use (1) approved online reprocessing antimicrobial systems or (2) offline reprocessing antimicrobial agents including chlorinated water containing 20 ppm to 50 ppm available chlorine or other antimicrobial substances that have been approved as safe and suitable for reprocessing poultry. Establishments will be required to address the use of online or offline reprocessing in their HACCP systems.
Table 1—Estimated Net Social Benefits From the Rule (Millions of Dollars), Annualized Over 10 Years With a 7% Discount Rate, for Varying Percent Changes That Switch to NPIS
[Percentage of Industry that Switches to NPIS]
0%
10%
25%
50%
75%
90%
100%
NPIS:
Benefits:
Public health benefits (10%, 90%)
0.0
1.0 (0.3 to 1.7)
2.4 (0.8 to 4.3)
4.8 (1.6 to 8.7)
7.2 (2.4 to 13.0)
8.6 (2.9 to 15.7)
9.6 (3.3 to 17.4)
FSIS net savings
0.0
2.3
5.7
11.4
17.1
20.5
22.8
Unquantified benefits
Increased flexibility for establishments to design and implement production measures tailored to their operations, in some cases possibly including increased line speed up to 140 chickens or 55 turkeys per minute
Costs:
Costs to establishments
0.0
1.6
4.0
8.0
12.0
14.4
16.0
Unquantified costs
Industry cost of responding to new NPIS inspections in a manner that may lead to public health benefits (e.g., discarding contaminated food or cooking it longer)
Mandatory Component:
Costs to establishments
9.1
9.1
9.1
9.1
9.1
9.1
9.1
Unquantified benefits
Potential additional public health benefits from documentation and testing
Unquantified costs
Industry cost of responding to information generated by documentation and testing in a manner that may lead to public health benefits (e.g., discarding contaminated food or cooking it longer)
Total benefits (10%, 90%)
0.0
3.3 (2.6 to 4.0)
8.1 (6.5 to 10.0)
16.2 (13.0 to 20.1)
24.3 (19.5 to 30.1)
29.1 (23.4 to 36.2)
32.4 (26.0 to 40.2)
Total costs
9.1
10.7
13.1
17.1
21.1
23.5
25.1
Net benefits (10%, 90%)
−9.1
−7.4 (−8.1 to −6.7)
−5 (−6.6 to −3.1)
−0.9 (−4.1 to 3.0)
3.2 (−1.6 to 9.0)
5.6 (−0.1 to 12.7)
7.3 (0.9 to 15.1)
FSIS presents the costs and cost savings that would be generated over a range of assumptions with respect to how much of the industry will choose to adopt NPIS within five years. These estimates are scaled from an illustrative calculation that assumes that all 219 small and large non-Traditional establishments adopt NPIS, which, while used to calculate potential maximum effect, is not necessarily FSIS's assumption of the most likely outcome. Later portions of the regulatory impact analysis section contain discussion of the uncertainty surrounding the net benefits associated with how much of the industry will choose to adopt NPIS.
Table of Contents
I. Background
II. Summary of Modifications Made to the Proposed Rule
III. Comments and Responses
A. NACMPI and Public Process
B. The HIMP Report
1. Data and Methods Used in the HIMP Report
2. HIMP as the Basis for the NPIS
3. Carcass Inspection Under HIMP
4. Public Health-Related Non-Compliances
5. OCP Standards Under HIMP
6.
Salmonella
Positive Rates in HIMP Establishments
C. The Risk Assessment
D. The New Poultry Inspection System (NPIS)
1. General Comments on the NPIS
2. Scope of the NPIS
3. Carcass Sorting and Inspection Under the NPIS
a. Carcass Sorting by Establishment Employees
b. Online Carcass Inspection
c. Inspection for Avian Visceral Leukosis
d. Verification Inspection
e. RTC Poultry Definition Under the NPIS
4. Facilities Requirements and Staffing for NPIS
a. Facilities Requirements
b. Staffing
5. Line Speeds Under the NPIS
a. Line Speeds and Process Control
b. Line Speeds and Online Carcass Inspection
E. Implementation of the NPIS
1. Background
2. Implementation Strategy
3. Comments on Proposed Implementation Plan
F. Line Speeds and Worker Safety
1. Collaboration With the National Institute for Occupational Safety and Health
2. Collaboration With OSHA
3. General Comments on Line Speed and Worker Safety
4. Inspection Line Speed, Processing Line Speed, and Production Volume
5. Factors Influencing Inspection Line Speed
6. Inspection Line Speed and Inspector Safety Under the NPIS
7. Industry Efforts To Address Worker Safety
8. Reporting of Work-Related Injuries
9. Attestation to FSIS on Work-Related Conditions
G. Changes That Affect All Establishments That Slaughter Poultry Other Than Ratites
1. Procedures and Recordkeeping Requirements for Preventing Contamination by Enteric Pathogens and Visible Fecal Contamination
2. Sampling and Testing Requirements To Monitor Process Control
a. Sampling Plan and Sampling Sites
b. Very Small and Very Low Volume Establishment Sampling
c. Sampling Frequency
d. Indicator Organisms and Baseline
3. Rescind Testing for Generic
E. coli
for Establishments That Slaughter Poultry Other Than Ratites
4. Rescind Codified
Salmonella
Performance Standards
H. Elimination of Time/Temperature Chilling Requirements
I. Online Reprocessing
J. Animal Welfare Considerations
1. Welfare of Live Birds
2. Line Speeds and Animal Welfare
3. Animal Welfare and the Reduction in Number of Online Inspectors
K. Environmental Impact
L. Economic Impact
1. General
2. Environmental Justice
3. Small Business Considerations
4. Implementation Costs
IV. Executive Order 12866 and 13563
V. Final Regulatory Flexibility Act
VI. Executive Order 12988
VII. E-Government Act
VIII. Executive Order 13175
IX. USDA Non-Discrimination Statement
X. Paperwork Reduction Act
XI. Additional Public Notification
XII. Final Regulatory Amendments
I. Background
On January 27, 2012, FSIS published the proposed rule, “
Modernization of Poultry Slaughter Inspection,”
to establish a new inspection system for young chickens and turkeys. Under the proposal, the new poultry inspection system (NPIS) would have replaced the current Streamlined Inspection System (SIS), the New Line Speed Inspection System (NELS), and the New Turkey Inspection System (NTIS). The NPIS that FSIS is adopting in this final rule is consistent with the inspection system that FSIS proposed in January 2012, with modifications, which are described below. However, in this final rule, FSIS is not eliminating SIS, NELS, or the NTIS, as was proposed. This final rule will leave all existing inspection systems in place to give establishments the flexibility to operate under the system that is best suited to their operations.
In the proposed rule, FSIS also proposed changes to the regulations that would apply to all establishments that slaughter poultry other than ratites. FSIS is adopting these proposed changes, with some modifications, which are also described below.
When FSIS issued the proposed rule, it initially gave the public until April 26, 2012, to submit comments. The Agency later extended the comment period until May 29, 2012. The public meeting and the Agency's decision to extend the comment period are discussed below.
Comment Period and Public Meeting
On March 21, 2012, FSIS held a public meeting with its National Advisory Committee on Meat and Poultry Inspection (NACMPI) via Web conference to discuss the January 2012 proposed rule to modernize poultry slaughter inspection. FSIS held the meeting in response to a request from certain members of the committee. At the meeting, FSIS provided an overview of the proposed rule and then held an open discussion with the committee members. A transcript of the public meeting is available on the FSIS Web site at:
http://www.fsis.usda.gov/wps/portal/fsis/topics/regulations/advisory-committees/nacmpi.
When the Agency held the public meeting, the comment period for the proposed rule was scheduled to close on April 26, 2012. At the public meeting, some of the committee members representing consumer advocacy organizations requested that FSIS extend the comment period. A coalition of consumer advocacy organizations also submitted a written request for the Agency to extend the comment period. On April 26, 2012, FSIS announced that it was extending the comment period until May 29, 2012 (77 FR 24873).
In the
Federal Register
document that announced the comment period extension, FSIS explained that during the comment period, the Agency had met with a coalition of consumer advocacy organizations and two trade associations representing the poultry industry to clarify certain aspects of the proposed rule to help inform their comments (77 FR 24873). Because the issues addressed in these meetings may have been relevant to the development of other stakeholders' comments, the
Federal Register
document summarized the issues raised at the meetings and the Agency's responses. In the
Federal Register
document, FSIS also requested additional comments on how it should implement the final rule resulting from the January 2012 proposal. The Agency also requested available data on potential worker safety issues associated with increased line speeds. In addition, the Agency explained that it had
received a request to hold a public technical meeting on the proposed rule, but that the Agency did not believe that such a meeting would be useful.
In developing this final rule, FSIS considered all comments submitted in response to the January 2012 proposed rule, as well as those provided at the NACMPI public meeting held in March 2012. Based on its analysis of the issues and of the information provided by the comments, FSIS made certain changes to, and clarified certain aspects of, the proposed regulations. Those revisions are summarized below and are discussed in detail in the Agency's responses to comments.
II. Summary of Modifications Made to the Proposed Rule
In this document, FSIS is finalizing, with some changes, the provisions in the January 27, 2012, proposed rule “
Modernization of Poultry Slaughter Inspection”
(77 FR 4408). The Agency is modifying the proposal to:
• Change the maximum line speed permitted under the NPIS to 140 bpm for young chickens, for entities that chose to operate under NPIS. The maximum line speed for turkeys will be 55 bpm, as was proposed;
• Leave all existing poultry inspection systems in place and allow young chicken and turkey slaughter establishments that do not choose to operate under the NPIS to continue to operate under their current inspection system;
• Continue to staff all establishments that do not choose to operate under the NPIS with the number of online inspectors that they currently have;
• Allow young chicken establishments that currently operate under HIMP through a
Salmonella
Initiative Program (SIP) waiver to continue to operate under a waiver to run at a maximum line speed of up to 175 bpm;
• Update the SIP waivers for young chicken establishments currently operating under HIMP to remove aspects of HIMP that are inconsistent with the NPIS;
• Establish a phased approach to implement the NPIS in geographic clusters;
• Establish separate applicability dates for large, small, and very small establishments to comply with the provisions in the rule that prescribe the new recordkeeping and microbiological sampling requirements that will apply to all establishments that slaughter poultry other than ratites. The applicability dates will provide additional time for small and very small establishments to comply with these provisions;
• Revise the facilities requirements for the NPIS to require that the online carcass inspection platform be height adjustable;
• Clarify that the records that establishments operating under the NPIS are required to maintain to document that the products resulting from their slaughter operations meet the definition of RTC poultry are subject to review and evaluation by FSIS personnel;
• Revise the proposed regulation that prescribes maximum line speed rates under the NPIS to emphasize establishments' existing legal obligation to comply with the Occupational Safety and Health Administration's regulations;
• Establish a new subpart in the regulations that requires each establishment that participates in the NPIS to submit on an annual basis an attestation to the management member of the local FSIS circuit safety committee stating that it maintains a program to monitor and document any work-related conditions of establishment workers. Current young chicken HIMP establishments that will be operating under the updated SIP waivers described above will be required to submit the annual attestation as a condition of their updated waivers;
• Permit very small and very low volume establishments to conduct sampling for microbial pathogens only at the post-chill point in the slaughter and dressing process to monitor their process control procedures instead of requiring sampling at pre- and post-chill, as was proposed;
• Prescribe a minimum frequency with which all establishments that slaughter poultry other than ratites will need to conduct testing for microbial organisms to monitor the effectiveness of their process control procedures; and
• Revise the definition for “air chill” to allow an antimicrobial intervention to be applied with water at the beginning of the chilling process if its use does not result in any net pick-up of water or moisture during the chilling process. The initial antimicrobial intervention may result in some temperature reduction of the product if the majority of temperature removal is accomplished exclusively by chilled air.
In addition, because the proposed pre-and post-chill sampling requirements will not apply to ratite slaughter establishments, FSIS is retaining the generic
E. coli
testing regulations as they apply to ratites only, but is rescinding the provisions in these regulations that apply to all other poultry classes. Poultry establishments other than establishments that slaughter ratites will be required to comply with the new sampling requirements prescribed in this final rule.
III. Comments and Responses
FSIS received over 250,000 comment letters in response to the January 2012 proposed rule. Most comments were submitted as part of organized write-in campaigns. The Agency also received a petition that included approximately 150,000 signatures and form letters before the comment period closed. The Agency received two petitions in November 2012, after the comment period had closed. One of these petitions included approximately 180,000 signatures and 13,000 comments, and the other included over 3,500 signatures. FSIS received an additional petition in September 2013 with approximately 43,000 signatures. All of the petitions requested that the Agency withdraw the proposed rule. The issues raised in the petitions and comments submitted in November 2012 and September 2013 are similar to the issues raised by the petition and comments submitted during the comment period. Therefore, the Agency will address the issues raised in all of the petitions and associated comments in this document.
Most of the individual comments were submitted as part of various write-in campaigns initiated by consumer advocacy organizations, labor unions, animal welfare organizations, and worker and human rights advocacy organizations. FSIS also received individual comments from private citizens, inspection personnel, and members of labor unions.
In addition to the individual comments, form letters, and petitions, the Agency also received approximately 120 separate comment letters from trade associations representing the poultry industry, companies that conduct poultry slaughter operations, consumer advocacy organizations, public health organizations, labor unions, animal welfare advocacy organizations, members of academia, a State Department of Agriculture, and worker/immigrant/human rights advocacy organizations. Following is a summary of the comments and FSIS's responses.
A.
NACMPI Meeting and Public Process
Comments:
Several consumer advocacy organizations expressed their concern that FSIS published the proposed rule in the
Federal Register
before it consulted with the NACMPI. According to the comments, the Agency
is required to consult with members of the NACMPI before proposing changes to its meat and poultry inspection program, and that the Agency should have consulted with the NACMPI before publishing the proposed rule to modernize poultry slaughter inspection.
Response:
FSIS held the March 21, 2012, NACMPI public meeting in response to a request from certain committee members representing consumer advocacy organizations that the Agency convene the committee to discuss the proposed rule. At the meeting, FSIS made clear that it was interested in the committee's comments and suggestions, but that the Agency was not seeking consensus from the committee.
FSIS disagrees that the Agency was required to consult with the NACMPI before proposing changes to its poultry inspection program. Under the Federal Meat Inspection Act (FMIA) and Poultry Products Inspection Act (PPIA), the Secretary is authorized to “appoint advisory committees consisting of such representatives of appropriate State agencies . . . to consult with him concerning State and Federal programs with respect to [meat and poultry] inspection and other matters within the scope of this chapter . . .” (21 U.S.C. 661(a)(4) and 21 U.S.C. 454(a)(4)). The Secretary of Agriculture established the NACMPI to provide advice concerning State and Federal programs with respect to meat and poultry inspection, food safety, and other matters that fall within the scope of the FMIA and PPIA. Under the NACMPI Charter, FSIS consults with the committee in carrying out its specific responsibilities under 21 U.S.C. 607(c), 624, 645, 661(a)(3), and 661(c) of the FMIA and 21 U.S.C. 454(a)(3), 454(a)(4), 454(c), 457(b), and 460(e) of the PPIA. These sections address: Type styles and sizes of labeling; definitions and standards of identity or composition; standards of fill of container; consistency of Federal and Federal-State standards; storage and handling regulations; exemption of establishments subject to non-Federal jurisdiction; Federal provisions applicable to State or Territorial business transactions of a local nature and not subject to local authority; scope of cooperation; and State meat inspection requirements. Thus, the NACMPI charter does not require that FSIS consult with the NACMPI before proposing changes to its poultry inspection program, although the Agency conducted a public meeting after the proposed rule was issued to seek feedback on the proposal.
Comment:
Some consumer advocacy organizations noted that FSIS decided not to hold a technical public meeting as requested by a coalition of consumer advocacy organizations.
Response:
As stated in the
Federal Register
comment period extension document, FSIS decided not to hold a public technical meeting on the proposed rule because the Agency did not believe that such a meeting would be useful (77 FR 24873). In April 2012, in response to a request from a group of consumer advocacy organizations, FSIS extended the comment period for the proposed rule. In the
Federal Register
document that announced the comment period extension, FSIS summarized issues that were raised in separate meetings with consumer and industry stakeholders and clarified certain aspects of the proposed rule to help inform stakeholder comments. In that document, the Agency also provided additional information on worker safety issues and its tentative strategy to implement the NPIS, and it solicited comments and data on both issues. As such, FSIS provided the public with all of the information it might have during a technical meeting, but through the public comment process. Thus, the process for developing this final rule was open and transparent and provided several opportunities for stakeholder input.
Comment:
One public health association said that FSIS failed to comply with E.O. 12866 and E.O. 13563 requirements with respect to public participation. The comment said E.O. 13563 requires that agencies make all of the documents they rely on to justify rules available to the public, and FSIS did not do so. According to the comment, as of May 19, 2012, more than 80 days after the proposal was published, there were only two documents in the public record posted by USDA at Regulations.gov, the January 27, 2012, and April 26, 2012,
Federal Register
document. The comment said that only 12 records are posted on the FSIS Web site. According to the comment, the public is unable to provide informed comments when the underlying records used to develop the proposed rule are not available for review.
A labor union criticized the Agency for publishing a complex statistical analysis while providing little raw data in the supporting documents. The comment also questioned whether the comment period for the proposed rule provided sufficient time for stakeholders to adequately consider the supporting data.
Response:
The Agency plans to post supporting documentation for this final rule and future Agency rulemakings on Regulations.gov. Although FSIS acknowledges that the underlying records used to develop the proposed rule were not posted on Regulations.gov, the proposed rule and all related documents, including supporting materials, were posted on the FSIS Web site when the proposed rule published in the
Federal Register
. The supporting materials included the Evaluation of the HACCP-Based Inspection Models Project; the draft 2011 FSIS Risk Assessment for Guiding Public Health-Based Poultry Slaughter Inspection; the Agency's response to Peer Review Comments on its draft 2008 Risk Assessment for Guiding Public Health Risk-Based Poultry Slaughter Inspection; and the On-Line and Off-Line Reprocessing In-Plant Trial Analysis. The supporting data for the analyses in the Evaluation of the HACCP-Based Inspection Models Project are presented in tables in the report and in the appendices. The data and modeling methods used in the 2011 FSIS Risk Assessment for Guiding Public Health-Based Poultry Slaughter Inspection are also fully described in the Appendix to that document.
The proposed rule and the
Federal Register
document extending the comment period for the proposed rule were posted on both the FSIS Web site and Regulations.gov when those documents published in the
Federal Register
. The preamble to the proposed rule includes the FSIS Web site link to the related materials and supporting documents, and it explains that these documents are also available in the FSIS docket room. These materials have been available on the Agency's Web site during the entire comment period and remain available at:
http://www.fsis.usda.gov/wps/portal/fsis/topics/regulations/federal-register/proposed-rules/proposed-rules-2012/!ut/p/a1/jZDBCoJAEIafpQeQnVURPdqCpaUikdleYsHVFsxdVuvQ06d0UpKcOf3w_XzMIIoKRFv2EjXrhWxZM2bq3CADB3sEojTwAwgTK8jdZIfBxgNwnQAeHoE8Sw-EgJtYK_sL48O_frRCYOqYxDWiivV3Q7SVREXFS65ZY2hei67nGhVKSyU7Xhr62fBung0TsIkuiE51gIcddCd7HyUWpPYc-PGPL7B8sHqci_dx64XC33wAFla5ew!!/?1dmy¤t=true&urile=wcm%3apath%3a%2Ffsis-content%2Finternet%2Fmain%2Ftopics%2Fregulatory-compliance%2Fhaccp%2Fhaccp-based-inspection-models-project%2Fhimp-study-plans-resources%2Fpoultry-slaughter-inspection.
With respect to the comment that said that FSIS did not provide sufficient time for public comment, E.O. 12866, as supplemented by E.O. 13563, states that agencies are to “afford the public . . .
with a comment period that should generally consist of not less than 60 days.” FSIS provided a 90-day comment period for the proposed rule and then extended it for an additional 30 days. The Agency believes that the public had ample time to consider the issues raised in the proposed rule and supporting documentation in order to develop their comments.
Comment:
A consumer advocacy organization criticized the Agency for including the anticipated cost savings from the proposal in the Agency's 2013 proposed budget to Congress before the public comment period for the proposal closed.
Response:
The Agency concluded that an open, transparent, and effective budgetary process requires that the Agency report on the rule and the associated estimated budget. In addition, the Appropriations Committee Report that accompanied the FY 2013 appropriations bill directs the Agency to notify the Committee of the status of the rule not later than September 15, 2012.
2
2
House Appropriations Committee report, p. 23 (
http://appropriations.house.gov/uploadedfiles/hrpt-112-ap-fy13-agriculture.pdf
).
B.
The HIMP Report
In the proposed rule, FSIS explained that it was proposing to establish a new system of inspection for young chickens and turkeys based on its experience under the HACCP-based Inspection Models Project (HIMP) pilot study (77 FR 4421). As discussed in the proposal, FSIS initiated the HIMP pilot study in 20 young chicken and 5 turkey slaughter establishments on a waiver basis after the Agency implemented the 1996 HACCP regulations. Similar to the NPIS, under HIMP, establishment personnel are responsible for sorting carcasses, disposing of carcasses affected with conditions that would require that they be condemned, and conducting any trim and reprocessing that they believe necessary to correct removable defects.
In the HIMP inspection system, a single FSIS online carcass inspector (CI) visually inspects every carcass at a fixed point on the evisceration line immediately before the chiller. Under HIMP, an offline verification inspector (VI) is responsible for conducting system verification activities that the Agency has concluded will be more effective in ensuring food safety, such as conducting offline carcass verification checks for septicemia/toxemia and visible fecal contamination, collecting samples for pathogen testing, and verifying the effectiveness of an establishment's HACCP system by, among other activities, reviewing the establishment's HACCP plan and HACCP monitoring records, observing establishment employees performing tasks specified in the HACCP plan, reviewing and determining the adequacy of the corrective actions taken by the establishment when a deviation occurs, and conducting measurements of critical control points (CCPs). The Agency analyzed the data collected from the HIMP study and prepared a written report that presents an evaluation of the model tested (see the “HIMP Report,” available on the Agency's Web site at:
http://www.fsis.usda.gov/wps/wcm/connect/fcd9ca3e-3f08-421f-84a7-936bc410627c/Evaluation_HACCP_HIMP.pdf?MOD=AJPERES
).
The HIMP Report assesses FSIS inspection findings across four interrelated inspection activities:
1. Inspection of each carcass by the CI to determine whether the carcass is not adulterated and thus eligible to bear the mark of inspection.
2. Verification by VIs of the establishment's execution of its HIMP process control plan, under which establishment employees sort acceptable and unacceptable carcasses and parts.
3. Verification of the establishment executing its sanitation SOPs and HACCP system.
4. Verification of the outcomes of the establishment's HIMP process control plan, both organoleptic and microbiological.
Inspection of each carcass by the CI to determine whether the carcass is not adulterated.
Based on an analysis of data collected from April 1, 2009, through March 31, 2011, the HIMP Report found that fewer than 0.0008 percent of the carcasses presented to the CI were affected with septicemia/toxemia, and fewer than 0.08 percent had visible fecal contamination. Despite these low rates, the CIs in HIMP establishments detected carcasses affected with septicemia/toxemia at a rate of 0.000004 percent or 4 per 100 million carcasses slaughtered and carcasses with visible fecal contamination at a rate of 0.0009 percent or 9 per 1 million carcasses slaughtered.
Verification by VIs of the establishment's execution of its HIMP process control plan.
The HIMP Report compares the ratio of all offline inspection procedures conducted in HIMP and non-HIMP establishments in calendar year (CY) 2010. FSIS inspectors in HIMP establishments perform offline inspection procedures to verify that the establishments are properly executing their HIMP process control plans. This comparison shows that overall in CY 2010, FSIS offline inspection personnel performed 1.6 times more offline inspection procedures in HIMP establishments than in non-HIMP establishments.
Verification of the establishment executing its sanitation SOPs and HACCP system.
The sanitation SOP and HACCP regulations are among the regulations most strongly related to public health. The HIMP Report's comparison of the ratio of offline inspection procedures performed in HIMP and non-HIMP establishments in CY 2010 shows that FSIS offline inspectors in HIMP establishments performed about 3.0 times more sanitation SOP and HACCP inspection procedures than offline inspectors performed in non-HIMP establishments. It also shows that offline inspectors in HIMP establishments performed 3.4 more HACCP procedures that include random verification of all HACCP requirements than inspectors in non-HIMP establishments.
The HIMP Report also compares health-related non-compliances in HIMP and non-HIMP establishments from CY 2006 through CY 2010. These data show that health-related non-compliance record (NR) rates at HIMP establishments are not statistically different from or are statistically lower for all inspection procedures considered. The HIMP Report also found that the rate of health-related non-compliances for visible fecal contamination from CY 2006 through CY 2010 is about 1.6 times lower in HIMP establishments than in non-HIMP establishments.
Verification of the outcomes of the establishment's HIMP process control plan, both organoleptic and microbiological.
To assess the outcomes of establishment's process control plans in addressing visible food safety defects and defects related to the wholesomeness or quality of the product, referred to as “other consumer protection” (OCP) defects, FSIS developed performances standards for these defects based on the performance of non-HIMP establishments. The performance standards allow the Agency to compare the performance of establishments operating under HIMP and non-HIMP inspection systems in controlling visible food safety and OCP defects.
A comparison of the findings of the offline VIs in HIMP establishments for the two-year period April 1, 2009, to March 31, 2011, with the HIMP food safety defect performance standards show that the rate of septicemia/toxemia in carcasses processed in HIMP establishments (8 per 1 million or 0.0008 percent) is 125 times lower than
the HIMP performance standard (0.1 percent). The HIMP Report also found that the rate of visible fecal material on carcasses processed in HIMP establishments (fewer than 0.8 per thousand or 0.08 percent) is 19 times lower than the HIMP performance standards (1.5 percent). A comparison of the findings of the offline VIs in HIMP establishments for the two-year period January 1, 2009 through December 31, 2010, with the HIMP OCP performance standards show that OCP defects identified on carcasses processed in HIMP establishments averaged about half the corresponding OCP HIMP performance standard.
To assess the microbiological outcomes of HIMP establishments' process control plans, the HIMP Report analyzed data from FSIS's
Salmonella
verification testing program collected from CY 2006 through CY 2010. The HIMP Report compares the
Salmonella
percent positive rates in 20 HIMP broiler establishments, 64 non-HIMP comparison establishments, and all 176 non-HIMP broiler establishments. The analysis shows that
Salmonella
positive rates in HIMP establishments average about 80 percent of those in non-HIMP establishments.
In the preamble to the proposed rule, FSIS explained that the Agency had concluded, based on analysis of the two-year data sets of food safety and OCP defects, that establishments operating under the HIMP inspection system performed better than establishments operating under non-HIMP inspection systems with respect to rates of food safety defects and OCP defects that may affect the wholesomeness or quality of the product (77 FR 4419). Data on health-related NRs collected from CY 2006 through CY 2010 show that non-compliances for fecal contamination are lower in HIMP than in non-HIMP establishments and that HIMP establishments have a higher compliance with sanitation SOP and HACCP regulations. HIMP establishments also had equivalent or lower
Salmonella
positive rates than non-HIMP establishments. The Agency explained that it was proposing to establish a new poultry inspection system informed by HIMP that would replace the SIS, NELS, and NTIS inspection systems for young chickens and turkeys (77 FR 4421).
FSIS received several comments on the HIMP Report and the Agency's analysis of the data collected under the HIMP study. Comments from the poultry industry and trade associations representing the poultry industry generally agreed with the findings of the HIMP Report and supported the Agency's decision to establish a new poultry inspection system. Comments from private citizens, consumer advocacy organizations, labor unions, and members of academia raised issues and concerns regarding the data collected under HIMP and the Agency's conclusions based on the HIMP study results.
1. Data and Methods Used in the HIMP Report
Comment:
Several comments from consumer advocacy organizations and private citizens questioned whether data collected under that HIMP study should be used to inform the NPIS. The comments said that the HIMP pilot has never been independently evaluated to determine whether the establishments operating under the HIMP inspection system are producing food that is as safe as product produced in establishments operating under non-HIMP inspection systems.
Response:
FSIS disagrees with the comment. In 2002, after the Government Accountability Office (GAO) issued its December 17, 2001, report on HIMP
3
(referred to as the “2001 GAO report”), FSIS contracted with a technical review team selected by the National Alliance for Food Safety to review and evaluate the data collected from young chicken establishments operating under HIMP. The review team focused on the validity of the HIMP study design and methodology to determine whether FSIS could use the organoleptic and microbial data collected under HIMP to compare the performance of establishments operating under HIMP and non-HIMP inspection systems. Overall, the review team found that the HIMP study design and methodology were valid and provided a useful and legitimate comparison of the HIMP and non-HIMP inspection systems. The review team's findings are described in the report: “
Review of the HACCP-Based Inspection Models Project by the National Alliance for Food Safety Technical Team
”
4
(also referred to as “The Hargis Report”).
3
GAO, 2001. Food Safety: Weaknesses in Meat and Poultry Inspection Pilot Should Be Addressed Before Implementation:
http://www.gao.gov/new.items/d0259.pdf
.
4
The Hargis Report is available for viewing by the public in the FSIS docket room and on the FSIS Web site at:
http://www.fsis.usda.gov/OPPDE/nacmpi/Nov2002/Papers/NAFS97.pdf
.
As stated in the report, “[t]he review team noted some issues related to optimal design and interpretation, but finds that overall the data collected were both meaningful and useful and that the study was designed and conducted under real-world conditions and limitations.” The review team also concluded that “the overall design and methodology . . . were perhaps the best available options to allow for comparison of organoleptic data between the traditional and HIMP systems.”
Comment:
One consumer advocacy organization noted that the HIMP Report said that the Agency's evaluations of microbiological and inspection findings are based on data for calendar years (CY) 2006 through 2010, with certain exceptions where only more recent data are available. According to the comment, the HIMP Report does not explain why certain data are missing or why time periods for comparisons are not uniform. The comment noted that the Agency only analyzed data from CY 2010 when comparing the ratio of offline inspection procedures performed in HIMP and non-HIMP establishments.
Response:
The time periods for the data that were analyzed for the HIMP Report vary because not all data were available as computerized data sets. Data on the number of carcasses affected with food safety and OCP defects were not available as computerized data sets. FSIS field personnel manually collected these data and recorded the results on paper forms. To reduce the burden on its field personnel, FSIS decided that an analysis of two years' worth of these non-computerized data sets would be sufficient. The HIMP report data for the number of carcasses affected with food safety defects is from April 1, 2009, through March 31, 2011, and data for carcasses affected with OCP defects is from January 1, 2009, through December 31, 2010.
In the body of the HIMP Report, the Agency used computerized data collected from CY 2010 to compare the ratio of offline inspection procedures performed in HIMP and non-HIMP establishments. The Agency used data from 2010 for this analysis because it was the most recent data available. Tables C-2 and C-3 in the Appendix of the HIMP Report contain summary information on non-compliances with sanitation SOP and HACCP regulations and on the number of inspection procedures in HIMP and non-HIMP establishments from CY 2006 through 2010. The data for these years are similar to the data from CY 2010.
Comment:
One comment noted that in the preamble to the proposed rule, the Agency compares findings (1) by VIs of OCP defects between January 1, 2009 and December 31, 2010; (2) by VIs of food safety defects between April 1, 2009 and March 31, 2011; and (3) by CIs of food safety defects between April 1, 2009 and March 31, 2011. The comment
said that while these time periods are not very different, it is possible that the slight shifts were made to conceal results that would be less supportive, or that would even contradict Agency claims.
Response:
The two-year period January 1, 2009 to December 31, 2010 was used to evaluate OCP defects, while the two-year period April 1, 2009, to March 31, 2011 was used to evaluate compliance with the HIMP food safety standards. Both of these comparisons used the most recent data available at the time. This is the reason for the different time periods.
2. HIMP as the Basis for the NPIS
Comment:
A trade association representing the poultry industry stated that the HIMP pilot program has been successfully carried out for the last 13 years. The comment said that during that time, food safety records in establishments operating under the HIMP inspection system have been as good as those in non-HIMP establishments. The comment stated that the equivalent or lower pathogen rates in HIMP establishments compared to non-HIMP establishments, as documented in the HIMP Report, are evidence that the program has been successful. The comment noted that this success is especially significant given that the review team selected by the National Alliance for Food Safety determined that food safety performance standards provide a scientifically valid measure by which performance of HIMP establishments can be evaluated (Hargis et al. 2002). The comment stated that, based on the data, the trade association agreed with the Agency's conclusion that the NPIS is a positive step toward enhancing food safety.
On the other hand, several consumer advocacy organizations questioned whether it is appropriate for FSIS to use the HIMP study results to predict how establishments will perform when operating under the NPIS. The comments noted that the 2001 GAO report criticized FSIS for not randomly selecting establishments for the HIMP pilot study and questioned whether the data generated by the pilot could be used to predict how all of the young chicken establishments would perform if FSIS were to adopt the HIMP inspection system nationwide.
Several comments stated that because participation in the HIMP study was voluntary and required that poultry establishments meet additional food safety and OCP performance standards, participating establishments could be viewed as high performers with respect to food safety. The comments asserted that for this reason, data from the HIMP pilot may not represent what FSIS is likely to see when the majority of young chicken and turkey slaughter establishments begin to operate under the NPIS.
Response:
The trade association comments support the agency proposal. With regard to concerns raised by the consumer advocacy organizations, FSIS addressed these issues in its comments on and response to the 2001 GAO Report. In that document, FSIS stated that although not randomly selected, there is evidence that volunteer establishments participating in the HIMP study are typical of the industry. The volunteer establishments represent diversity in geography, corporate structure, management styles, number of evisceration lines, product distribution patterns, inspection system in use prior to the pilot, and other variables. In addition, the Hargis Report, discussed above, noted that the establishments selected for the HIMP pilot represent the States supplying the majority of domestic chicken production and the size range of establishments included in the study are representative of almost 90 percent of chickens slaughtered in federally-inspected facilities in the United States. The Hargis Report noted that establishment design, equipment, and procedures within poultry establishments are relatively uniform. The report concluded that “[i]t is very difficult to hypothesize a geographic or plant-selection bias in this study.”
Comment:
Two consumer advocacy organizations stated that the NPIS is not an exact replica of the HIMP pilot, which raises further concerns about whether results from the HIMP pilot accurately reflect how establishments will perform under the NPIS.
Response:
Although the NPIS is not an exact replica of HIMP, the NPIS was informed by the data collected under HIMP. These data demonstrate that an inspection system that combines the features described in this document, which include carcass sorting by establishment employees, a CI that conducts an inspection of each carcass before the chiller, and, most important, a VI that conducts more offline inspection activities that specifically focus on food safety, does not reduce the effectiveness and may, in fact, lead to better compliance with sanitation and HACCP regulations and in carcasses with lower levels of fecal contamination and equivalent or lower levels of
Salmonella
contamination.
In addition, as discussed in detail below, in the 2014 risk assessment, analysis of historical data shows a statistically significant correlation between specifically targeted unscheduled offline inspection procedures and reductions in
Salmonella
positive samples in young chicken slaughter establishments and
Campylobacter
positive samples in young turkey slaughter establishments. Modeled scenarios involving an increase in targeted inspection activities (specifically unscheduled offline inspection activities, rather than a randomly selected set of activities) suggest that implementing the NPIS would likely result in public health benefits. Assuming that the number of offline inspection procedures performed in all poultry slaughter establishments increase proportionately to the number of such procedures currently performed in HIMP establishments, FSIS's risk model predicts a likely public health benefit. Consistent with the underlying assumptions of the model, it is reasonable to conclude that inspection systems in which Agency resources continue the core online inspection activities while enhancing the frequency and focus of unscheduled offline activities directly related to food safety, such as HIMP and the NPIS, would likely result in a lower prevalence of carcasses contaminated with
Salmonella
and
Campylobacter,
which in turn would likely lead to fewer human illnesses.
Comment:
One consumer advocacy organization criticized the Agency's evaluation of HIMP. The comment stated that the HIMP Report compares the current performance of HIMP establishments with performance levels observed from 1998-2000 when FSIS collected baseline data from establishments that later joined the HIMP pilot.
The comment also stated that the Agency failed to explain how the performance level of the bottom four establishments that entered the HIMP pilot is representative of approximately 200 other establishments more than a decade later.
Response:
The Hargis Report, described above, concluded that the design of the HIMP pilot “is generally appropriate for a field study of this nature, and the methodologies employed generally allow for interpretation and comparison of [HIMP versus non-HIMP inspection systems.]” The Hargis Report also concluded that comparison of HIMP food safety and OCP performance levels with performance standards does provide a scientifically valid measure by which changes in food safety and OCP performance under HIMP can be assessed.
With respect to the comment that suggests that the HIMP OCP performance standards represent the performance level of the bottom four establishments that entered the HIMP pilot, the HIMP OCP performance standards are set at the 75th percentile of what was achieved under the Research Triangle Institute (RTI) baseline study of 16 young chicken establishments under non-HIMP inspection systems before they entered the HIMP study. Thus, the performance standards were set so that 25 percent of the establishments that entered HIMP would have to improve upon their baseline results in order to meet the more stringent standards.
3. Carcass Inspection Under HIMP
In the preamble to the proposed rule, FSIS explained that the Agency concluded that establishments operating under the HIMP inspection system performed better than establishments operating under non-HIMP inspection systems with respect to rates of food safety and OCP defects (77 FR 4419). With respect to food safety-related defects, the Agency noted that data collected from the HIMP study show that the levels of carcasses affected with septicemic or toxemic conditions (also referred to as “septicemia/toxemia”) or visible fecal contamination in HIMP establishments is very low (77 FR 4415). The HIMP Report concluded that notwithstanding these very low levels, the data demonstrate that CIs in HIMP establishments effectively identify carcasses affected with septicemia/toxemia and visible fecal contamination. Several consumer advocacy organizations commented on this conclusion.
Comment:
Some consumer advocacy organizations stated that the CI detection rate for visible fecal contamination and septicemia/toxemia is based on the assumption that the rates at which VIs detect these food safety-related conditions represents the level at which these conditions occur in the establishment. The comments questioned this assumption. The comments noted that in HIMP establishments, the VI collects eight 10-bird verification samples per line per shift. The comment asserted that there is no evidence to indicate that this sample size is sufficient to represent the true level of food safety defects on carcasses throughout the shift.
Response:
FSIS disagrees that the CI detection rate is based on the assumption that the rate at which VIs detect carcasses affected with septicemia/toxemia or visible fecal contamination represents the level at which these conditions occur in the establishment. The CI detection rate is the rate at which CIs in HIMP establishments detected carcasses with these food safety-related conditions before the carcasses entered the chiller. It is not based on the VI detection rate.
FSIS believes that its sampling for food safety defects under HIMP is sufficient to reflect the level of food safety defects on carcasses processed in HIMP establishments. Statistically, given the sample design, the precision of an estimate of an establishment's level of food safety defects depends primarily on the total number of samples for an establishment collected over time.
The food safety performance standards, which are based on thousands of samples collected by a 3rd party contractor and reflect the level of food safety defects on carcasses processed in establishments before they entered the HIMP pilot, vary by defect category. The performance standard for septicemia/toxemia is 0.1 percent, and the performance standard for visible fecal contamination is 1.5 percent. When deciding the number of samples that FSIS should take to reflect an establishment's level of food safety defects over time, FSIS determined that collecting 80 birds per line per shift would provide an estimated defect rate that was close to the true defect rate.
For example, if the true defect rate for visible fecal contamination was 0.1 percent at an establishment that operated one line for two shifts, 300 days per year, taking an 80 bird sample per line per shift would give a total of 48,000 samples a year, per line. This number of samples, assuming a random distribution of defects throughout the year, would give FSIS an estimated defect rate between 0.72 and .128 percent with about 95 percent probability. Thus, FSIS believes that the specified sample size is sufficient to make general comparisons of average defect rates among establishments or lines.
Comment:
One consumer advocacy organization stated that another reason that the VI detection rate may not represent the actual level of food safety-related defects in HIMP establishments is that statements it obtained from HIMP inspectors indicate that establishment employees take greater care to prevent and remove visible fecal contamination and to identify and remove septicemic/toxemic carcasses when they know that the VI inspector is getting ready to take a sample.
Response:
The comments seem to suggest that establishment employees are able to manipulate the results of the VI's verification checks. FSIS disagrees.
As noted above, VIs in HIMP establishments collect scheduled verification samples that consist of eight 10-bird samples per line per shift. VIs also collect targeted, unscheduled 10-bird samples in response to VI or CI findings of excessive food safety or OCP carcass defects.
VIs in HIMP establishments collect scheduled and unscheduled verification samples for septicemia/toxemia and visible fecal contamination using the same offline verification methodology that offline inspectors in non-HIMP establishments use to collect samples for visible fecal contamination checks. In both HIMP and non-HIMP establishments, offline inspectors do not inform establishment employees when they collect verification samples and, equally important, take care to ensure that the samples represent the operating conditions in the establishment. Thus, there is no reason to believe that employees in HIMP establishments have any significant opportunity, and certainly no additional opportunity, to affect the results of the verification checks.
Comment:
Some consumer advocacy organizations said that the data in the HIMP Report do not support the Agency's conclusion that CIs are able to identify carcasses affected with visible fecal contamination and septicemia/toxemia.
With respect to visible fecal contamination, one consumer advocacy organization stated that the data presented in the HIMP Report indicate that CIs did not detect 88 out of 89 birds with fecal contamination going down the line. The comment stated that the inspectors in the VI position who were able to examine both the inside and the outside of the bird detected visible fecal contamination on the carcass at approximately 90 times the rate that the CIs detected it. Another said that based on the data, it is reasonable to calculate that CIs failed to detect over a quarter of a million carcasses with fecal contamination in the 20 HIMP establishments within the two-year period of data collection.
With respect to septicemia/toxemia, one comment said that data presented in the HIMP Report indicate that CIs detect approximately 1 of every 200 carcasses affected by septicemia/toxemia. The comment said that this means that the CI does not detect 199 of every 200 carcasses affected with septicemia/toxemia.
Response:
FSIS disagrees with the commenters' conclusions. The commenters' assessments are based on a
comparison of the results of the CI's carcass inspection and the VI's carcass verification checks and do not take into account the difference between the role of the CI and VI under HIMP. Under HIMP, the inspections performed by the VI and CI serve different purposes and are not done in the same way. Thus, the rate at which VIs identify food safety defects when conducting offline verification checks is not an appropriate basis for assessing whether the CI is conducting an effective inspection of each carcass leaving the slaughter line.
Under the HIMP inspection system, the VI and CI have different but complementary roles in ensuring that poultry products leaving the slaughter line are safe and wholesome. CIs are responsible for conducting a continuous online inspection of each carcass to determine whether it is not adulterated. The VI's role is very different. VIs collect carcass samples before the CI inspection station after the establishment has conducted sorting, trimming, and reprocessing activities to monitor and evaluate the establishment's process controls. The samples collected by VIs may be either “scheduled” or “unscheduled.”
On the one hand, VIs collect eight randomly selected 10-bird samples per line per shift. These are referred to as the “scheduled” samples because the IIC schedules the collection of the eight sample sets before each shift. On the other hand, VIs also collect targeted, unscheduled 10-bird samples as directed by the IIC in response to VI or CI findings of excessive food safety or OCP carcass defects. These samples are in addition to the 80-bird scheduled samples. Because the VI's unscheduled samples are collected when excessive carcass defects have been identified, the results typically show higher rates of carcass defects than the VI's scheduled sampling results.
The VI detection rates in the HIMP report reflect the combined results of the VI's scheduled and unscheduled sampling and are thus are much higher than the rates that would have resulted had the VI only performed scheduled carcass sampling. Because CIs under HIMP perform an online inspection of each carcass, the CI detection rates are not subject to the same sampling bias introduced by the unscheduled sampling that VIs perform during high defect periods. Significantly, FSIS has not captured what percent of the defects found by VIs were found in scheduled as opposed to unscheduled sampling. Thus, the VI and CI detection rates are not comparable. Therefore, as stated above, the comparisons of the VI and CI detection rates cited by the comments do not provide a valid assessment of the CI's ability to conduct an effective online carcass inspection.
4. Public Health-Related Non-Compliances
In the proposed rule, the Agency noted that the HIMP Report shows that HIMP establishments have public health-related non-compliance record(NR) rates that are not statistically different from or that are statistically lower than the rates for non-HIMP establishments (77 FR 4416-4417). The Agency also noted that HIMP establishments had fewer NRs for visible fecal contamination than non-HIMP establishments. Several consumer advocacy organizations, FSIS inspectors, and a labor union commented on these conclusions.
Comment:
Comments from inspectors, labor unions, and consumer advocacy organizations stated that the location of the establishment's critical control point (“CCP”) for food safety defects may prevent a CI from issuing an NR even if the CI detects such a defect. The comments noted that at the start of the HIMP pilot, the CCPs for visible fecal contamination and septicemia/toxemia were located before the FSIS carcass inspection station. The comments stated that before FSIS began collecting data to support the proposed rule, the Agency allowed the HIMP establishments to move their CCPs for fecal contamination and septicemia/toxemia to points after the CI. One comment said that the timing for allowing establishments to move CCPs to a point after the CI suggests that the primary purpose was to reduce the number of NRs issued to HIMP establishments for these conditions. Another comment said that the fact that CIs cannot issue an NR if they observe food safety defects before the CCP, affects the HIMP Report's CI detection rate statistics.
Response:
FSIS disagrees with the suggestion that the location of the CCP with respect to the CI affects the comparison of NR rates between HIMP and non-HIMP establishments. The HIMP Report's analysis of NRs for visible fecal contamination in HIMP and non-HIMP establishments is based on a comparison of visible fecal NRs detected through offline verification activities, not on the CI detection rate, as suggested by one of the comments.
As noted above, the VI under HIMP collects carcass samples after establishment employees have sorted and trimmed the carcasses, but before the carcasses are presented to the CI. If the VI detects visible fecal contamination offline, the VI issues an NR because the establishment violated the Agency's zero tolerance for visible fecal contamination. If a CI observes a carcass with visible fecal contamination the CI stops the line to prevent the carcass from entering the chiller. The location of the establishment's CCP for food safety defects does not affect the CI's or VI's duties under HIMP. Thus, because the NR rate for visible fecal contamination under HIMP is based on the VI detection rate, the location of the CCP with respect to the CI inspection station does not affect the HIMP Report's analysis of visible fecal NRs.
With respect to the comment that suggested that the location of the CCP affects the CI detection rate statistics, the CI detection rate reflects the rate at which CIs stop the line to prevent carcasses with food safety defects from entering the chiller. Thus, contrary to the commenter's suggestion, the location of the CCP after the CI inspection station does not affect the CI detection rate.
Comment:
One comment stated that the Agency provided no information to demonstrate that documentation policies and opportunities for documenting public health-related NRs were the same in HIMP and non-HIMP establishments. The comment stated that the 2001 GAO report on HIMP noted that after the switch to HIMP, a substantial number of establishments saw increased fecal NR rates. The comment said that the GAO report cited increased line speeds under HIMP as a potential factor for the increased rate of fecal NRs. The comment said that these findings suggest that the transition to HIMP may result in increased rates of fecal contamination.
Response:
As noted in the Agency's comments on the 2001 GAO report, under HIMP, the Agency performs verification checks on approximately 80 carcasses per line per shift as opposed to verification on approximately 20 carcasses per line for fecal contamination under non-HIMP broiler inspection. In addition, VIs under HIMP perform more offline inspection activities that FSIS has concluded are more effective in ensuring food safety than inspectors perform in non-HIMP establishments. Thus, FSIS inspectors in HIMP establishments have more opportunities for detecting non-compliances with regulatory requirements that are directly related to public health than inspectors do in non-HIMP establishments. The procedures for documenting public-health related NRs are the same for both HIMP and non-HIMP establishments.
Although the GAO report cited increased line speeds in HIMP establishments as a potential factor for
the increased rate of fecal NRs, the Agency is not aware of any data to support this hypothesis. The increased rates of fecal NRs that occurred at the beginning of the HIMP pilot could just as easily be the result of increased monitoring under the HIMP inspection system rather than an increase in fecal contamination. Further, the final rule includes a maximum line speed of 140 bpm under the NPIS rather than the 175 bpm allowed in the HIMP pilot.
Comment:
A consumer advocacy organization stated that it had recently acquired records of NRs written for visible fecal contamination within the last year from two HIMP establishments and two non-HIMP establishments. The comment stated that to the best of the commenter's knowledge, all of the establishments are large establishments with two production lines and two production shifts. The comment said that the non-HIMP establishments had 19 and 23 NRs for visible fecal contamination, respectively, and the HIMP establishments had 93 and 173 visible fecal NRs, respectively. The comment stated that these comparisons add to the concerns that the lower NR rates for HIMP establishments described in the HIMP Report may not be good indicators of the actual level of food safety defects on carcasses.
Response:
Because the consumer advocacy organization did not indicate where it obtained the data or which establishments the data are from, FSIS is unable to respond to the comment in detail.
The HIMP Report's comparison of visible fecal NRs issued from offline verification checks in HIMP and non-HIMP establishments compares `rates,' which adjust for the number of samples taken. The report shows that fecal NR rates at HIMP establishments are statistically lower than those in both the control set of 64 non-HIMP establishments and the 176 all non-HIMP comparison set. In addition, the rate of visible fecal material contamination on carcasses in HIMP establishments is about half that in non-HIMP establishments. Thus, when the sample is viewed as a whole and rates are the unit of comparison, the data show that HIMP establishments have both slightly lower visible fecal NR rates and slightly lower rates of visible fecal contamination than non-HIMP establishments.
The comparison included in the comment is based on NR rates from two HIMP establishment and two non-HIMP establishments and does not necessarily reflect the average NR rates for all HIMP establishments.
Comment:
Another consumer advocacy organization stated that it had received records for the first shift of production for 11 young chicken and 3 young turkey HIMP establishments from FSIS through a Freedom of Information (FOIA) request. The organization analyzed documents that covered the period of January 2011 through August 2011. According to the comment, the overwhelming number of NRs filed for the 14 establishments was for visible fecal contamination found on the carcasses. The comment stated that out of 229 NRs filed from March to August 2011, 208 (90 percent) were for visible fecal contamination. Other comments referenced this finding.
Response:
The analysis conducted by the consumer advocacy organization is not inconsistent with the conclusions in the HIMP Report. While it is true that a large percentage of public health-related NRs in poultry slaughter establishments are for visible fecal contamination, the occurrence of fecal contamination on carcasses in HIMP establishments is fewer than 8 per ten thousand carcasses, which is about 19 times lower than the HIMP performance standards. In addition, the rate of visible fecal material contamination on carcasses in HIMP establishments averages about half that in non-HIMP establishments (Table 3-7 in HIMP Report).
5. OCP Standards Under HIMP
In the preamble to the proposed rule, FSIS noted that data from the HIMP Report show that OCP defects identified on carcasses processed in HIMP establishments averaged about half the corresponding OCP HIMP performance standards (77 FR 4418). Based on the HIMP data, the Agency concluded that establishments operating under the HIMP inspection system performed better than establishments operating under non-HIMP inspection systems with respect to OCP defects. Several consumer advocacy organizations and some private citizens commented on this conclusion.
Comment:
Some consumer advocacy organizations asserted that the OCP standards under HIMP were not stringent. The comments said that even with these less than rigorous OCP defect levels, HIMP establishments were still just meeting the standards.
Response:
While there is likely to be some variation in performance among establishments, for the two year period from CY 2009 through 2010, FSIS verification data show that OCP defect levels in HIMP establishments averaged about half the corresponding OCP performance standards.
In addition, the HIMP OCP performance standards are set at the 75th percentile of what was achieved under the RTI's baseline study of the performance of 16 establishments before they entered the HIMP study. Thus, 25 percent of the establishments that entered HIMP have had to improve upon their baseline results to meet the more stringent standards.
Comment:
One comment noted that the HIMP study's statistics on compliance with OCP performance standards are based on a sampling of up to 80 carcasses per slaughter line per shift of production. The comment asserted that when each slaughter line is processing upwards of 100,000 chickens per eight hour shift, this sample size is likely to be too small to accurately reflect the level of OCP defects on RTC carcasses produced by the establishment.
Response:
FSIS disagrees with the comment. FSIS believes that its sampling for OCP defects under HIMP is sufficient to reflect an establishment's level of OCP defects. Statistically, given the sample design, the precision of an estimate of an establishment's level of OCP defects depends primarily on the total number of samples for an establishment collected over time.
The OCP performance standards, which are based on a tightening of the FPS for removable animal diseases and trim and dressing defects for establishments before they entered the HIMP pilot, vary by OCP defect category. For example the performance standard for OCP-1, Condition-Animal Diseases, is 1.7 percent, and the performance standard for OCP-3, Digestive Content (Ingesta), is 18.6 percent. When deciding the number of samples that FSIS should take to reflect an establishment's level of OCP defects over time, FSIS determined that collecting at most 80 birds per line per shift would provide an estimated defect rate that was close to the true defect rate. For example, if the true defect rate for OCP-1 defects was 1 percent at an establishment that operated one line for two shifts, 300 days per year, taking an 80 bird sample per line per shift would give a total of 48,000 samples a year, per line. Eighty samples are not always collected; but in general, close to this number were collected daily. It is reasonable to assume that the total number of samples would not be less than 90 percent, or 43,200 samples. This number of samples, assuming a random distribution of defects throughout the year, would give FSIS an estimated defect rate between 0.905-1.095 percent with about 95 percent probability. Thus, FSIS believes that the specified sample
size is sufficient to make general comparisons of average defect rates among establishments or lines.
6. Salmonella Positive Rates in HIMP Establishments
The HIMP Report compares
Salmonella
positive rates for HIMP young chicken slaughter establishments with a control set of 64 non-HIMP establishments and all 176 non-HIMP broiler establishments (77 FR 4418-4419). The data show that
Salmonella
positive rates are equivalent or lower in HIMP establishments than they are in non-HIMP establishments. The Agency concluded that the increase in offline inspection activities provided for under HIMP resulted in the initial lower levels of
Salmonella
contamination in HIMP establishments. Several consumer advocacy organizations and private citizens commented on the HIMP Report's analysis of
Salmonella
positive rates in HIMP and non-HIMP establishments and on the Agency's conclusions with respect to this analysis.
Comment:
A consumer advocacy organization stated that the HIMP Report's analysis of the
Salmonella
positive rates for HIMP establishments may not reflect the rates for all establishments operating under HIMP. The comment noted that data from the Agency's
Salmonella
testing program show that the Agency collected data on
Salmonella
positive rates from only 14 HIMP establishments in 2006, 17 HIMP establishments in 2007, and 15 HIMP establishments in 2008. The comment noted that the Agency collected
Salmonella
data from only 10 of the 20 HIMP broiler establishments in 2010. The comment also said that the Agency provided no comparison on
Salmonella
results in the turkey establishments. One member of academia said that the Agency's microbial sampling and analysis under the HIMP pilot were not performed with adequate frequency or power to detect sporadic low-level contamination of carcasses.
Response:
FSIS uses the same methodology to schedule and conduct verification sampling for
Salmonella
in both HIMP and non-HIMP establishments. Under the FSIS risk-based methodology for scheduling
Salmonella
verification sample sets, not all establishments are sampled every year. FSIS schedules up to 75 new sample sets each month. The establishments and products selected for sample sets are chosen according to a risk-based algorithm that involves sorting the list of eligible establishments and their respective products by certain criteria and selecting the top 75 from this list. Depending on the frequency of production, product type, and availability of resources, the time to complete a sample set ranges from less than two months to over a year. In establishments that produce more than one product subject to
Salmonella
verification testing, only one product is tested at a time. However, since the same method is used in both HIMP and non-HIMP establishments,
Salmonella
positive levels represent a valid means of comparing the performance of HIMP and non-HIMP establishments.
With respect to the comment that said that the Agency's microbial sampling and analysis under the HIMP pilot were not performed with adequate frequency or power to detect sporadic low-level contamination of carcasses, the sampling and analysis for
Salmonella
under the HIMP pilot was used to compare performance of both HIMP and non-HIMP establishments, not to detect sporadic, low-levels of contamination in HIMP establishments.
Comment:
One consumer advocacy organization said that the Agency's conclusion that HIMP establishments have lower
Salmonella
positive rates than non-HIMP establishments is misleading because the HIMP Report compared
Salmonella
positive rates for HIMP establishments with all establishments operating under non-HIMP inspection systems. According to the comment, the Agency should have compared rates for HIMP establishments with the rates for comparably sized non-HIMP establishments.
Response:
The HIMP Report compared
Salmonella
positive rates in HIMP establishments with both comparable non-HIMP establishments and all young chicken slaughter establishments. The first comparison set of establishments was a subset of 64 non-HIMP establishments selected to be comparable to HIMP establishments with respect to total slaughter volume, line speeds, and geographic distribution. The second comparison set was all 176 non-HIMP establishments that slaughtered young chicken in all 5 years considered in the study. The analysis shows that with respect to
Salmonella
positive rates, the HIMP establishments performed better than or as well as both the comparison set of 64 non-HIMP establishments and the set of all 176 non-HIMP establishments from CY 2006 through 2010.
Comment:
A consumer advocacy organization asserted that the
Salmonella
positive rates in HIMP establishments do not support the Agency's claim that HIMP establishments have consistently performed better under HIMP than they did under non-HIMP inspection systems. The comment stated that the Agency's own
Salmonella
data from 1998-2007 demonstrate that 14 of the 20 HIMP establishments had lower
Salmonella
positive rates under the non-HIMP inspection systems than they did under the HIMP, and that the average
Salmonella
positive rate for all 20 of the HIMP establishments was better when the establishments were operating under non-HIMP inspection systems. The organization conducted its own analysis of the Agency's
Salmonella
data from January 1, 2006 through September 20, 2007 and said that its analysis shows that the HIMP establishments had an average
Salmonella
positive rate of 8.9 percent, while the non-HIMP establishments had an average rate of 6.5 percent.
Response:
In CY 2006 through 2008, the
Salmonella
positive rate in HIMP establishments was statistically significantly lower than in the 64 non-HIMP comparison set, and there was no statistically significant difference in CY 2009 and CY 2010. A comparison of HIMP establishments with all non-HIMP broiler establishments shows that the
Salmonella
positive rate in HIMP establishments was statistically significantly lower in CY 2006 through 2009 and not statistically significantly different in CY 2010. This analysis demonstrates that with respect to
Salmonella
positives rates, HIMP establishments are performing at least as well as current non-HIMP establishments.
With respect to
Salmonella
data from January 1, 2006, through September 20, 2007, referenced by the comment, FSIS has analyzed the most recent data from that time period and found
Salmonella
positive rates of 7.55 percent and 9.61 percent for HIMP and non-HIMP establishments, respectively.
Comment:
A consumer advocacy organization stated that in CY 2009 and CY 2010, HIMP establishments had higher
Salmonella
positive rates than the 64 non-HIMP comparison establishments. The comment noted that the HIMP Report shows that the rates for the HIMP establishment were 4.9 percent and 4.7 percent in CY 2009 and CY 2010, respectively, and the rates for the non-HIMP establishments for these years were 4.3 percent and 4.0 percent, respectively. The comment suggested that before moving forward with the NPIS, FSIS should first try to understand why this happened.
Response:
FSIS disagrees with the comment's suggestion that HIMP establishments had higher
Salmonella
rates than non-HIMP establishment in CY 2009 and CY 2010. The differences
in
Salmonella
positive rates in CY 2009 and CY 2010 noted by the comment are not statistically significant. On the other hand, in CY 2006 through 2008, the
Salmonella
positive rate in HIMP establishments was statistically significantly lower than in the 64 non-HIMP comparison set.
Comment:
Some consumer advocacy organizations stated that reductions in
Salmonella
positive rates may be the result of factors other than increased offline inspection procedures performed under the HIMP inspection system. The comments noted that from CY 2006 through 2008,
Salmonella
positive carcass rates in HIMP establishments were statistically significantly lower than in the non-HIMP comparison establishments, but that in CY 2009 and CY 2010, there was no statistically significant difference. The comments also noted that both HIMP and non-HIMP establishments lowered their
Salmonella
positive rates considerably between CY 2006 and CY 2010.
The comments asserted that because the Agency did not report any changes to the HIMP or non-HIMP inspection systems during that time, it is reasonable to assume that factors other than increased offline inspection activities in HIMP establishments may have caused such a significant decrease in
Salmonella
positive rates. One comment noted that in 2008 FSIS began publishing the names of establishments in Categories 2 and 3 under the Agency's new
Salmonella
performance standards. The comment stated that the data for CY 2009 and CY 2010 may indicate that the industry as a whole reduced its
Salmonella
positive rates as a result of this initiative. Another comment stated that the decline in
Salmonella
positive rates may have been caused by an increase in the use of online reprocessing technology throughout the industry.
Response:
In the preamble to the proposed rule, the Agency explained that results in CY 2010 most likely reflect the effects of the
Salmonella
initiatives that FSIS began implementing in 2006 to reverse the multi-year trend of persistently higher percent positive rates for
Salmonella
detected through the Agency's HACCP verification testing each year (77 FR 4419). As a result of these initiatives, the industry reduced the incidence of positive
Salmonella
results, particularly those establishments with the highest
Salmonella
positive rates. Nonetheless, before these initiatives were fully implemented, the HIMP report shows that HIMP establishments performed better than non-HIMP establishments with respect to
Salmonella
positive rates. The reduction in
Salmonella
positive rates in both HIMP and non-HIMP establishments reflects the effectiveness of FSIS's initiatives to reduce
Salmonella
industry-wide.
Comment:
One member of academia said that the Agency needs to conduct more frequent sampling for a broader range of pathogens to assess the impact of the HIMP inspection system.
Response:
Salmonella is a key pathogen of concern in poultry products. FSIS conducts
Salmonella
verification sampling in both HIMP and non-HIMP establishments. Thus,
Salmonella
positive rates are a valid means of comparing the performance of both HIMP and non-HIMP establishments.
C. The Risk Assessment
The preamble to the proposed rule explained that in June 2011, FSIS completed a quantitative risk assessment to model how performing a greater number of sanitation, sampling, and other offline inspection procedures in young chicken and turkey slaughter establishments might affect the number of human illnesses from
Salmonella
and
Campylobacter
(77 FR 4420). FSIS updated the 2011 Risk Assessment in response to public comments received on the January 2012 proposed rule; that version of the risk assessment was subsequently posted to the FSIS Web site in August 2012 (referred to as the August 2012 version). In addition, the 2011 risk assessment was subjected to independent external peer review; the risk assessment was further updated in response to the peer review comments. It has also benefited from editing consistent with the Office and Management memorandum,
Final Guidance on Implementing the Plain Writing Act of 2010
(M-11-15), striving to make the risk assessment report language “clear, concise, well-organized. The most recent version of the risk assessment, which reflects the revisions made in response to public and peer review comments, is referred to as the July 2014 version. Both the August 2012 version and the July 2014 versions have been posted to the FSIS Risk Assessment Web page at:
http://www.fsis.usda.gov/wps/portal/fsis/topics/science/risk-assessments
.
The HIMP Report explained that FSIS inspectors performed more offline inspections to verify compliance with sanitation SOP and HACCP regulations in HIMP establishment than they do in non-HIMP establishments. The regression analysis of historical data that was included in the risk assessment showed a statistically significant correlation between unscheduled offline inspection procedures and reduction in the prevalence of
Salmonella
and
Campylobacter
positive samples. Based on these results, FSIS thinks it is reasonable to conclude that the redeployment of Agency resources to unscheduled offline activities is likely to contribute to improved food safety resulting from a lower prevalence of carcasses contaminated with
Salmonella
and
Campylobacter,
which in turn we expect to lead to fewer human illnesses.
Comment:
Several comments requested that the Agency clarify the status of the 2011 risk assessment's peer review. The comments noted that the Agency had prepared a risk assessment in 2005 that was peer reviewed. The comments said as a result of the peer review, the Agency prepared a revised risk assessment in 2008 but, according to the comments, the docket for the proposed rule contains neither the 2008 risk assessment nor a peer review of that risk assessment.
Response:
The FSIS “Risk Assessment for Guiding Public Health-Based Poultry Slaughter Inspection” has been available to the public on the FSIS Risk Assessment Web site since 2008 at:
http://www.fsis.usda.gov/wps/wcm/connect/07c57a64-932f-4ebb-977b-2b10e45a1830/Poultry_Slaughter_Risk_Assess_Jan2008.pdf?MOD=AJPERES
. The analysis was originally peer reviewed in 2006 by an independent group of mathematical modeling specialists. The risk assessment was modified and improved based on the initial peer review. Because the model and analysis has continued to evolve, the 2011 version of both the model and analysis have undergone a peer review. The 2011 risk assessment has been updated based on the peer review comments. The 2011 risk assessment, the peer review comments, FSIS's response to those comments, and the current version of the risk assessment are available on the FSIS Web site at:
http://www.fsis.usda.gov/wps/wcm/connect/8f374626-ee06-49d3-9d41-6eb65ad32cbb/Poultry_Slaughter_Risk_Assess_Aug2012.pdf?MOD=AJPERES
.
Comment:
A consumer advocacy organization said that the risk assessment provides little raw data, little explanation of how it was analyzed, and is largely silent on the assumptions upon which it was based. A comment from a labor union was also critical of the FSIS risk assessment.
Response:
FSIS generally disagrees with the comments. The risk assessment uses all relevant data taken from FSIS's inspection database paired with
Salmonella
and
Campylobacter
regulatory and baseline sampling data for young chickens and turkeys. Overall,
substantial amounts of empirical data were used in this risk assessment. It uses the Young Chicken Baseline and PR/HACCP
Salmonella
verification data from July 2007-September 2010 and the Young Chicken Baseline
Campylobacter
data from July 2008-September 2009. It also uses the Young Turkey Baseline and PR/HACCP
Salmonella
verification data from July 2007-September 2010 and the Young Turkey Baseline
Campylobacter
data from August 2008-July 2009. There are about 40,900 raw data samples collected on 94 inspection procedures taken from the computerized Performance-Based Inspection System (PBIS).
Although FSIS thinks that the 2011 version of the risk assessment is fully documented, the July 2014 version has benefited from the addition of language that more clearly describes how the model works and articulates more it clearly the underlying assumptions. As noted above, this version also was updated in response to peer review comments. As discussed above, the 2011 version of the risk assessment, the peer review comments on that version, FSIS's response to the peer review comments, and the updated 2014 version of the risk assessment are posted on the FSIS Web site.
Comment:
A commenter said that one of the major assumptions in the risk assessment is that if performing more unscheduled offline inspection procedures “either reduces (or does not change) the occurrence of foodborne pathogens such as
Salmonella
and
Campylobacter
on finished poultry products, then a net public health benefit may result.” The comment questioned how there could be a “net public health benefit” if there is no change to the incidence of pathogens on poultry carcasses. The comment said that FSIS should not predicate a significant restructuring of the poultry slaughter inspection program based on a finding that there will be no change to the incidence of contamination of poultry products. According to the comment, any substantial change to meat or poultry inspection should result in significant improvements to public health.
Response:
The Agency agrees that the statement in the risk assessment may not fully articulate how a net public health benefit may result if performing more unscheduled offline inspection procedures reduces (or does not change) the occurrence of foodborne pathogens. To clarify, the risk assessment estimates that if more unscheduled offline inspection procedures reduces the occurrence of a specific foodborne pathogen, such as
Salmonella,
but does not change the occurrence of a different pathogen, such as
Campylobacter,
there will be an overall reduction in pathogens on finished poultry products. This aggregate reduction of pathogens and the subsequent reduction in human illnesses is what was hypothesized to result in a net public health benefit.
The risk assessment characterizes a negative correlation between the frequency of unscheduled offline inspection activities and the prevalence of both
Salmonella
and
Campylobacter
positive samples. Based on these modeling results, FSIS thinks it is reasonable to conclude that redeployment of Agency resources from online inspection activities to targeted unscheduled offline activities is likely to produce an improvement in the food safety system resulting from a lower prevalence of carcasses contaminated with
Salmonella
and
Campylobacter,
which could in turn result in a net reduction in the number of human illnesses.
Comment:
Several comments noted that the Agency conceded that “substantial uncertainty about forecasted changes in illness rates” results from uncertainty about the change in future inspection activities and the rates of human illnesses attributable to poultry.
Response:
The risk assessment analyzed data on specific types of inspection activities and the prevalence of
Salmonella
and
Campylobacter
in young chicken and turkey slaughter establishments. The results suggest that, because inspection personnel assigned to the NPIS will conduct more of the type of inspection activities that were correlated with lower
Salmonella
and
Campylobacter
prevalence, the NPIS will likely result in fewer human illnesses than would be expected if not implemented. In addition to the expected values, the analysis provides the statistical uncertainty of the estimated number of averted illnesses by reporting the upper and lower 80 percent confidence bounds around the estimates to acknowledge that uncertainty always will exist in such models.
Comment:
Several consumer advocacy organizations noted that the 2011 version of the risk assessment predicts that additional unscheduled offline procedures could lead to as many as 986 fewer
Campylobacter
-related illnesses per year. The comment noted that the risk assessment states that “this analysis suggests ambiguous effects of the proposed rule with respect to
Campylobacter
occurrence on chicken carcasses” and thus does not show a clear public health benefit.
Some comments noted that the Agency recently established a performance standard for
Campylobacter.
The comment said that the Agency does not have enough experience with the
Campylobacter
performance standards to assess industry efforts to reduce
Campylobacter
in poultry to make any reasonable predicted public health benefits. The comments said that if the Agency's proposed changes to poultry slaughter inspection are truly intended to improve public health, the Agency needs a much better understanding of
Campylobacter
rates in poultry establishments and of how the Agency's proposal will impact those rates.
One comment added that the risk assessment suggests that “the positive
Salmonella
implications of HIMP” could be applied to
Campylobacter,
but the Agency provides no justification for this statement. The comment said that several studies point to the difficulty of making correlations between controlling for
Salmonella
and controlling for
Campylobacter.
The comments asserted that FSIS should postpone implementation of the proposed rule until it has collected additional data on
Campylobacter
and is better able to estimate the impacts of the proposed rule on reducing this pathogen.
Response:
The Risk Assessment presented the results of two scenarios—one that was based on only increasing unscheduled offline procedures (referred to as the “discriminate scenario”) and one that did not specify the particular activities to be increased (referred to as the “indiscriminate scenario”). The former (discriminate scenario), which was based on the type inspection procedures performed more often in the HIMP establishments, suggested larger improvements to public health than the indiscriminate model. FSIS peer-reviewed risk assessment (July 2014), results suggest that the
discriminate scenario
of increased off-line inspection could decrease the number of positive
Salmonella
and
Campylobacter
samples in young chicken and young turkey establishments with high probability. This is the scenario upon which this rule is based.
As noted by the comments, the Agency recently established performance standards for
Campylobacter
for young chicken and turkey slaughter establishments. Because the Agency has not been collecting and analyzing samples for
Campylobacter
as long as it has been collecting and analyzing samples for
Salmonella,
there are fewer
Campylobacter
sampling results available for analysis. Thus, although the trends for the
Salmonella
and
Campylobacter
results are the same, the
Campylobacter
results are less robust because of the smaller sample size. The updated risk assessment estimates that there would be a reduction of 3,980
Salmonella
illnesses attributable to young chicken and turkey establishments combined. This in itself would be a positive public health outcome. Because an increase in unscheduled offline inspection activities is expected to result in fewer
Salmonella
illnesses, FSIS believes that there is no reason to delay implementation of the rule until the Agency collects and analyzes more samples for
Campylobacter.
Additionally, Agency responses to
Campylobacter
sample set failures will continue to follow procedures for Salmonella set failures, i.e. immediate follow-up testing for both organisms and, in most instances, Food Safety Assessments, regardless of whether an establishment adopts the NPIS or not.
Comment:
A consumer advocacy organization said that the risk assessment relies heavily on the data collected through the HIMP pilot and the microbiological verification testing programs. The comment asserted that, because these programs are not representative of all poultry establishments, data collected through these programs cannot be generalized to the entire poultry industry. The comment said that the microbiological verification testing programs were not designed to estimate the incidence of foodborne pathogens in meat and poultry products, nor were they designed to evaluate trends over time. The comment said that despite these limitations, the risk assessment has used these data to evaluate the public health impact of reassigning online inspectors to offline activities and has concluded that there is a public health benefit to doing so. The comment suggested that FSIS conduct a pilot study in a representative sample of poultry establishments to ensure that there is a public health benefit before implementing the proposed rule in all poultry establishments.
Response:
The assertion that the risk assessment relies on data that are not representative of all poultry establishments is not accurate. The risk assessment uses a volume-weighted model to account for the fact that the microbiological sampling is not proportional to volume. The risk assessment relies on
Salmonella
data collected from 189 young chicken and 25 turkey slaughter establishments and on
Campylobacter
data collected from 181 young chicken and 65 young turkey slaughter establishments from July 2007 to September 2010. There are 20 young chicken establishments and 5 turkey establishments operating under the HIMP inspection system.
The risk assessment does not use the results of microbiological verification testing programs to estimate the prevalence of foodborne pathogens in poultry products or to evaluate trends over time, as suggested by the comment. The risk assessment uses FSIS microbiological verification testing results to analyze correlations between observed positive samples and offline inspection activities in young chicken and turkey slaughter establishments. These correlations are then used as one input to the model that characterizes changes in attributable human illness. The risk assessment showed that the greatest effect on
Salmonella
and
Campylobacter
prevalence and related illness would occur when inspection activities are concentrated on increased unscheduled offline procedures. Thus, FSIS disagrees with the comment's suggestion that the Agency should not implement the proposed rule until it conducts a pilot study in a representative sample of poultry establishments to ensure that there is a public health benefit. The Agency has ample evidence to support its conclusions that there is a solid basis to allow for the NPIS.
Comment:
Comments from a consumer advocacy group and a labor union said that the risk assessment is based on the assumption that the Agency's
Salmonella
verification data accurately reflect the performance of the establishments. The comments questioned whether the Agency's
Salmonella
verification results reflect the typical operating conditions in establishments. According to the comments, establishments know when FSIS is about to collect
Salmonella
verification samples because the test kit is mailed to the establishment right before the inspectors are to collect the samples. According to the comments, on days when inspectors collect samples for
Salmonella
testing, it is not unusual for the establishments to increase the concentration of available chlorine in the chiller. The comments asserted that the results of the risk assessment are not reliable because the predictions are not based on typical operating conditions in establishments. As a result, the comments said that FSIS's claims that the proposed rule may reduce the number of
Salmonella
and
Campylobacter
related illnesses are highly speculative and unlikely to be realized.
Response:
FSIS disagrees with the comments. The available data from FSIS's microbiological baseline studies and the Agency's
Salmonella
verification results indicate that FSIS's
Salmonella
verification sampling results do reflect typical operating conditions in the establishment.
The Agency compared its most recent baseline data for
Salmonella
prevalence in young chicken and turkey slaughter establishments collected under its National Microbiological Data Collection Programs completed in 2008 and 2009, respectively, with the results of the
Salmonella
samples that it collected and analyzed under its HACCP
Salmonella
verification program for similar time periods. The estimated
Salmonella
prevalence associated with the two sets of data, when volume weighted and adjusted for other establishment characteristics, were not significantly different. FSIS has documented this conclusion in a series of Agency reports
5
and written material associated with the
Federal Register
notice, “New Performance Standards for
Salmonella
and
Campylobacter
in Young Chickens and Turkey Slaughter Establishments, Response to Comments and Announcement of Implementation Schedule,” which announced the new
Salmonella
and
Campylobacter
performance standards (76 FR 15282).
5
http://www.fsis.usda.gov/wps/portal/fsis/topics/regulations/federal-register/federal-register-notices/notices-2011
;
Salmonella
and
Campylobacter
Notice and comparisons of HACCP and baseline report.
In addition, under both HIMP and non-HIMP inspection systems, the protocol is for inspectors to randomly collect scheduled
Salmonella
verification samples and do not inform establishments when they collect the samples. FSIS uses the best available data and has taken steps to enhance data quality going forward. For example, FSIS authorizes its inspectors to request that the Agency schedule additional
Salmonella
verification sampling if they have evidence to demonstrate that an establishment altered its food safety system to coincide with the FSIS
Salmonella
verification sample set.
6
Since FSIS implemented this policy, there have been 10 requests, from which 3 were found to be process changes during
Salmonella
sampling that justified an additional verification set. As of July 21, 2014, there have been no requests since December 2013. Thus,
FSIS has no basis to think that establishments are regularly making changes to their processes that would substantially affect the Agency's
Salmonella
verification results or, in turn, affect the conclusions of the risk assessment or the HIMP report.
6
See FSIS Notice 66-12, which reissued the policy in former FSIS Notice 42-11 at:
http://www.fsis.usda.gov/OPPDE/rdad/FSISNotices/66-12.pdf.
Comment:
A consumer advocacy organization said that the risk assessment's assumptions regarding unscheduled inspection procedures were based on procedures assigned under the PBIS. The comment said that now that FSIS has implemented the Public Health Inspection System (PHIS), the number of pre-operational sanitation procedures that inspectors conduct on a monthly basis was reduced to accommodate other inspection procedures under PHIS. According to the comment, the risk assessment is flawed in that it is not based on the inspection tasks that FSIS inspectors will actually be performing under PHIS.
Response:
The risk assessment is based on the data that were available at the time that FSIS conducted the analysis. At that time, the available data on offline inspection procedures reflected the number of such procedures scheduled under PBIS. The analysis of historical data that is presented in the risk assessment showed a relationship between lower
Salmonella
in young chicken and
Campylobacter
in turkey prevalence and the type of inspection activities that will be conducted more frequently under the NPIS. FSIS inspectors will continue to conduct both unscheduled and scheduled offline inspection activities under PHIS. Thus, the Agency thinks that the risk assessment's results are valid under PHIS.
Comment:
Two consumer advocacy organizations said that while the risk assessment details the uncertainty about the change in human illness rates when offline inspection activities are intensified, there is no comparable examination of the human illness changes from reducing online Federal inspection activities. One of the comments asserted that the risk assessment also did not fully consider the other changes to the inspection system that the Agency was proposing. This comment specifically noted that the risk assessment did not consider the increase in line speeds that had been proposed under the NPIS. Both comments asserted that the Agency should withdraw the rule until an analysis of all of the modifications and variables provides certainty that the inspection changes will not increase the risk to human health.
Response:
FSIS disagrees with the suggestion that the Agency withdraw the rule until it conducts an additional analysis. The modifications noted by the comments were addressed in the HIMP pilot study. FSIS thinks that the performance of establishments under HIMP, as documented in the HIMP report, represents what would be achieved under the NPIS. These results support moving forward with this final rule.
As under HIMP, under the NPIS, establishment employees will be responsible for conducting online sorting activities that are currently conducted by FSIS online inspectors. Based on the results of the HIMP pilot, FSIS thinks that establishment employees can perform these activities as effectively as FSIS inspectors do. To ensure that they do, FSIS inspectors in establishments operating under the NPIS will verify that establishment employees are effectively sorting carcasses on an ongoing basis. As they do under HIMP, VIs under the NPIS will collect samples and conduct verification checks and CIs will perform a visual inspection of each carcass at the end of the line before the chiller. If inspection personnel find food safety-related defects or the presence of persistent, unattended trim and dressing defects or removable animal diseases on carcasses and parts, FSIS will require that the establishment take appropriate action to ensure that establishment employees are effectively sorting carcasses and that the establishment is operating under conditions needed to produce safe, wholesome, and unadulterated product. Therefore, the Agency believes that establishment employees operating under FSIS inspection can effectively perform the sorting activities that they will be responsible for under the NPIS.
FSIS also disagrees with the comment that suggested that the Agency conduct an additional risk assessment to estimate the effects of line speeds on food safety and public health. The focus of the risk assessment is to determine how performing a greater number of sanitation, sampling, and other offline activities in young chicken and turkey slaughter establishments might affect the number of human illnesses from
Salmonella
and
Campylobacter.
Although the regression analysis used in the risk assessment did include a categorical variable representing line speed as a structural (fixed effect) variable in the regression model that predicts prevalence, the results do not reflect measures that establishments typically implement in response to a given line speed in order to maintain process control.
7
The Agency believes that the performance of establishments under HIMP, as documented in the HIMP report, represent what would be achieved under the NPIS at similar line speeds.
7
See Appendix Tables 6-9 in the July 2014 Risk Assessment.
Comment:
One comment said that the risk assessment concludes that more unscheduled offline procedures are the key to lowering
Salmonella
levels. The comment noted that the risk assessment did not consider whether this would be the case if inspectors also did not perform all of the scheduled food safety verifications, which were the only inspection tasks that inspectors performed more in HIMP establishments than in non-HIMP establishments. According to the comment, this is important because there are no scheduled offline food safety checks in the NPIS. The comment questioned the Agency's “assum[ption] that offline inspection activities after the voluntary implementation of the new inspection system will parallel offline inspection activities in current HIMP establishments.”
Response:
Inspection procedures that will be performed in establishments operating under the NPIS will be determined by protocols currently required under PHIS. Under PHIS, inspectors perform both routine (scheduled) procedures and directed (unscheduled) procedures. Thus, inspectors assigned to establishments operating under the NPIS will perform both scheduled and unscheduled offline procedures, just as they currently do in both HIMP and non-HIMP establishments. These offline procedures include, verifying compliance with HACCP and Sanitation SOP requirements, performing carcass verification checks for septicemia/toxemia and visible fecal contamination, verifying sanitary dressing requirements, and collecting samples. The offline inspection activities conducted under the NPIS are intended to be the same rather than parallel the procedures of the existing inspection systems, yielding the same or better public health outcomes.
D. The New Poultry Inspection System (NPIS)
In the preamble to the proposed rule, the Agency explained that, based on its experience under HIMP, it was proposing to establish the NPIS for young chickens and turkeys (77 FR 4421). The proposed rule would have eliminated SIS, NELS, NTIS, and the HIMP pilot and would have required
that all young chicken and turkey slaughter establishments operate either under the NPIS or the Traditional Inspection System, as modified in the proposed rule. The proposed rule would also have limited the number of online inspectors under Traditional Inspection to two for each evisceration line, with an exception for existing establishments that slaughter poultry other than young chickens and turkeys that are currently operating with more than two online inspectors.
As discussed below, after considering the comments, FSIS has decided to modify the proposed rule to leave in place all of the existing poultry inspection systems. FSIS has also decided to allow the 20 young chicken establishments that have been granted SIP waivers to operate under HIMP to continue to operate under a SIP waiver to run at line speeds of up to 175 bpm. However, FSIS will update these SIP waivers to remove aspects of HIMP that are inconsistent with the NPIS, such as the OCP performance standards. If an establishment operating under a SIP waiver described above goes out of business or decides to give up its waiver, FSIS will select another establishment to take its place. Thus, as under the current HIMP protocol, FSIS will continue to provide SIP waivers for up to 20 young chicken establishments to operate at 175 bpm. Under this final rule, the maximum line speed under the NPIS for turkeys will be 55 bpm, as was proposed. Thus, there is no need for the five HIMP turkey establishments to continue to operate under an updated SIP waiver because they will be able to achieve the same results by operating under the NPIS. FSIS has also decided that it will not limit the number of online inspectors under Traditional Inspection to two. Under this final rule, FSIS will continue to staff all establishments that do not choose to operate under the NPIS with the number of online inspectors currently assigned to the establishment.
The preamble also explained that FSIS would allow establishments that slaughter classes of poultry other than young chickens and turkeys to operate under the NPIS under a waiver through the Salmonella Initiative Program (SIP). Under the SIP, meat and poultry slaughter establishments receive waivers of regulatory requirements on condition that they will conduct regular microbial testing and share the resulting data with FSIS.
1. General Comments on the NPIS
Comment:
Comments from producers of poultry products and trade associations representing the poultry industry expressed general support for the NPIS. Comments from some FSIS inspection personnel and some private citizens also expressed support for the NPIS. Some comments noted that the existing inspection systems were designed before FSIS implemented HACCP and were developed to identify visual defects that affect the quality of the product. The comments agreed that Agency resources are better spent performing activities that are more effective in ensuring food safety rather than performing functions that establishments can effectively accomplish under FSIS inspection by both VIs and CIs. Another comment said that the NPIS will give establishments the flexibility to investigate and develop new and more efficient technologies. The comment agreed with the Agency's conclusion that the new inspection system will improve the effectiveness of poultry slaughter inspection and overall food safety, remove unnecessary regulatory obstacles to innovation, and make better use of the Agency's resources. Another comment said that the NPIS is the next logical step in protecting public health through modern, science-based food safety technology.
One comment that supported the NPIS proposal stated that it should be considered as part of a comprehensive food safety program that includes the recently implemented PHIS and performance standards for
Salmonella
and
Campylobacter
in broilers and turkeys. The comment said that the proposal should not be considered separate and apart from other regulatory food safety programs.
Response:
The Agency agrees with the thrust of these comments and concurs. Certainly the NPIS is part of the initiatives that contribute to the Agency's comprehensive food safety program. As noted by one comment, among these initiatives are the
Salmonella
and
Campylobacter
performance standards, the SIP, PHIS, as well as the NPIS.
Comment:
Comments from consumer advocacy organizations, labor unions, FSIS inspectors, public health organizations, animal welfare advocacy organizations, members of academia, human and worker rights advocacy organizations, and some private citizens objected to the NPIS for various reasons. Many of these comments objected to the NPIS because the commenters view the NPIS as a system that “privatizes” inspection by replacing USDA online inspectors in part with establishment employees. The petitions submitted in response to the proposed rule express these same views.
Response:
The NPIS will not privatize poultry inspection; this system makes Federal inspection of poultry more effective and carcass inspection by FSIS inspectors more efficient.
Under the existing poultry slaughter inspection systems, FSIS inspectors check each carcass for defects and disease and direct establishment employees to take corrective actions. Under the NPIS, a well-trained FSIS CI will conduct a carcass-by-carcass inspection after establishment employees have sorted, trimmed, and conducted any necessary reprocessing. Thus, under the NPIS the CI will be able to conduct a more effective and efficient carcass-by-carcass inspection because carcasses will only be presented for inspection by the CI if they have been sorted by the establishment and are likely to pass inspection.
As discussed earlier in this document, the VIs under the NPIS will conduct offline food safety-related inspection activities and will monitor and evaluate establishment process controls. The VIs will conduct carcass verification checks on carcass samples collected before the CI station to ensure that the establishment is effectively sorting carcasses and that it is producing products that comply with the Agency's zero visible fecal tolerance and other performance standards. VIs will also perform offline activities in addition to carcass verification checks, such as verifying compliance with sanitation SOPs, SPS, and HACCP regulatory requirements, and ensuring that the establishment is meeting all regulatory requirements and is effectively preventing contamination by enteric pathogens and fecal material throughout the entire slaughter and dressing process.
2. Scope of the NPIS
Comment:
One comment said that it interprets the proposed rule to limit establishments that slaughter mature fowl to operate under the NPIS only if they participate in the SIP. The comment noted that the only other alternative for establishments that slaughter mature fowl would be to operate under Traditional Inspection. The comment stated that FSIS should expand the scope of the NPIS to include classes of poultry other than young chickens and turkeys without additional qualifications. According to the comment, requiring that establishments that slaughter poultry classes other than young chickens and turkeys operate under a SIP waiver places them at a competitive disadvantage because they must incur costs associated with the
additional testing and data collection required under the SIP.
Response:
The NPIS was informed by the Agency's experience under the HIMP pilot, which, for poultry, was limited to young chicken and turkey slaughter establishments. Thus, the Agency would need additional data to support an expansion of the NPIS to classes of poultry other than young chickens and turkeys. As noted by the comment, FSIS would permit establishments that slaughter classes of poultry other than young chickens and turkeys to operate under the NPIS under a waiver through the SIP. At a later time, the Agency would consider the data collected in such poultry slaughter establishments operating under a SIP waiver to determine whether to expand the NPIS to other classes of poultry.
Comment:
Comments from two labor unions and a worker rights advocacy organization stated that although the proposed rule allows young chicken and turkey slaughter establishments to choose whether they will operate under the NPIS or under Traditional Inspection, there is no real choice because the Agency proposed to limit the number of online inspectors in establishments operating under Traditional Inspection to two. The comments noted that because most of the establishments that slaughter young chickens and turkeys are large automated operations, it is unlikely that these establishments will choose the traditional method of inspection with slower line speed and two inspectors per line. A comment from an individual questioned why establishments cannot choose to continue to operate under their current inspection systems. The comment stated that FSIS did not require that establishments operate under SIS, NELS, or NTIS when the Agency established those inspection systems. The comment said that allowing establishments to choose to keep their current inspection system gives them a true choice and maintains competition in the marketplace.
A comment from a member of academia said that the proposed rule gives establishments that slaughter young chickens and turkeys the flexibility to decide whether the benefits of switching to the NPIS exceed their estimated costs to operate under such a system. The comment said that many very small establishments are likely to choose to remain under Traditional Inspection because, unlike larger establishments, the benefits of operating under the NPIS may not exceed their costs.
Response:
After careful consideration of these comments, FSIS has decided to revise the proposed rule to allow establishments that do not choose to operate under the NPIS to continue to operate under their current inspection system, i.e., SIS, NELS, NTIS, or Traditional Inspection. FSIS has also decided to allow the 20 young chicken establishments that have been granted SIP waivers to operate under HIMP to continue to operate under a SIP waiver to run at line speeds of up to 175 bpm. However, FSIS will update these SIP waivers to remove aspects of HIMP that are inconsistent with the NPIS, such as the OCP performance standards. If an establishment operating under a SIP waiver described above goes out of business or decides to give up its waiver, FSIS will select another establishment to take its place. Thus, as under the current HIMP protocol, FSIS will continue to provide SIP waivers for up to 20 young chicken establishments to operate at 175 bpm. Under this final rule, the maximum line speed under the NPIS for turkeys will be 55 bpm, as was proposed. Thus, there is no need for the five HIMP turkey establishments to continue to operate under an updated SIP waiver.
This final rule will give establishments the flexibility to operate under the system that is best suited to their operations. FSIS has also determined that allowing establishments to continue to operate under their current inspection system instead of converting to the modified Traditional Inspection with two online inspectors will create less disruption to the industry when FSIS begins to implement the NPIS. As noted by the comments, large establishments will likely choose to operate under the NPIS, while very small establishments are likely to choose to operate under the modified Traditional Inspection System. Some establishments may be interested in operating under the NPIS but are not prepared to make the capital investments needed to convert right away. Under this final rule, these establishments will have the option to switch to the NPIS at a later date without having to convert to a modified Traditional Inspection first.
3. Carcass Sorting and Inspection Under the NPIS
a. Carcass Sorting by Establishment Employees
Comment:
Several comments from consumer advocacy organizations, FSIS inspectors, labor unions, and private citizens objected to the NPIS's requirement that establishment employees properly sort carcasses before they are presented to the CI for inspection because the comments believe that establishment employees will miss many food safety and OCP defects. Many of the comments referenced the analysis conducted by the consumer advocacy organization that obtained FSIS inspection records from 14 establishments participating in the HIMP pilot presented in an earlier comment. According to the comments, the analysis shows that establishment employees missed food safety and wholesomeness defects at high rates.
Another comment stated that it had secured affidavits from three USDA inspectors who have worked in HIMP establishments who report that because of excessive line speeds and lack of training, company employees routinely miss many food safety and wholesomeness defects. The comments stated that FSIS must more thoroughly evaluate the proposal to allow establishment employees to perform preliminary carcass sorting before it implements the NPIS.
Response:
The overall performance of HIMP establishments measured by the findings of offline inspections by VIs was as good as or better than non-HIMP establishments. Results from the VI inspections in HIMP establishments, which are conducted after establishment employees have completed the initial carcass sorting, show that the rates of carcasses with septicemia/toxemia and visible fecal contamination in HIMP establishments were very low, well below the levels set by the HIMP performance standards. These results were discussed in detail above. In addition, as discussed above, OCP defect rates identified on carcasses in HIMP establishments average about half the corresponding OCP HIMP performance standard. Thus, the data from the HIMP pilot show that establishment employees do effectively sort carcasses, dispose of carcasses that must be condemned, and conduct necessary trimming and re-processing activities before the carcasses are presented to the CI for online carcass inspection.
Comment:
Several comments from consumer advocacy organizations and private citizens noted that the NPIS does not require that establishment employees performing the sorting function receive training or prove proficiency in performing their duties. The comments noted that the 2001 GAO report on the HIMP pilot program criticized FSIS for not requiring that establishment employees complete training before assuming carcass sorting activities. The comments said that FSIS should accept the GAO recommendation for FSIS to develop a
training and certification program in conjunction with industry.
Response:
FSIS is not prescribing specific sorter training or certification. However, the Agency has developed guidance documents to assist establishments in training their sorters. This guidance is available on the FSIS Web site at:
http://www.fsis.usda.gov/wps/portal/fsis/topics/regulatory-compliance/compliance-guides-index.
The guidance that the Agency has developed is based on the training that FSIS provides to online inspection personnel that are responsible for sorting carcasses under the existing inspection systems.
FSIS agrees with the comment that training of sorters is important to ensure that they are able to properly perform their duties. Proper training is necessary if sorters are to make accurate decisions on how to address animal disease conditions and trim and dressing defects. Under the NPIS, if sorters do not make these decisions correctly, FSIS inspection personnel will take appropriate action such as stopping the production line, issuing NRs, and directing the establishment to reduce the line speed to ensure that the establishment is able to maintain process control, that establishment sorters are able to successfully perform their duties, and that FSIS CIs are able to conduct a proper inspection.
Comment:
A comment from an animal welfare advocacy organization said that by requiring establishment employees to sort out damaged carcasses before FSIS conducts online inspection, employees remove the evidence, i.e., the carcasses themselves, that birds may have died from causes other than slaughter. The comment asserted that this eliminates one means by which FSIS can verify that establishments are employing good commercial practices.
Response:
Inspectors in both HIMP and non-HIMP establishments verify that poultry is being slaughtered in accordance with good commercial practices. Compliance with these requirements ensures that poultry are handled humanely prior to FSIS online inspection. On a daily basis, FSIS offline inspectors observe operations in the receiving, hanging, stunning, bleeding, and pre-scalding areas in both HIMP and non-HIMP establishments. Compliance and enforcement actions are taken as warranted and necessary.
b. Online Carcass Inspection
Comment:
Several consumer advocacy organizations expressed concern that online inspectors will only look at the back of the bird under the NPIS. The petitions submitted in response to proposed rule also raised this issue. The comments stated that it is necessary to inspect the front and inside of the carcass in order to detect food safety defects. According the comments, under the NPIS, most inspectors will only look at the back of the bird as it quickly moves down the line and are therefore less likely to identify food safety defects in each carcass.
Response:
FSIS disagrees with the comments. The CI carcass presentation under the NPIS allows the CIs to focus their inspection on the same condemnable diseases and conditions that online inspectors focus on under the current inspection systems. Therefore, the Agency believes that the CI carcass presentation under the NPIS will allow the CI to conduct an effective online inspection to detect food safety defects.
Under all four existing inspection systems, i.e., SIS, NELS, NTIS, and Traditional Inspection, FSIS online inspectors focus their inspection on identifying and condemning carcasses with septicemic and toxemic animal diseases and other condemnable conditions that cannot be corrected through trimming or reprocessing. Unlike septemia/toxemia, visible fecal material on the surfaces of a carcass is a food safety defect that can be corrected through reprocessing. Therefore, all poultry slaughter establishments have an online or offline reprocessing system for carcasses accidentally contaminated with fecal material.
Under the current inspection systems, FSIS online inspectors do not issue NRs or condemn carcasses if they observe visible fecal contamination on the interior or exterior carcass surfaces. The Agency ensures that the establishment reprocesses the carcasses after online inspection to remove any fecal contamination before the carcasses enter the chiller. Unlike the NPIS, after such reprocessing, none of the current inspection systems provide for an additional online carcass inspection before the reprocessed birds enter the chiller.
FSIS online CIs under the NPIS will continue to focus on identifying and condemning carcasses with septicemic and toxemic animal diseases and other condemnable conditions that cannot be corrected through trimming or reprocessing. In addition, while the Agency will continue to ensure that the establishments operating under the NPIS reprocess carcasses to remove any visible fecal contamination before the carcasses enter the chiller, the FSIS online CI will also inspect all of the carcasses after they have been sorted, washed, trimmed, and reprocessed, before the carcasses enter the chiller. If there is evidence of fecal material on a carcass, or that the carcass is affected with septicemia or toxemia, the CI will stop the line to prevent the affected carcass from entering the chiller. In addition, the CI will issue an NR because the establishment's procedures for preventing visible fecal contamination and for addressing carcasses with septicemia/toxemia were not effective.
Poultry diseases and conditions, except for avian visceral leukosis, are readily identified by observing the carcass alone; pathogens require testing. Inspection of the outside of the carcass for signs of septicemia/toxemia is sufficient to determine whether the carcass and corresponding viscera must be condemned. Carcasses affected with systemic septicemic or toxemic conditions are darker in color due to dehydration or hemorrhaging and may be smaller or have less body fat due to inappetence or increased metabolic rate. Accordingly, inspection of the exterior of the carcass in accordance with the presentation required under the NPIS is sufficient for CIs to effectively identify and condemn carcasses affected with septicemia/toxemia, along with their corresponding viscera. As discussed elsewhere in this document, an FSIS offline inspector will determine the leukosis status of each flock slaughtered. Viscera in leukosis positive flocks will be inspected by FSIS inspectors.
Thus, online inspection under the NPIS is at least as good, if not better, than online inspection under the current inspection systems. CIs under the NPIS will focus their inspection not only on detecting septicemic and toxemic animal diseases, but on detecting visible fecal contamination as well. In addition, as discussed throughout this document, the VI under the NPIS will conduct carcass verification checks on carcass samples collected before the CI station to ensure that the establishment is effectively sorting carcasses and that it is producing products that comply with the Agency's zero visible fecal tolerance and other performance standards. The VI and CI will work with the IIC to ensure that food safety defects or other conditions do not impair the CI's ability to effectively inspect each carcass.
Comment:
Several labor unions and consumer advocacy organizations expressed concern that the NPIS does not require that an inspector examine the viscera of each bird or be able to identify each bird's viscera for inspection. These comments asserted
that an examination of the viscera is important in determining whether or not a bird is diseased, contaminated, or otherwise adulterated.
Response:
All poultry diseases and conditions, except for avian visceral leukosis, are readily identified by observing the carcass alone. If the CI identifies a carcass with a condemnable condition, the viscera associated with that carcass must also be condemned. When a carcass is condemned, establishments that have maintained the identity of the corresponding viscera must dispose of that viscera as inedible or, where the identity has not been maintained, must dispose of all viscera harvested within the time period related to the condemned carcass. In either case, the CI's visual examination of each carcass also determines the disposition of the corresponding viscera. The CI's online carcass inspection serves as an inspection of the viscera.
Additionally, FSIS inspectors conduct verification checks on all harvested giblets and necks and will apply the RTC standards under the NPIS. These inspection activities ensure that carcasses and parts, including viscera, have been inspected and are determined by FSIS inspectors to be not adulterated. Inspection procedures for avian visceral leukosis are discussed below.
Comment:
A few labor unions expressed concern that under the Traditional Inspection System retained by the proposed changes to part 381, there is no guarantee that an inspector will be able to inspect a carcass along with its viscera. The unions stated that under the current inspection regulations, the carcass and its viscera are inspected together, as the viscera is required to be “uniformly trailing or leading.” One union was of the view that while proposed 9 CFR 381.76(c) requires that the identity of each bird's viscera be “maintained in a manner satisfactory to the inspector until such inspection is made,” this seems to depart from the current requirements in 9 CFR 381.76 because the “new” Traditional Inspection System does not ensure that the viscera and the corresponding carcasses can be inspected by a government inspector.
Response:
As discussed above, under this final rule, the regulations that prescribe requirements for the existing inspection systems will remain in place, with some modifications. Thus, the regulations for all inspection systems except the NPIS and HIMP will continue to require that carcasses and viscera be inspected together.
c. Inspection for Avian Visceral Leukosis
As discussed in the preamble to the proposed rule, avian visceral leukosis is a rare manifestation of the viral disease leukosis that is not transmissible to humans (77 FR 4421-2). Avian visceral leukosis can only be detected by observing the viscera. Avian leukosis does not present a human health concern, but it may render poultry unwholesome or otherwise unfit for human food.
As explained in the preamble to the proposed rule, it is common commercial practice to vaccinate each chicken flock for viral leukosis. On rare occasions, the vaccine is not effective. If it is not, visceral leukosis is present on a flock basis. Thus, under the NPIS, an offline inspector will observe the viscera of the first 300 birds slaughtered from each young chicken flock to determine whether the disease is present in the flock. FSIS has followed this practice in young chicken HIMP establishments, and it has been effective. In the HIMP report, FSIS explained that “[i]t is FSIS's experience that when a flock has avian visceral leukosis, 10 to 15 percent of the birds in the flock have detectable leukosis lesions. For a flock in which 10% of the birds have detectable avian leukosis, a 300 bird sample provides a greater than 95% probability of detecting 22 or greater more birds with visible leukosis lesions” (HIMP Report, p. 26). From these calculations, the Agency concluded that a 300-bird sample is adequate to detect avian leukosis in a flock.
FSIS received several comments on the proposed avian visceral leukosis inspection procedures.
Comment:
A trade association and a poultry producer argued that FSIS should eliminate the proposed avian visceral leukosis check. According to the trade association, the check serves no meaningful public health purpose, is not scientifically sound, and is an outdated inspection approach. The trade association stated that when avian leukosis inspection procedures were originally designed, scientists did not know that the condition is caused by Marek's Disease and the Avian Leukosis Complex. According to the trade association, modern treatment and flock handling practices have effectively eliminated these diseases in commercial poultry operations. The comment stated that comprehensive literature reviews of these conditions, including a statement by the National Institutes of Health, have concluded that neither disease presents an apparent risk to public health. Additionally, the trade association stated that as early as 1984, Agency data has shown that avian visceral leukosis was present in only .017 percent of young chickens slaughtered, and that number is lower today.
Response:
As noted by the comment, avian visceral leukosis is not transmissible to humans and does not present a human health concern. However, it may render poultry unwholesome or otherwise unfit for human food. Thus, carcasses affected with the condition need to be condemned. Because avian visceral leukosis, if present, will be present throughout an entire flock, inspecting the first 300 birds of each flock is an appropriate and effective procedure for identifying the disease.
Under the NPIS, an establishment must ensure that it can identify viscera and parts corresponding with each carcass. This final rule also requires that establishments operating under the NPIS provide a location along the production line at which an inspector can inspect for avian leukosis “the first 300 carcasses of each flock together with associated viscera either uniformly trailing or leading, or otherwise identified with the corresponding carcass.”
Comment:
One labor union stated that FSIS is going to require that establishments notify the IIC when they intend to slaughter a new flock so that FSIS may inspect the viscera, but “flock” is not defined anywhere in the regulations. The union stated that FSIS's clarification of flock, “In general, a flock constitutes birds raised under similar circumstances on the same premises” in the preamble to the proposed rule is imprecise and the clarification is not included in the PPIA or Federal regulations. The union asserted that this lack of a definition of “flock” makes the process for detecting avian leukosis problematic.
Response:
Establishments are able to identify which birds belong to the same flock because birds from the same flock, i.e., birds that have been raised under similar circumstances on the same premises, arrive at slaughter together. Establishments operating under the NPIS will identify when a new flock arrives and are required to notify the IIC when they intend to slaughter a new flock.
d. Verification Inspection
Comment:
A consumer advocacy organization and a trade association requested that FSIS clarify the role of the VI under the NPIS. The consumer advocacy organization requested that FSIS explain how the NPIS will enable
inspectors to conduct more food safety checks; whether more VIs will be assigned to each slaughter line; whether VIs will have to cover more than one slaughter line in an establishment; and whether VIs will have more than one establishment to cover on a given shift, similar to processing assignments.
Response:
There will be one CI and one VI assigned to each evisceration line per shift in establishments that chose to operate under the NPIS. As stated throughout this document, because the establishment's employees will be responsible for sorting carcasses, disposing of carcasses that must be condemned, and conducting any trimming or reprocessing activities before carcasses are presented to the online CI, the CI will be better able to focus on detecting carcasses with visible defects that impact food safety, such as visible fecal contamination and septicemia/toxemia.
In addition to online inspection performed by CIs, VIs under the NPIS will conduct offline food safety-related inspection activities and will monitor and evaluate establishment process controls. The VIs will conduct carcass verification checks on carcass samples collected before the CI station to ensure that the establishment is effectively sorting carcasses and that it is producing products that comply with the Agency's zero visible fecal tolerance and other performance standards. As in HIMP, VIs under the NPIS will also conduct an array of other inspection activities that are important to ensure food safety, such as performing ante-mortem inspection; collecting samples for pathogen testing; verifying the effectiveness of the establishment's HACCP system; and verifying that the establishment is meeting sanitary dressing requirements. As noted throughout this document, the VI and CI will work with the IIC to ensure that food safety defects or other conditions do not impair the CI's ability to effectively inspect each carcass.
Comment:
A consumer advocacy organization argued that the NPIS decreases the protections that are part of the HIMP pilot program. The comment stated that under HIMP, VIs collect and examine 10-bird samples for food safety defects every hour, and examine at least two of the 10-bird samples for wholesomeness defects. Because of the decreased role of online FSIS inspectors, the consumer advocacy organization stated that these 10 bird samples are the only hands-on verification of poultry carcasses under HIMP. This comment expressed concern that the NPIS does not provide for scheduled verification checks, i.e., food safety or wholesomeness checks, and the Agency has been unwilling to commit to any specific number of scheduled checks.
Response:
FSIS agrees with the consumer advocacy organization's assertion that the verification checks that VIs conduct for food safety defects under HIMP are necessary to ensure that establishment employees are effectively sorting carcasses and disposing of carcasses that must be condemned before the carcasses are presented to the CI. Effective carcass sorting by establishment employees is essential for the CI to conduct an efficient and effective online carcass-by-carcass inspection. Therefore, under the NPIS, VIs will continue to conduct carcass verification checks for food safety defects at a point in the slaughter process before the CI's online fixed position. VIs will also verify that establishments are effectively addressing OCP defects through review of establishment records documenting that the establishment is producing RTC poultry and through observation of carcasses when conducting verification checks.
Because HIMP was a pilot study, the activities for offline VIs needed to be more controlled and prescriptive to ensure that the data collected from each establishment participating in the study were consistent. Under the NPIS, the carcass verification checks will be more risk-based to reflect the performance of the establishment. Thus, for some establishments, VIs may conduct more carcasses verification checks under the NPIS than they do under HIMP.
Under the NPIS, the Agency will follow the same procedure used under HIMP to schedule VI carcass checks for food safety defects to ensure that VIs collect an appropriate number of verification samples to assess each establishment's performance under the NPIS. The Agency will monitor and analyze the ongoing results of its verification activities to assess the effectiveness of the establishment's carcass sorting and other process control procedures. The Agency will modify carcass verification checks and other verification activities as needed to respond to findings through the same data-driven process that FSIS uses for all in-plant verification.
The inspection results recorded in PHIS provide FSIS with the information it needs to ensure that verification activities are targeted at identified public health risks. Under PHIS, FSIS is able to modify verification activities to respond to specific situations in individual establishments, to findings in a particular type of establishment, or across the entire regulated industry. In-plant inspection personnel use PHIS to initiate additional verification tasks if their inspection findings raise concerns about an establishment's compliance with regulatory requirements. FSIS managers use PHIS to initiate additional verification and sampling tasks in individual establishments in response to certain criteria, such as not meeting the
Salmonella
performance standard. They are also able to adjust the frequencies and priorities of verification tasks on a national level to quickly shift inspectors' focus to verify requirements where findings indicate problems may be occurring.
Comment:
One trade association requested that the Agency clarify where in the process a finding of fecal contamination would result in a regulatory noncompliance.
Response:
Similar to HIMP, under the NPIS, the VI will issue an NR for visible fecal contamination if the VI detects such contamination when performing carcass verification checks.
In addition, this final rule requires that all poultry slaughter establishments develop, implement, and maintain written procedures to ensure that carcasses with visible fecal contamination do not enter the chiller and to incorporate these procedures into their HACCP systems. It also requires that all poultry slaughter establishments develop, implement, and maintain written procedures to prevent fecal contami
This text is long and has been trimmed here. Open the source document for the complete record.
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.