[Federal Register: January 28, 2008 (Volume 73, Number 18)]
[Page 4767-4774]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]



Food Safety and Inspection Service

[Docket No. FSIS-2006-0034]

Salmonella Verification Sampling Program: Response to Comments 
and New Agency Policies

AGENCY: Food Safety and Inspection Service, USDA.

ACTION: Notice, response to comments, and request for comments.


SUMMARY: The Food Safety and Inspection Service (FSIS) is announcing 
new policies for the Agency's Salmonella Verification Sampling Program 
and related activities conducted in meat and poultry establishments. 
These changes include publication of completed FSIS verification sample 
set results for establishments that show inconsistency in their ability 
to meet Salmonella performance standards, beginning with those from 
young chicken slaughter establishments; a voluntary incentive-based 
program for meat and poultry establishments that should yield 
significant data on attribution of human illness to FSIS-regulated 
products; and increasing the Agency's use of targeted sampling 
approaches and collaborative serotype and subtype data. FSIS is taking 
these actions to advance its efforts to achieve the Agency's public 
health goal of significantly reducing human cases of salmonellosis. 
This Notice explains the basis for these actions and responds to 
comments received by the Agency on its Federal Register Notice of 
February 27, 2006 (71 FR 9772-9777) on Salmonella policy.

DATES: Submit comments by February 27, 2008. The policies described in 
the Notice are planned to take effect March 28, 2008.

ADDRESSES: FSIS invites interested persons to submit comments on this 
Notice. Comments may be submitted by any of the following methods:
    Federal eRulemaking Portal: This Web site provides the ability to 
type short comments directly into the comment field on this Web page or 
attach a file for lengthier comments. FSIS prefers to receive comments 
through the Federal eRulemaking Portal. Go to http://www.regulations.gov and, in the ``Search for Open Regulations'' box, 
select ``Food Safety and Inspection

[[Page 4768]]

Service'' and ``Notices'' from the agency drop-down menu and then click 
on ``Submit.'' In the Docket ID column, select the FDMS Docket Number 
to submit or view public comments and to view supporting and related 
materials available electronically. After the close of the comment 
period, the docket can be viewed using the ``Advanced Search'' function 
in Regulations.gov.
    Mail, including floppy disks or CD-ROM's, and hand- or courier-
delivered items: Send to Docket Clerk, U.S. Department of Agriculture, 
Food Safety and Inspection Service, 1400 Independence Avenue, SW., Room 
2534 South Building, Washington, DC 20250.
    Electronic mail: fsis.regulationscomments@fsis.usda.gov.
    All submissions received must include the Agency name and FDMS 
docket number FSIS-2006-0034.
    All comments submitted in response to this Notice will be posted to 
the regulations.gov Web site. Comments will also be available for 
public inspection in the FSIS Docket Room at the address listed above 
between 8:30 a.m. and 4:30 p.m., Monday through Friday.

FOR FURTHER INFORMATION: For further information contact Daniel 
Engeljohn, PhD, Deputy Assistant Administrator for Office of Policy, 
Program and Employee Development, FSIS, U.S. Department of Agriculture, 
Room 349-E, South Building, 14th and Independence SW., Washington, DC 
20250-3700, (202) 205-0495, fax (202) 720-2025, e-mail: 



    The actions announced in this Notice result from an Agency re-
evaluation of its verification sampling program for Salmonella in meat 
and poultry establishments. Pertinent issues were raised in two 
previous FSIS Federal Register Notices published in April 2003 (68 FR 
18593-18596) and February 2006 (71 FR 9772-9777). The Notice of April 
2003 asked for public comment on FSIS practices and its then-current 
thinking about Salmonella. The February 2006 Notice responded to public 
comments on the 2003 Notice, announced changes in how the Agency 
schedules Salmonella verification sample sets and in how it reports 
sample results, and discussed outstanding concerns and possible new 
    The Salmonella verification sampling program formally began with 
the Agency's Final Rule on Pathogen Reduction; Hazard Analysis and 
Critical Control Point (PR/HACCP) Systems that issued on July 25, 1996 
(61 FR 38805-38989; http://www.fsis.usda.gov/OPPDE/rdad/FRPubs/93-
016F.pdf). Among other things, the PR/HACCP rule set Salmonella 
performance standards for establishments that slaughter selected 
classes of food animals or that produce selected classes of raw ground 
products. FSIS uses the Salmonella performance standards to ensure that 
each establishment is consistently achieving an acceptable level of 
performance with regard to controlling and reducing harmful bacteria on 
raw meat and poultry products. The microbiological performance 
standards for reduction of Salmonella in raw products, coupled with 
performance criteria for use with E. coli testing allow FSIS to verify 
the effectiveness of process controls in slaughter establishments.
    Raw products with established PR/HACCP performance standards 
include carcasses of cows/bulls, steers/heifers, market hogs, and young 
chickens (broilers). Processed products measured by performance 
standards include ground beef, ground chicken, and ground turkey. The 
performance standards for these product classes are based on the 
prevalence of Salmonella as determined from nationwide microbiological 
baseline studies that FSIS conducted before PR/HACCP was implemented. 
In addition, FSIS established a guideline level for young turkey 
carcasses, relying on a baseline study that the Agency published in 
2005 (70 FR 8058-8060; http://www.fsis.usda.gov/OPPDE/rdad/FRPubs/02-
    Under the Salmonella verification program as conducted from 1996 to 
2006, the Agency verified that establishments are meeting a Salmonella 
standard or guideline by having FSIS inspection program personnel 
collect randomly-selected product samples from randomly-selected 
individual establishments over the course of a defined number of 
sequential days of production to complete a sample set. Generally, 
these tests were conducted once each year for each establishment. 
Procedures for doing this are described in Appendix E of the PR/HACCP 
Final Rule (61 FR 38917-38928). The product samples are sent to FSIS 
laboratories for analysis. The number of samples in a sample set varies 
by product class. The maximum number of positive samples allowed in a 
set is based on data from the nationwide baseline studies. The 
standards were defined on a product class basis so that an 
establishment operating at the baseline level would have an 80% chance 
of meeting the standard.
    The Agency's February 2006 Notice announced specific actions that 
the Agency intended to take. FSIS has effected these actions over the 
last eighteen months. The following is a list of these actions and 
their current state of implementation:
    1: Reporting results of each individual Salmonella test to 
establishments as they become available.
    Implemented in July 2006.
    2: Posting quarterly nationwide Salmonella data by product class.
    Implemented in June 2006 beginning with the first quarter of 2006. 
See http://www.fsis.usda.gov/Science/Quarterly_Salmonella_Results/index.asp.
    3: Collecting swab samples from turkey carcasses for Salmonella 
    Implemented in June 2006.
    4: Classifying establishments in three process control categories 
according to their performance in completed sample sets relative to the 
regulatory performance standard or baseline guidance level for 
Salmonella percent positive in their product class, and scheduling 
sample collection according to these categories:
    Category 1:--Establishments with percent positive Salmonella 
samples at 50% or less of the performance standard or guideline in the 
two most recent completed sample sets.
    Category 2:--Establishments with percent positive Salmonella 
samples above 50% but not exceeding the standard/guideline in the most 
recent completed sample set.
    Category 3:--Establishments with percent positive Salmonella 
samples exceeding the standard/guideline in the most recent completed 
sample set.
    Implemented in July 2006 when a new scheduling algorithm, based on 
these process control categories, went into effect. The February 2006 
Notice explained that a Category 1 ``establishment will [now] be tested 
no more than once a year, but at least once every two years, unless it 
gets a result that puts it in Category 2 or 3'' (71 FR at 9776). This 
algorithm has the effect of placing greater emphasis on scheduling 
young chicken establishments, which are currently the Agency's top 
    5: Conducting Food Safety Assessments (FSAs) in establishments in 
Category 3 for Salmonella before sample set failure.
    Implemented by FSIS Notice in August 2007.
    6: Issuing a new compliance guideline on control of Salmonella for 
young chicken slaughter establishments.
    Implemented in August 2006 with publication of ``Compliance 

[[Page 4769]]

for Controlling Salmonella in Poultry, First Edition'' at http://www.fsis.usda.gov/PDF/Compliance_Guideline_
Controlling_Salmonella_Poultry.pdf. A second edition, updated to include guidance for 
controlling Campylobacter as well, is in development.
    7: Obtaining serotype (species) data from isolates and timely 
sharing of serotype data with establishments.
    Implemented in May 2006. After one of the FSIS field service 
laboratories reports the analysis results, isolates of Salmonella-
positive samples are serotyped to identify species at the USDA Animal 
and Plant Health Inspection Service's National Veterinary Services 
Laboratories in Ames, Iowa. In recent years virtually all samples 
positive for Salmonella have been serotyped. Identified serotypes are 
communicated to establishments as soon as they are reported by APHIS to 
FSIS, usually within two weeks after a sample has been reported as 
positive. A report listing aggregate identified serotypes by year was 
posted in August 2007 on the Agency Web site at http://www.fsis.usda.gov/Science/Serotypes_
    8: Obtaining timely access to pulsed field gel electrophoresis 
(PFGE) subtype data identifying specific strains of Salmonella 
    Implemented in August 2007 when FSIS and the Agricultural Research 
Service (ARS) signed a cooperative Memorandum of Agreement for subtype 
data sharing. Under this agreement, PFGE subtype information on 
Salmonella isolates collected by FSIS from meat and poultry products is 
matched with subtype information from isolates associated with human 
illness in PulseNet, a data base maintained by the Centers for Disease 
Control and Prevention (CDC). FSIS has routine access to data for all 
subtype isolates maintained by ARS, in a time frame short enough to be 
relevant to in-plant and public health investigations.
    9: Updating baseline studies to better measure improvements in 
pathogen reduction in all classes of raw product.
    Implemented; a new young chicken baseline began in June 2007, and a 
young turkey baseline is scheduled for 2008. Additional baselines are 
being planned for the future.
    10: Monitoring process control changes and movement between 
Categories 1, 2, and 3 by tracking Salmonella percent positive in 
verification samples by product class over the course of a full year 
beginning in July 2006.
    Implemented. In the February 2006 Notice, the Agency established 
its goal of having 90% of all meat and poultry establishments 
manufacturing raw classes of product subject to Salmonella testing in 
Category 1 by October 1, 2010. In addition, FSIS explained that it 
would consider taking action to provide incentives for enhanced public 
health protection if that benchmark had not been met by July 2007. As 
of the benchmark date of June 30, 2007, Category 1 included 73% of 
turkey establishments, 72% of young chicken establishments, 60% of 
market hog establishments, 59% of cow/bull establishments, and 45% of 
steer/heifer establishments. As of September 30, 2007, Category 1 
included 84% of turkey establishments, 73.5% of young chicken 
establishments, 63% of market hog establishments, 61% of cow/bull 
establishments, and 45% of steer/heifer establishments. While these 
results did not reach the July 2007 benchmark, they are encouraging 
signs of progress. Though the Agency is encouraged by this progress, it 
has continuing concerns about the adverse effects of Salmonella on 
public health. FSIS has determined, therefore, that more robust 
incentives, including publication of establishment verification 
sampling results, are necessary to encourage the industry to improve 
its performance in controlling Salmonella.

Public Health Concerns

    To put the 2006-2007 results of the Salmonella verification program 
cited above in perspective, CDC has reported that in 2006 Salmonella 
serotypes accounted for 38.6% of all human foodborne infections, making 
it the most common human foodborne pathogen (http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5614a4.htm?s--cid=mm5614a4--e). Moreover, the U.S. 
Department of Health and Human Services National Food Safety Objectives 
``Healthy People 2010''--which USDA and FSIS use to guide strategic 
planning for public health--set a goal of 6.8 cases of salmonellosis 
per 100,000 persons by 2010 (http://www.healthypeople.gov/document/tableofcontents.htm).
The 2006 rate for human Salmonella infections 
reported by CDC is 14.81 cases per 100,000 persons, more than double 
the target rate for Healthy People 2010. This represents only a slight 
decrease in overall human infections from Salmonella serotypes from an 
incidence of approximately 16.8 cases per 100,000 persons in the CDC 
reference period of 1996-98.
    Meat and poultry products contain some serotypes of Salmonella that 
are rarely found in human patients suffering from salmonellosis, but 
other serotypes found in various meat and poultry products are among 
those often identified in human cases. The seven most commonly 
identified Salmonella serotypes causing human infection in the United 
States according to the CDC are Typhimurium, Enteriditis, Newport, 
Javiana, Montevideo, Heidelberg, and I 4, [5], 12:i:-. Combined, these 
serotypes accounted for 64% of human cases of salmonellosis. The CDC 
notes that of these seven most common Salmonella serotypes found in 
human infections in 2006, only Typhimurium has declined since the 1996-
98 baseline study, and its incidence since 2003 has been stable. S. 
Enteriditis, however, has increased by 28% since 1996-98. S. Newport, 
which has shown signs of mutation toward multi-drug resistance, has 
increased by 42% since 1996-98.
    Most importantly, the CDC has determined that poultry is an 
effective vector for S. Enteriditis, a cause of human infections. A 
FoodNet case-control study found that human infections that were 
associated with eating chicken displayed Enteriditis bacteriophage type 
(PT) strains that matched the predominant strains that are isolated 
from chickens. This match in PT strains indicates that the association 
in the cases between eating chicken and human illness was causal.
    FSIS has reported its own data showing an increase during calendar 
years 2000-2005 in the frequency of isolation in young chickens of 
Salmonella, particularly S. Enteriditis (http://www.fsis.usda.gov/Science/Progress
_Report_Salmonella_Testing/index.asp). The percent 
positive of Salmonella samples in young chicken establishments tested 
by the Agency's PR/HACCP verification program increased from 11.5% in 
2002 to 12.8% in 2003, 13.5% in 2004, and 16.3% in 2005. In 2006, 
however, the Agency's testing program showed the lowest percentage of 
young chickens that had tested positive for Salmonella in several 
years--11.4%. This 30% decrease in percent positive samples from 2005 
is noteworthy, particularly given that beginning with 2006 FSIS has 
directed its resources toward testing young chicken establishments with 
variable or highly variable process control in order to reverse the 
upward trend in percent positives seen in such establishments. This 
one-year reversal is promising, though the change in FSIS sampling 
direction in 2006 makes comparisons of percent positive samples with 
previous years uncertain. The trend will need to be confirmed and of 
course sustained.

[[Page 4770]]

    During the 2000-2005 period, the annual number of S. Enteriditis 
isolates found by FSIS increased more than four-fold, and the 
percentage of establishments with sample isolates positive for 
Enteriditis increased by nearly three-fold (http://www.cdc.gov/ncidod/
EID/vol12no12/06-0653.htm). During 2000-2005, the number of states with 
Enteriditis isolates found in young chickens sampled by FSIS increased 
by 71%. Moreover, the predominant Enteriditis bacteriophage types found 
were PT 13 and PT 8, strains isolated from human infections associated 
with eating chicken as described in the CDC report cited above. FSIS 
data from 2006 show that the proportion of Enteriditis found among all 
poultry isolates of Salmonella jumped from 7.71% in 2005 to 13.66%, a 
77.2% increase (http://www.fsis.usda.gov/PDF/Serotypes_Profile_Salmonella_Tables_&_Figures.pdf#page=6).
    In addition, an emerging concern for the Agency is the Salmonella 
serotype having the antigenic formula I 4,[5],12:i:-, which has been 
increasingly recognized by CDC since the mid-1990's in human illness 
cases. In 2006 the serotype was the 7th most commonly identified 
serotype in U.S. human surveillance data. FSIS began reporting 
antigenic formulas in 2004. In 2004 and 2005, I 4,[5],12:i:- was the 
sixth most commonly identified serotype in young chicken isolates, and 
in 2006 it was the fifth most commonly identified serotype in young 
chicken isolates.
    FSIS notes that product classes other than young chickens have not 
shown such persistent upward trends, and the percentage of positive 
verification samples has declined for all three beef product classes. 
The Agency has become concerned, however, about the rise in human 
illnesses attributable to S. Newport beginning in the late 1990's. S. 
Newport, which has shown signs of developing resistance to antibiotics, 
is found most notably in the cattle classes and especially ground beef 
but has been detected in all FSIS-regulated commodities.

Posting of Completed Verification Sample Set Results

    Given the concerns described above with serotypes of human health 
significance and the fact that the goal of having 90% of establishments 
in Category 1 was not reached by July 1, 2007, FSIS will begin 
publishing completed Salmonella sample set results from young chicken 
slaughter establishments with variable (Category 2) and highly variable 
(Category 3) process control. This decision was foreshadowed in the 
February 2006 Notice when the Agency stated that by providing 
information on Salmonella performance:

    FSIS believes it would be providing the appropriate incentive 
for the meat and poultry slaughter industry to attain consistent, 
good control for Salmonella. (71 FR at 9776)

The Agency believes that it is appropriate to begin by publishing the 
results from young chicken slaughter establishments because the 
Agency's consistent priority has been to address the adverse trends in 
Salmonella percent positive samples seen in young chickens during 2000-
2005 and the serotypes of human health significance found in them. A 
prototype table for Web posting of these results has been published on 
the Agency Web site (http://www.fsis.usda.gov/PDF/Salmonella_Initiative_Prototype.pdf). Posting of these results will begin 60 days 
after the publication of this Notice. Posting of results from Category 
2 and 3 establishments slaughtering other product classes will begin 
120 days after the publication of this Notice.
    FSIS believes it important to publish results from establishments 
in these categories of greater concern because targets set by the 
agency have not been met, despite the Agency's allocation of a greater 
amount of its verification sampling program resources to establishments 
in Categories 2 and 3 since 2006. FSIS is still considering, however, 
whether publishing completed verification sample set results from young 
chicken slaughter establishments showing consistent (Category 1) 
process control, as well as from other meat and poultry product 
classes, would be useful in terms of creating incentives for greater 
process control. FSIS will closely monitor Freedom of Information Act 
(FOIA) requests related to this issue, as well as comments received, 
and expand or modify the publication process as necessary. As with the 
individual sample results that have been reported to establishments 
since 2006, all unpublished completed sample set results are available 
upon request under FOIA.

Salmonella Initiative Program (SIP)

    FSIS stated in the February 2006 Federal Register Notice (71 FR at 

    [T]he Agency is aware that limits on linespeeds are a concern to 
both the young poultry slaughter and the hog slaughter industries. 
If widespread action within these industries controlled Salmonella 
contamination such that the Agency, in its testing of carcasses, 
consistently found industry-wide results at half or below half the 
current standard/baseline guidance, FSIS would be prepared to 
consider allowing the industries to study whether linespeeds could 
be increased above the current regulatory limits. * * * Such studies 
could be conducted through existing regulatory provisions for a 
waiver of the meat and poultry regulations (9 CFR 303.2 and 381.3).

    Since the February 2006 Notice, poultry establishments have 
expressed interest in specific waivers that would either permit an 
increase in the volume of birds slaughtered or allow the chilling of 
eviscerated carcasses in a different manner than is allowed by existing 
time/temperature requirements.
    The Agency has developed the Salmonella Initiative Program (SIP) to 
offer such waivers to Category 1 establishments, which have already met 
the current FSIS goals, as an incentive for volunteer meat and poultry 
slaughter and processing establishments to increase process control 
efforts for Salmonella and Campylobacter. Although the February 2006 
Federal Register notice discussed specifically allowing young poultry 
and hog slaughter industries to study whether linespeeds could be 
increased above the current regulatory limits, all slaughter or 
processing establishments that produce product subject to Salmonella 
standards or guidelines would be eligible to apply for the SIP. The SIP 
was published as a prototype on August 31, 2007 (http://
The SIP is designed to facilitate definite improvements in ongoing 
control for Salmonella in all classes of raw product subject to 
Salmonella standards or guidelines, as stated in 9 CFR 303.1(h) and 
381.3(b). The SIP will have the further benefit of providing key 
microbial data to the Agency from sampling and analysis conducted by 
establishments that would volunteer for the SIP. For SIP, the Agency 
will waive provisions of the meat and poultry regulations so that new 
procedures, equipment, or processing techniques can be tested. To 
participate in the SIP, establishments must submit a request in 
accordance with specific criteria.

Program Criteria

    To be eligible, an establishment must be in Category 1 as a result 
of FSIS Salmonella verification sample testing, with at least one of 
the sample sets being completed since February 2006. An establishment 
selected for the Program must provide FSIS with copies of relevant 
information about its food safety system (e.g., interventions in use, 
volume of production, evisceration equipment type, line speeds, 
laboratory analytic procedures, and establishment sample set percent 
positive rate).
    Such establishments would be expected to operate under a waiver to 

[[Page 4771]]

CFR 310.25 or 381.94 (Contamination with microorganisms; process 
control verification criteria and testing; pathogen reduction 
standards). An establishment selected for the SIP will be required to 
take samples for microbial analysis on each line every day and during 
each shift. The sample set is to be the same size as that used by FSIS 
for verification testing of the specific product class, but, unlike 
current FSIS practice, the establishment may take multiple samples on 
one day. Each month establishments selected for the Program and 
maintaining a Category 1 level of performance must, upon FSIS request, 
provide the Agency with the Salmonella isolates from a randomly 
selected establishment-collected set.
    Each week poultry slaughter establishments selected for the SIP 
will collect at least one sample at both rehang and post chill. 
Establishments will collect the postchill sample at the approximate 
time the carcass sampled at re-hang would move to postchill, so as to 
reflect the time it takes for a carcass to pass from rehang to post 
chill. The establishment is to analyze the samples for Campylobacter, 
generic Escherichia coli, and Salmonella. At least monthly, at least 
one rehang sample and one matched post-chill sample positive for any of 
these microorganisms must be enumerated. The results must be provided 
to FSIS at least quarterly.
    The Salmonella percent positive rate for all sample sets collected 
by either FSIS or the establishment will need to be maintained at or 
below half the acceptable number of positive samples in the current 
standard/guideline--e.g., 6 or fewer positive results from a randomly 
selected set of 51 young chickens. If at any time the percent positive 
rate, with either FSIS or establishment testing, exceeds half the 
acceptable number of positive samples in the current standard/
guideline, the establishment must immediately suspend operating under 
any waiver of the regulations, as well as intensify the frequency of 
daily testing. A Category 1 level must be re-established within one 
quarter for the establishment to remain in the SIP.
    A selected establishment that at any time falls to a Category 2 or 
3 level will have to provide FSIS with all Salmonella and Campylobacter 
isolates from all establishment-collected sets for each week until it 
meets the requirements for a Category 1 establishment, which must be 
accomplished within one quarter. FSIS will conduct serotyping and PFGE 
subtype pattern identification for the isolates and then, in 
collaboration with CDC, assess the patterns against clinical isolate 
patterns in PulseNet. FSIS will provide information to the 
establishment about its samples monthly.
    The Agency will check on establishment process control under SIP by 
evaluating the laboratory performing microbial testing for an 
establishment to ensure that the laboratory is following appropriate 
procedures, reviewing establishment data to ensure that the 
establishment is operating at a Category 1 level, and conducting 
unannounced sampling. For such unannounced sampling, FSIS would obtain 
a portion of the establishment's samples collected that day. For 
samples resulting in the collection of rinsate, FSIS would gather 200 
milliliters of the 400 milliliters of the rinse liquid used in 
establishment sampling per the PR/HACCP final rule. For ground product 
samples, FSIS would take one-half of the establishment collected 
sample, of a weight sufficient for testing. These ``split samples'' 
would be analyzed by FSIS for Salmonella and used to determine whether 
the establishment is maintaining effective and consistent process 
control and whether its testing procedures are valid. Internal Agency 
statistical analysis suggests that use of 12 samples per quarter would 
not provide definite verification that the establishment is in fact 
operating at a Category 1 level, but that given a certain number of 
positive samples (e.g., 3 out of 12 for young chickens), it would be 
sufficient to indicate that FSIS can no longer be confident that the 
establishment is operating at a Category 1 level. Such results would 
indicate that the processes to control Salmonella may be out of control 
and therefore warrant further investigation and action by FSIS. As with 
the individual sample results that have been reported to establishments 
since 2006 and any unpublished completed sample set results, the 
results from unannounced sampling are available upon request under 
    FSIS will terminate the Program in establishments that cannot 
regain a Category 1 status within one quarter--i.e., an establishment 
must have percent positive Salmonella samples at 50% or less of the 
performance standard or baseline guideline.
    Establishments must agree that they will provide FSIS with at least 
30 days notice if they decide to terminate their participation in the 

Application Process

    All SIP applications and requests for waivers should follow the 
guidelines for submitting notifications and protocols for new 
technologies posted on the FSIS Web site at http://www.fsis.usda.gov/
index.asp. Applications should be addressed to Dr. Isabel Arrington at: 
    FSIS foresees no limit on the number of slaughter or processing 
establishments selected for the Program provided that no changes in 
inspection would be required. The Agency, however, will select no more 
than five establishments in which any waiver of regulatory requirements 
may affect inspection whereby additional inspectors are needed. 
Preference for selection of establishments where FSIS inspection 
personnel would be increased will be given to those establishments 1) 
with a positive rate for all sample sets collected by FSIS since 
February 2006 of half the rate required to be in Category 1 (e.g., 5% 
for young chickens), as well as for establishment-collected sample sets 
completed within the past quarter, and 2) that in their HACCP plans 
identify Salmonella as a hazard reasonably likely to occur or have 
written controls in place to address Salmonella within the Sanitation 
Standard Operating Procedures or other HACCP prerequisite programs. 
Establishments requesting waivers that would require additional 
inspectors will have 15 days after the publication of this Notice to 
submit their requests. Establishments that have already submitted 
applications before the publication of this Notice do not need to re-
submit their applications.
    Establishments participating in the SIP can request waivers to FSIS 
regulations if they can demonstrate that operating under such waivers 
would facilitate definite improvements, as provided for in 9 CFR 
303.1(h) or 381.3(b), whereby half the Category 1 criterion for 
Salmonella is maintained. An establishment may further request a waiver 
to any other regulatory requirement that it can demonstrate can be 
appropriately waived under 9 CFR 303.1(h) or 381.3(b). The Agency will 
respond to requests for waivers within 60 days.
    Establishments that are currently operating under waived 
regulations for On-Line Reprocessing (OLR) will have six months from 
the date of publication of this Notice to decide if they wish to 
continue under a waiver by applying for the SIP or otherwise revert to 
operating without a waiver. After that six-month period, all OLR 
waivers will be terminated. An establishment that chooses to terminate 
its OLR waiver or has an OLR waiver terminated at six months after 
publication of this Notice can apply for a waiver under SIP after

[[Page 4772]]

a waiting period of nine months after termination of the old waiver.
    Establishments that are currently operating under waived 
regulations for the HACCP-based Inspection Models Project (HIMP) will 
also have six months from the date of publication of this Notice to 
decide if they wish to continue receiving waivers by applying for the 
SIP or otherwise revert to operating without a waiver. After that six-
month period, all HIMP waivers will be terminated. An establishment 
that chooses to terminate its HIMP waiver or has an HIMP waiver 
terminated at six months after publication of this Notice can apply for 
a waiver under SIP after a waiting period of nine months after 
termination of the old waiver.
    FSIS plans to continue developing the SIP to collect data that 
would allow the Agency to determine the appropriateness of a regulation 
incorporating elements of the SIP into a public health based poultry 
slaughter inspection system.

Restructuring the Salmonella Verification Program

Sampling and Testing Initiatives

    To meet the complex challenge for public health protection 
presented by Salmonella, FSIS has decided that it will need to sample 
meat and poultry products targeted through the use of available data. 
This type of targeted approach represents a shift from randomly based 
sampling of all meat and poultry establishments, regardless of the 
relative risk posed, that the Agency conducted before July 2006 when it 
defined the three process control categories for meat and poultry 
establishments according to their performance relative to a performance 
standard or guideline.
    Using the new sample scheduling algorithm adopted in 2006, each 
month FSIS now schedules approximately 75 new verification sample sets 
for Salmonella in raw classes of product. FSIS allocates its sampling 
resources within classes of raw product based on consideration of 
specified criteria. If criterion 1 (All new plants regardless of 
product class) does not obligate all 75 available sample sets, then 
criterion 2 (All Category 3 plants regardless of product class) is 
fulfilled; when criterion 2 does not obligate all available sample 
sets, then criterion 3 (All Category 2 plants, depending upon product 
class) would be fulfilled, etc. (For a full discussion of the sample 
scheduling algorithm, see http://www.fsis.usda.gov/Science/Scheduling_
Criteria_Salmonella_Sets/index.asp.) As a result of allocating 
resources in this way, rather than sampling randomly, FSIS is now able 
to fulfill many of the higher priority criteria, i.e. the Agency is 
able to complete current sampling of establishments of greater concern.
    Because Category 1 establishments are normally not scheduled for 
verification sampling until at least a year after their last set, the 
Agency is developing an algorithm for random unannounced sampling to be 
conducted at all Category 1 establishments during the period that may 
extend up to two years between full Salmonella verification sets. The 
purpose of this unannounced sampling algorithm is to ensure that 
establishments currently in Category 1 maintain their performance. The 
Agency's current thinking is that it will conduct quarterly, random 
sampling at Category 1 establishments. As with the individual sample 
results that have been reported to establishments since 2006 and any 
unpublished completed sample set results, the results from unannounced 
sampling are available upon request under FOIA.
    The Agency has recognized that low-volume raw ground beef 
producers, producing less than 1000 pounds of product per day, 
constitute a large part of the sampling frame for establishments 
eligible for verification sample set scheduling though they account for 
a very small proportion of the raw ground beef supply. Since production 
of ground beef at these establishments may not occur throughout a week 
or month, sampling them for Salmonella may be extended for a year or 
more before a set is completed, as opposed to no more than a couple of 
months of sampling for higher volume establishments.
    Consequently, FSIS is developing plans for modifying its sampling 
procedures at very low-volume establishments. The Agency has determined 
that a more practical approach for low volume establishments would be 
to test the samples that are already being collected at these 
establishments for E. coli O157:H7 for Salmonella as well. As a result, 
these establishments will be removed from the PR/HACCP verification 
sample set scheduling frame. The FSIS field service laboratories will 
perform qualitative testing for the presence or absence of Salmonella 
using the same methodology, discard criteria, and reporting as those in 
place for Salmonella ground beef HACCP samples. Samples that screen 
positive will be quantitatively analyzed, i.e. the Salmonella organisms 
present will be enumerated, using the MPN (Most Probable Number) 
    Import samples from ground beef or trim may also be tested for 
Salmonella as well as other pathogens of public health concern, with 
the PFGE patterns being entered into PulseNet for attribution purposes.
    FSIS will exclude from the Salmonella verification testing program 
schedule any slaughter establishment that processes all carcasses 
slaughtered into ready-to-eat (RTE) product or diverts all of its raw 
products to another official federally-inspected establishment for 
further processing into a RTE product. If the establishment is 
undergoing sampling for Salmonella, but then elects to send all 
affected product to RTE, FSIS will continue to sample until the set is 
completed. At the end of the set, FSIS will verify that all product 
will be designated for further processing into RTE product within the 
establishment or in another Federal establishment. If and when such 
establishments again produce raw product for sale, they will be re-
scheduled for Salmonella verification sets.
    In addition, FSIS will identify and sample all sub-groups of raw 
classes of product that are not currently tested in full verification 
sets. For example, FSIS intends to identify and sample all ratite or 
religious-exempt operations, at least quarterly. FSIS will report 
individual results back to establishments, determine the aggregate 
percent positive rate for each sub-group of product classes, and post 
the aggregate percent positive rate for the subgroup at the Agency Web 
site. FSIS will also identify and implement a data analysis plan to 
ensure that the Agency is aware of adverse trends in percent positive 
samples. As with all pathogen-positive samples, isolates will be 
further assessed for their PFGE patterns and the patterns will be 
included in PulseNet for attribution purposes.

Data Usage

    As noted above, the Agency committed in the last several years to 
serotyping all isolates of Salmonella that it finds in sampling meat 
and poultry products. In the past, sampling for Salmonella showed only 
positive or negative results, indicating the presence or absence of 
detectable Salmonella but not the specific dominant serotype present in 
a positive sample. Salmonella includes at least 2500 serotypes, or 
subspecies, and many of them are rarely associated with human illness.
    As discussed above, isolates from Salmonella-positive samples are 
sent from the FSIS Field Service laboratories to the USDA-APHIS 
National Veterinary Services Laboratories (NVSL)

[[Page 4773]]

for bacteriophage serotyping. FSIS reports the identified serotypes 
that NVSL finds in the product to the originating establishment after 
it receives the results from APHIS. Obtaining and sharing serotype 
information in this way enables both FSIS and meat and poultry 
establishments to determine and monitor the serotypes of human health 
significance that are appearing in Agency sampling. All identified 
serotypes are now aggregated by year and reported on the Agency Web 
site at http://www.fsis.usda.gov/Science/Serotypes_Profile_
    The Agency is beginning to use the level of common serotypes of 
human health significance, based on percentile distributions of 
available serotype data, in evaluating establishment performance. For 
young chickens, for example, 0-1 such serotypes of human health 
significance in a verification set of 51 samples would be considered a 
low level, 2-4 would be a medium level, and 5 or more would be a high 
level. FSIS is now including the level of serotypes of human health 
significance in the End of Set letter it sends to each establishment 
upon completion of a verification sample set for Salmonella. This 
serotype information will help an establishment in evaluating and 
improving its process control performance.
    The Agency schedules a Food Safety Assessment (FSA) at any 
establishment with either a medium or high level of serotypes of human 
health significance found in the most recent Salmonella verification 
sample set. An FSA is scheduled as soon as possible for any 
establishment that has failed a Salmonella verification set before the 
full set has been completed, as was announced in the February 2006 
Federal Register Notice on Salmonella policy.
    Because the Agency is now able to complete sampling for many high 
priority establishments, it is beginning to advance sample set 
scheduling of Category 1 establishments based on the presence of 
serotypes of human health significance in the product. In some cases, 
this factor may mean that, even though an establishment is in Category 
1, it will be tested within one year of the completion of its last 
sample set.
    Identifying the specific serotypes found in FSIS sampling was an 
important advance in estimating the impact of meat and poultry products 
on public health. A fundamental issue for the Agency has been what the 
public health community calls ``the attribution gap''--the question of 
the relationship between findings of Salmonella in meat and poultry 
products and human cases of salmonellosis. In serotyping all samples 
that are positive for Salmonella, FSIS has taken a key initial step 
toward answering this question and intends to take further steps in the 
near future.
    An even more targeted approach that is now being used extensively 
is subtyping, which identifies the specific genetic strains found in 
positive samples. All FSIS sample isolates are PFGE-subtyped by the 
Agricultural Research Service (ARS), which also tests isolates for 
Anti-Microbial Resistance (AMR) as mentioned above in reference to S. 
    As of August 2007, FSIS has continuous direct access to the ARS 
VetNet database for PFGE subtypes. FSIS is now able to compare the PFGE 
patterns from its positive samples to those available in the CDC 
PulseNet database of cases of human illness. As a result, FSIS should 
be able to get a clearer picture of whether, and the extent to which, 
the products that it regulates are contributing to human illness.
    The Agency is also exploring the possibilities for sharing and use 
of industry data as a supplement to data gathered by FSIS and other 
public organizations. On August 8-9, 2007, FSIS hosted a public meeting 
of the National Advisory Committee on Meat and Poultry Inspection 
(NACMPI) to review and discuss, among other things, mechanisms for 
sharing industry data with FSIS (see http://www.fsis.usda.gov/Frame/
0027.htm). The SIP described above is designed in part to obtain data 
from the industry beyond what FSIS is able to gather itself. The 
serotyping and enumeration data obtained from the SIP may be 
particularly useful in answering the question of to what degree human 
illnesses result from FSIS-regulated products.

Response to Comments on the Federal Register Notice of February 27, 

    In deciding how to proceed, the Agency considered the nine comments 
it received during the comment period on the Notice it published in 
February 2006 (one comment was sent twice in different formats and 
registered twice). One additional comment that was received well after 
the comment period closed involved an industry proposal for a specific 
incentive program.


    One comment stated that the performance categories should be 
expanded to capture more nuances in plant performance over time. This 
point was echoed by other commenters, who expressed concern that the 3-
category scheme is too broad.
    FSIS Response: The Agency continues to believe that the 3-category 
scheme is appropriate. It is easy to administer, and it captures the 
most important information regarding Salmonella control in 

Quarterly Summary of Salmonella Positives

    Several comments asserted that data presented as a quarterly 
aggregate could be misleading since testing is ``biased'' toward plants 
with marginal or substandard performance and because of seasonal 
variation in Salmonella positives. On the other hand, another comment 
stated that, since establishments will now receive individual sample 
results as soon as they are available, they can adjust their processes 
so as to avoid ``failing'' sets and thus may make performance appear to 
be better than it is.
    FSIS Response: The Agency's decision to sample a higher percentage 
of establishments, as described above, and to present establishments in 
performance categories relative to a standard or guideline removes 
inherent ``bias'' that could have been introduced by simply presenting 
gross percentages of positive samples. Sampling a higher percentage of 
establishments captures better the relative performance of different 
product classes, and categorizing establishments according to a 
performance standard or guideline better defines the differences 
between establishments.

Posting of Completed Sets by Establishment

    One comment stated that posting such results will be confusing to 
the public.
    FSIS Response: The Agency believes that the 3-category system is 
easily understandable, that it appropriately and effectively describes 
significant differences between categories of establishments, and that 
posting results will provide a valuable incentive to improve process 
control. Should we find that the public is confused by the information, 
we will revise our approach to ensure that the information is 
meaningful and easily understood.

Testing Frequency

    One comment stated that the current annual frequency of testing 
will not provide adequate evidence of sustained process control, 
particularly for Category 1 establishments, which may only be tested 
once every two years. The

[[Page 4774]]

same comment raised questions as to how the Agency adjusts sampling for 
establishment size and volume. Two other comments urged the Agency not 
to reduce testing frequency for Category 1 establishments.
    FSIS Response: The Agency has carefully considered all pertinent 
factors to maximize testing and data productivity. FSIS intends to 
stagger testing of Category 1 establishments so that a full 24 months 
will not pass before an establishment is tested again. Further, the 
Agency intends to conduct random, unannounced sampling of Category 1 
establishments during the period between full verification sample sets. 
For these reasons, FSIS believes that its frequency of testing will 
ensure that the status of a Category 1 establishment is appropriately 
tracked. It is also important to note that FSIS does not regard the 
Salmonella results alone as an indication of ``adequate evidence of 
sustained process control''. Verification of process control will rely 
on an establishment's ability to meet Salmonella performance standards, 
the establishment's own generic E. coli test results, FSIS inspectional 
observations, reports of illness associated with product produced at an 
establishment, and other factors.

Performance Standards

    One comment noted that the performance standards should be 
reevaluated through regularly updated baseline studies. Another comment 
stressed that continual improvement sought by statistical process 
control approaches requires the tightening of standards. On the other 
hand, one comment argued that the Agency's focus on reducing 
performance to a fraction of the standard or guideline ignores the 
validity of the baseline-derived standard/guideline as an index of 
realistic process capability.
    FSIS Response: The Agency is committed to updating baseline studies 
when needed. The Agency does not agree that establishing performance 
objectives at one-half of the performance standard/guideline ignores a 
baseline standard and that the Agency's objectives for process control 
are realistic and necessary. FSIS believes that further knowledge of 
attribution factors will show that continual improvement in reducing 
occurrence of human pathogens in meat and poultry will reduce the 
incidence of human salmonellosis.

Salmonella Subtyping Methodology

    One commenter recommended phage-typing over pulsed field gel 
electrophoresis (PFGE) for subtyping, and another commenter said more 
discussion was needed before the Agency should choose to use PFGE data. 
Another comment, on the other hand, recommended PFGE for subtyping.
    FSIS Response: Phage-typing is primarily used in reference 
laboratories and is impractical for regulatory purposes. The Agency 
believes that PFGE has proven to be a valid and appropriate methodology 
for obtaining subtype information from verification sampling and 
baseline studies.

Enumeration of Salmonella and Attribution Questions

    One comment urged the Agency to conduct enumeration analysis of its 
verification samples in order to investigate the causal factors in 
human salmonellosis related to dose level.
    FSIS Response: Enumeration is very expensive and of doubtful value 
for practical regulatory purposes that are qualitative in nature. FSIS, 
however, is committed to exploring questions of attribution for human 
disease and recognizes that enumeration of Salmonella would have a 
positive role to play in such an investigation. For this reason, the 
Agency is requiring participants in the Salmonella Initiative Program 
to enumerate a portion of their Salmonella and Campylobacter isolates.

Incentives and Further Actions

    One comment stated that the Agency should begin posting all 
completed sample sets immediately rather than first allowing a one-year 
period for collecting data to determine whether publication of 
establishment results was necessary. Two comments urged the Agency not 
to consider any modifications in inspection practices without strong 
evidence of superior establishment performance.
    FSIS Response: The Agency believes that the lead-time announced in 
the February 2006 Notice of one year (from July 2006 to July 2007) for 
tracking results was appropriate. The key point is that the Notice 
informed the industry that process control improvements were crucial 
and needed to be accomplished in a timely manner. The Agency agrees 
that modifications in inspection should only occur if there is strong 
evidence of superior establishment performance, and it is exploring 
such possibilities in the Salmonella Initiative Program for Category 1 
establishments described above.

Additional Public Notification

    Public awareness of all segments of rulemaking and policy 
development is important. Consequently, in an effort to ensure that 
minorities, women, and persons with disabilities are aware of this 
notice, FSIS will announce it online through the FSIS Web page located 
at http://www.fsis.usda.gov/regulations/2008_Notices_Index/. FSIS 
will also make copies of this Federal Register publication available 
through the FSIS Constituent Update, which is used to provide 
information regarding FSIS policies, procedures, regulations, Federal 
Register notices, FSIS public meetings, and other types of information 
that could affect or would be of interest to constituents and 
stakeholders. The Update is communicated via Listserv, a free 
electronic mail subscription service for industry, trade groups, 
consumer interest groups, health professionals, and other individuals 
who have asked to be included. The Update is also available on the FSIS 
Web page. Through the Listserv and Web page, FSIS is able to provide 
information to a much broader and more diverse audience. In addition, 
FSIS offers an e-mail subscription service which provides automatic and 
customized access to selected food safety news and information. This 
service is available at http://www.fsis.usda.gov/news_&_events/Email_Subscription/index.asp. 
Options range from recalls to export information 
to regulations, directives and notices. Customers can add or delete 
subscriptions themselves, and they have the option to password protect 
their accounts.

    Done in Washington, DC on: January 22, 2008.
Alfred V. Almanza,
[FR Doc. E8-1432 Filed 1-25-08; 8:45 am]