[Federal Register Volume 69, Number 7 (Monday, January 12, 2004)]
[Rules and Regulations]
[Pages 1874-1885]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 04-626]


-----------------------------------------------------------------------

DEPARTMENT OF AGRICULTURE

Food Safety and Inspection Service

9 CFR Parts 301, 318, and 320

[Docket No. 03-038IF]
RIN 0583-AC51


Meat Produced by Advanced Meat/Bone Separation Machinery and Meat 
Recovery (AMR) Systems

AGENCY: Food Safety and Inspection Service, USDA.

ACTION: Interim final rule and request for comment.

-----------------------------------------------------------------------

SUMMARY: The Food Safety and Inspection Service (FSIS) is issuing this 
interim final rule on meat produced by advanced meat recovery (AMR) 
systems. This new regulation is a prophylactic measure designed, in 
part, to prevent human exposure to the Bovine Spongiform Encephalopathy 
(BSE) agent by ensuring that AMR systems are not a means of introducing 
central nervous system tissue into product labeled as ``meat.'' In 
addition to the measures related to BSE, FSIS is finalizing 
restrictions related to bone solids and bone marrow for livestock 
products. This rule articulates the criteria that FSIS will use to 
ensure that AMR products can be represented as ``meat'' and thus are 
not adulterated or misbranded. Finally, the Agency is requiring that 
Federally-inspected establishments that process the carcasses or parts 
of cattle develop, implement, and maintain written procedures for the 
removal, segregation, and disposition of specified risk materials 
(SRMs), including non-complying product from beef AMR systems. 
Establishments must incorporate these procedures into their HACCP plans 
or in their Sanitation SOPs or other prerequisite program. FSIS is 
issuing this document as an interim final rule because of the discovery 
of a BSE-positive cow in this country.

DATES: This interim final rule is effective January 12, 2004. Comments 
on this interim final rule must be received by April 12, 2004.

ADDRESSES: Submit written comments to: FSIS Docket Clerk, Docket 
03-038IF, Room 102, Cotton Annex, 300 12th Street, SW., 
Washington, DC 20250-3700. Reference materials cited in this document 
and any comments received will be available for public inspection in 
the FSIS Docket Room from 8:30 a.m. to 4:30 p.m., Monday through 
Friday. Reference materials that are not copyrighted will also be 
available on the FSIS Web site at http://www.fsis.usda.gov. All 
comments will be available for inspection in the FSIS Docket Room or on 
the FSIS Web site at http://www.fsis.usda.gov.

FOR FURTHER INFORMATION CONTACT: Daniel L. Engeljohn, Ph.D., Executive 
Associate, Policy Analysis and Formulation, Office of Policy and 
Program Development, Food Safety and Inspection Service, U.S. 
Department of Agriculture, Washington, DC 20250-3700; (202) 205-0495.

SUPPLEMENTARY INFORMATION:

Table of Contents

Background

[[Page 1875]]

BSE
Previous Rulemaking
Discussion of Public Comments on Docket 96-027P
Overview of this Interim Final Rule and Requests for Comment
Emergency Action
Codified Text

Background

    The mission of the Food Safety and Inspection Service (FSIS) is to 
ensure that meat and meat food products are wholesome, not adulterated, 
and properly marked, labeled and packaged. Under the Federal Meat 
Inspection Act (FMIA) (21 U.S.C. 601 et seq.), FSIS has the authority 
to determine that product is unfit for human food, i.e., adulterated, 
within the meaning of section 1(m)(3) of the FMIA (21 U.S.C. 
601(m)(3)). Furthermore, a meat or meat food product is misbranded 
under any of a number of circumstances, including if its labeling is 
false or misleading in any particular; if it is offered for sale under 
the name of another food; if it is an imitation of another food, unless 
its label bears (in type of uniform size and prominence) the word 
``imitation'' and, immediately thereafter, the name of the food 
imitated; or if it purports to be or is represented as a food for which 
a definition and standard of identity or composition is prescribed by 
regulations, unless it conforms to the regulations and its label bears 
the name of the food specified in the definition and standard (21 
U.S.C. 601(n)(1), (n)(2), (n)(3), and (n)(7)). This interim final rule 
addresses both the adulteration and misbranding provisions of the FMIA.

BSE

    Bovine Spongiform Encephalopathy (BSE) is a slowly progressive 
degenerative disease that affects the central nervous system (CNS) of 
adult cattle and is a member of the family of diseases known as 
transmissible spongiform encephalopathies (TSEs). TSEs also include 
scrapie in sheep and goats, chronic wasting disease in elk and deer, 
and variant Creutzfeldt-Jakob Disease (vCJD) in humans.
    The typical incubation period (the time from when an animal becomes 
infected until it first shows signs of disease) is believed to be from 
two to eight years. BSE was first documented in the United Kingdom in 
1986, and has since been identified and confirmed in a number of other 
European and non-European nations.
    The agent that causes BSE and other TSEs has yet to be fully 
characterized. The theory that is most accepted in the scientific 
community is that the agent is a prion, which is an abnormal form of a 
normal protein known as cellular prion protein, although other types of 
agents have been implicated. FSIS has determined that this interim 
final rule is necessary to ensure that AMR systems are not a means of 
introducing CNS-type tissues (including brain, trigeminal ganglia, 
spinal cord, and dorsal root ganglia (DRG)), which have been identified 
as a potential source for the BSE infective agent into the food supply.

Animal Age and BSE Infectivity

    Age-of-onset was known and recorded for approximately 135,000 
cattle with confirmed clinical BSE in the United Kingdom between 1988 
and August 2003. The age distribution data show that, of the cattle 
that developed clinical BSE in the field, only 0.01 percent were less 
than 30 months of age. Therefore, cattle younger than 30 months of age 
are less likely to be in the later stages of BSE incubation than older 
BSE-infected cattle and are less likely to contain high levels of BSE 
infectivity. For additional information about the onset of clinical 
BSE, see the interim final rule ``Prohibition of the Use of Specified 
Risk Materials for Human Food and Requirements for the Disposition of 
Non-Ambulatory Disable Cattle,'' Docket No. 03-025IF, also in this 
issue of the Federal Register.
    FSIS is providing a method for its inspection program personnel in 
slaughter establishments to use to determine the age of cattle when 
supporting documentation is not provided by the establishment. This is 
relevant to this rulemaking on advanced meat/bone separation machinery 
and meat recovery (AMR) systems because AMR systems generally are 
operated separate from slaughter operations. Thus, establishments will 
need to process skulls and vertebral columns under control programs 
(i.e., Hazard Analysis Critical Control Point (HACCP) plans, Sanitation 
Standard Operation Procedures (Sanitation SOPs), or prerequisite 
programs) separate from their slaughter operation controls. To ensure 
that the skulls and vertebral columns are appropriately handled, the 
slaughter establishment will need to provide documentation associated 
with the age of the skulls and vertebral columns to the receiving 
processing operation. Establishments using AMR systems will need to 
ensure that the skulls and vertebral columns are not from cattle 30 
months of age and older.

Infective Tissue

    In 2001, the European Commission's Scientific Steering Committee 
(SSC), an advisory committee for the European Union, considered the 
amount and distribution of BSE infectivity in a typical case of BSE and 
estimated that, in an animal with clinical disease, the brain contains 
64.1 percent of the total infectivity in the animal, and the spinal 
cord contains 25.6 percent. According to the SSC, the highest remaining 
proportion of infectivity in a typical animal with clinical BSE is 
found in the DRG (3.8 percent). In experimentally infected cattle with 
clinical BSE, infectivity has been demonstrated in the brain, spinal 
cord, DRG, trigeminal ganglia, and the distal ileum of the small 
intestine. For additional information about BSE infectivity, see Docket 
No. 03-025IF.

The Harvard BSE Risk Assessment

    In 1998, USDA commissioned the Harvard Center for Risk Analysis to 
conduct an analysis and evaluation of the current measures implemented 
by the government to prevent the introduction and spread of BSE in the 
United States and to reduce the potential exposure of consumers to the 
BSE agent.
    Using a probabilistic simulation model to characterize the 
consequences of introducing BSE into the country through a variety of 
pathways, the Harvard study concluded that the risk to consumers in the 
United States was low, and that the country is highly resistant to the 
spread of the disease, if introduced.\1\
    In evaluating the potential risk mitigation actions that could be 
taken to further reduce the likelihood that BSE could spread to cattle 
or humans, the risk assessment recommended three courses of action. The 
first is to prevent infected or potentially infected animals or 
contaminated feed from entering the country. The second is to ensure 
compliance with Food and Drug Administration's (FDA's) ruminant feed 
ban. The third is to prohibit the infective materials of BSE-infected 
animals from entering both the human food and animal feed chains.
    The Harvard study divided potential sources of human exposure to 
BSE infectivity into two categories: Specific high-risk tissues and 
contamination of low-risk tissues. The former include, in order of 
infectivity, brain, spinal cord, DRG, distal ileum, trigeminal ganglia, 
and other tissues found in the head (e.g., eyes and tonsils). As for 
the latter, the Harvard study indicated that the most important means 
by which low-risk tissue can become contaminated is through the use of 
AMR systems that can leave spinal cord and DRG in the recovered meat 
product.

[[Page 1876]]

The AMR Process

    AMR systems are newer models of systems that have been used since 
the 1960s. The new systems emulate the physical action of hand-held 
high-speed knives for the removal of skeletal muscle tissue from bone 
through the use of hydraulic pressure. AMR systems apply pressure to 
detach the meat (skeletal muscle) tissue from the bones in a ``hard 
separation'' process. Desinewers that typically use belt pressure 
against a rotating perforated steel drum then separate meat from 
connective tissue, sinews, and other non-meat components in a ``soft 
separation'' process. In addition to vertebrae, typical bones processed 
by piston-driven AMR systems are brisket bones (breast or lower chest), 
rib bones, flat bones (scapulas), and hip bones (pelvis).
    AMR product is an intermediate product that is typically blended at 
about 5 to 12 percent of the formulation of ground products derived 
from manufacturing trimmings. Descriptive labeling for the product of 
AMR includes ``(species) trimmings, finely textured,'' ``finely ground 
(species),'' or any other term that accurately reflects its form.
    AMR technology enables processors to remove attached skeletal 
muscle tissue from livestock bones without incorporating significant 
amounts of bone and bone products into the final meat product. When 
produced properly, product from AMR systems is comparable to meat 
derived by hand deboning and can be labeled as ``meat'' (9 CFR 301.2). 
Under the FSIS regulations, spinal cord is not a component of meat, and 
therefore, product from AMR systems identified as ``meat'' that 
contains spinal cord is misbranded. Until today, FSIS has not taken 
regulatory action against ``meat'' containing DRG and other CNS-type 
tissues.
    From January through August 2002, FSIS conducted a survey of AMR 
products derived from the vertebral column of cattle to establish a 
baseline for the prevalence of spinal cord and DRG in beef AMR products 
(referred to as the 2002 Beef AMR Survey). In the 2002 Beef AMR Survey, 
the Agency found that while some establishments were able to 
consistently produce beef AMR product that was free of spinal cord and 
DRG, a majority of the establishments had difficulty keeping spinal 
cord and DRG out of their AMR products. Overall, FSIS found that that 
approximately 76% (25 of 34) of the establishments whose AMR product 
was tested had positive laboratory results for spinal cord, DRG, or 
both in their final beef AMR products. The survey also found that 
approximately 35% (89 of 256) of all final AMR product samples that 
were tested had positive laboratory results for spinal cord, DRG, or 
both.
    In March 2003, after completion of the 2002 Beef AMR Survey, FSIS 
implemented a routine regulatory sampling program of beef products from 
AMR systems as an additional measure to prevent misbranding of beef AMR 
products. Prior to the implementation of this regulatory sampling 
program, FSIS inspection program personnel collected AMR product 
samples for analysis for the presence of spinal cord tissue only if 
they believed that the establishment was not completely removing spinal 
cord from the vertebral column before the vertebral bones entered the 
AMR system (FSIS Directive 7160.2, April 14, 1997). Under the revised 
regulatory sampling program, FSIS inspection program personnel take 
samples of beef AMR product on a routine basis to verify that spinal 
cord tissue is not present in such product (FSIS Directive 7160.03, 
Revision 1, August 25, 2003). If spinal cord tissue is detected in beef 
AMR product, FSIS inspection program personnel take regulatory control 
action against the AMR product and equipment to prevent misbranded 
product from entering commerce. If the establishment has distributed 
misbranded beef AMR product, FSIS requests a voluntary recall.
    Removal of the spinal cord before the vertebral columns enter the 
AMR system does not always ensure that spinal cord or DRG will not be 
incorporated into the final product. The Harvard study (discussed 
below) found that, if a beef carcass is mis-split when the spinal cord 
is removed, a portion of the spinal cord may remain encapsulated in the 
spinal canal of the vertebral column, and, if it is not removed before 
the vertebral bones enter the AMR system, the spinal cord could 
contaminate the final AMR product. Even when the spinal cord is 
completely removed from the vertebral column, the DRG of cattle are 
firmly attached to the bones of the vertebral column and are not 
removed along with the spinal cord. Thus, removing the spinal cord from 
the vertebral column does not prevent the DRG from entering an AMR 
system and becoming incorporated into the final AMR product.
    Although FSIS and the regulated industry have recently taken 
actions to prevent the incorporation of spinal cord and, in some 
instances, DRG, in beef AMR products, FSIS continues to detect spinal 
cord and DRG in its routine regulatory sampling of beef AMR products, 
although to a lesser extent than it did in the 2002 Beef AMR Survey. In 
its routine regulatory sampling conducted from March to December in 
2003, FSIS found spinal cord in 23 of 340 randomly scheduled samples, 
an estimated prevalence of 6.8 percent. In addition, the prevalence in 
follow-up samples was 13.6 percent, indicating that establishments with 
an initial positive continued to have some problems controlling for 
spinal cord in beef AMR systems. While FSIS was testing samples for 
spinal cord, FSIS also recorded the results for DRG. The prevalence for 
DRG was found in 10.9 percent of the samples in which DRG was 
recorded.\2\
    Under the current regulations, AMR product that contains DRG, or 
any other CNS tissue except spinal cord, is not misbranded and can be 
identified as meat. However, given the nature of DRG and other CNS 
tissue except spinal cord, and the fact that BSE has been confirmed in 
a cow in the United States, FSIS has reconsidered its approach to the 
presence of all CNS tissues, particularly from cattle, as further 
discussed below. In addition, for a more complete explanation as to why 
skulls and vertebral columns of cattle 30 months of age and older are 
designated as specified risk materials (SRMs) and cannot be used in AMR 
systems, see Docket No. 03-025IF in this issue of the Federal Register.
    In addition to the measures identified to address BSE through 
restrictions associated with SRMs, FSIS also is identifying additional 
measures to restrict the use of beef product and spent bone materials 
associated with CNS-type tissues from cattle younger than 30 months of 
age, as described below. Finally, FSIS is finalizing new bone solids 
and bone marrow restrictions that are slightly modified from those 
previously proposed for livestock product labeled as ``meat.''

Previous Rulemaking

    In 1994, the Agency published a final rule (59 FR 62551) to amend 
the definition of ``meat'' to include product resulting from AMR 
systems. The 1994 rule reflected the Agency's position that calcium 
limits and the physical conformation of the bones exiting the system 
were sufficient to ensure that the production process was in control, 
and that the characteristics and composition of the resulting product 
were those of meat.
    The rule required that product resulting from the bone separation 
process not exceed a calcium content of 0.15 percent or 150 milligrams/
100

[[Page 1877]]

grams of product (150 mg/100 g) within a tolerance of 0.03 percent or 
30 mg/100 g of product for each sample analyzed. The rule also required 
that the bones emerging from the AMR machinery be comparable to those 
resulting from hand deboning; that is, they must be essentially intact 
and in their natural physical conformation, such that they are 
recognizable as, for example, loin bones and rib bones, when they 
emerge from the machinery.
    Shortly after FSIS issued the 1994 rule, consumer groups expressed 
concern that the regulatory requirements for meat produced by AMR 
systems were not being met consistently. Consumer groups alleged that, 
in certain AMR operations, the starting materials and machinery were 
being manipulated to produce a product that conformed to the 
requirements for Mechanically Separated (Species) (MS(Species)), a 
finely comminuted meat food product that may include spinal cord and 
dorsal root ganglia (DRG), but not to the requirements for meat. (At 
the time, FSIS considered spinal cord to be central nervous system 
(CNS) tissue. However, FSIS did not include DRG within the meaning of 
CNS tissue. Rather, it considered DRG to be more a part of the 
peripheral nervous system instead of a CNS-type tissue because it was 
contained within the nexus between the spinal cord and the muscle 
tissue.)
    In 1995, FSIS conducted a survey of federally inspected meat 
establishments using AMR systems. Inspection program personnel in 13 of 
the 48 surveyed establishments reported results that were not in 
compliance with the requirements for AMR established in the 1994 
rule.\3\
    To determine whether the product that was being produced by AMR 
systems was compositionally consistent with hand-deboned meat, in 1996, 
FSIS began conducting a survey to profile the chemical and histological 
composition of meat derived from beef neck bones. Beef neck bones from 
the upper vertebral column are split during the slaughter dressing 
process, as opposed to long bones which generally are not split, and 
thus are inherently likely to contribute bone content (e.g., marrow) to 
the product resulting from the AMR system. Samples were found to 
contain spinal cord and fragments of other CNS-type tissue. FSIS 
concluded that the AMR product produced was likely not comparable to 
corresponding hand-deboned product, even when the calcium criterion of 
the 1994 rule was for the most part met.
    The results of the 1996 survey demonstrated that the provisions of 
the 1994 rule, if met, were not sufficient to ensure that AMR product 
would be comparable to hand-deboned meat in composition. A final report 
on the 1996 survey results is available in the Docket Room and on the 
FSIS web site.\4\
    After considering information from consumer groups about compliance 
concerns, reviewing the 1995 field survey and the response to a 1996 
notice soliciting public comment on that survey, and studying the 
results of the 1996 neck bone survey, FSIS concluded that it was 
necessary to propose amending its regulations and to issue a directive 
to inspection personnel to ensure that manufacturers were not 
incorporating spinal cord into AMR product labeled as meat. In 1997, 
FSIS published Directive 7160.2 to instruct inspection program 
personnel that establishments must completely remove spinal cord from 
any neck or back bones before the bones enter the AMR system. The 
directive emphasized that the definition of ``meat'' in 9 CFR 301.2 
does not apply when the use of AMR systems results in product that 
contains spinal cord. FSIS did not address DRG in the directive 
because, at that time, FSIS did not have validated methodology to 
identify DRG, and DRG was not yet identified as a potential risk 
material.
    On April 13, 1998, FSIS issued a proposed rule (63 FR 17959), in 
which it stated that provisions in the 1994 final rule needed revision 
to prevent misbranding and economic adulteration of AMR product labeled 
as ``meat.'' Specifically the Agency proposed to: (1) Adopt performance 
standards for bone solids and bone marrow; (2) adopt a zero tolerance 
for the presence of spinal cord; and (3) delete the provision that 
focused upon the condition of the bones emerging from the AMR systems 
to determine whether or not the production process was in control. The 
Agency's objective was to ensure that the regulations provided clear 
standards for industry to meet.
    Prior to December 23, 2003, FSIS had not addressed AMR systems in 
the context of BSE, although FSIS had taken numerous steps to limit the 
presence of spinal cord in product derived from AMR systems. In 
particular, in March 2003, FSIS announced the results of the 2002 Beef 
AMR Survey and stated that FSIS soon would clarify its intent by 
rulemaking on AMR to ensure that DRG was excluded from the definition 
of product labeled as ``meat.''
    By 2002, FSIS had a validated methodology to detect and discern 
DRG, there was widespread agreement within the scientific community 
that DRG was included within the meaning of CNS-type tissue, and there 
was scientific evidence that DRG carried the BSE infective agent. FSIS 
did not contemplate addressing tissues of brain and trigeminal ganglia 
in product from AMR systems because FSIS was not aware of any 
establishments using bone material, such as skulls, that would contain 
these tissues in the production of meat. Brain and trigeminal ganglia, 
along with spinal cord and DRG, all fit within the meaning of CNS-type 
tissues for purposes of further discussion in this document. Currently, 
FSIS does not analyze meat for tissues of brain and trigeminal ganglia. 
However, since skulls may in the future be used in AMR systems, FSIS is 
reassessing whether it should validate its testing methodology to 
detect and discern brain and trigeminal ganglia in product recovered 
from AMR systems.
    FSIS has concluded that the 1994 rule, the 1998 proposed rule, and 
the FSIS Directives will not keep spinal cord and other CNS-type tissue 
out of product derived from livestock, particularly cattle, that is 
labeled as ``meat.'' FSIS concludes that restrictions for CNS-type 
tissues need to be explicitly stated in the regulations, along with a 
requirement to have written process control procedures and testing by 
the establishment, to ensure that the process control procedures are 
effective in producing product labeled as ``meat.''
    Furthermore, FSIS has initiated a survey on pork AMR products and 
believes that the lack of process control regarding the presence of 
CNS-type tissues in pork product recovered from AMR systems also may be 
a concern. The new requirements in this interim final rule are 
applicable, for the most part, to products derived from pork bones.
    FSIS has decided to publish this new AMR regulation as an interim 
final rule and to address both CNS-type tissues and the restrictions 
related to bone solids and bone marrow. The presence of spinal cord or 
other CNS-type tissue in AMR product, that is, in meat, particularly 
from cattle, represents a potential threat to the public health of the 
United States. The Administrator thus finds that there is good cause to 
make this new AMR regulation effective immediately. It is especially 
designed to prevent the occurrence of spinal cord and other CNS-type 
tissues in ``meat'' and meat food products derived from cattle, and to 
prevent the occurrence of spinal cord and other CNS-type tissues in 
``meat'' derived from livestock other than cattle.
    Before explaining in more detail the provisions of this interim 
final rule, a

[[Page 1878]]

brief discussion of the comments received on the proposal and FSIS' 
responses follows.

Discussion of Public Comments on Docket 96-027P

    The 60-day comment period on the 1998 proposed AMR rule ended on 
June 12, 1998. Forty-five comments were received from food and 
equipment manufacturers, professional and industrial trade 
associations, consumers and consumer advocacy organizations, academia, 
and consultants.
    On December 16, 1999, FSIS issued a notice (64 FR 70200) reopening 
the comment period for an additional 30 days to give the public an 
opportunity to review and comment on the methods and results used by 
Agricultural Research Service (ARS) scientists to derive new iron-to-
protein values. The Agency also sought comment on a report submitted by 
a meat industry group regarding economic and worker safety issues 
relevant to the proposed rule. The reopened comment period closed on 
January 18, 2000. Twenty-six additional comments were received in 
response to the notice. The two sets of comments and FSIS' responses 
are merged in this ``Comment'' section.

Bone Solids

    Comment: Many commenters disagreed with the proposed calcium 
requirement that was established as a measure of the bone solids 
content of AMR product, to ensure that AMR product is meat. One 
commenter stated that the limit was too high, and another suggested 
that the limit should be lowered to approximate the calcium level in 
hand-deboned meat, with a reasonable allowance for variation. Another 
commenter pointed out that FSIS asserted in the 1994 final rule that 
its purpose was to ensure that the characteristics and composition of 
AMR are consistent with those of meat. Another commenter claimed that 
the proposed reduction in the calcium level was arbitrary and 
determined on the basis of a limited data set and not based on actual 
process data. Another commenter requested that the calcium performance 
standard account for differences among meat species.
    Response: FSIS does not agree that the calcium standard should be 
based only on actual process data and does not agree that the calcium 
level for AMR products needs to approximate that of hand-deboned 
products. The calcium level in hand-deboned products is nearly 
negligible. The increased amount in the AMR product that the Agency 
proposed to allow represented a small amount of calcium that would not 
in any appreciable way affect the safety or quality of the product. 
When the vertebrae are split, increased bone dust (i.e., material high 
in calcium) is created and may accumulate in the AMR product. In hand-
deboning, such material is less likely to be incorporated into the 
product. The calcium limit that FSIS proposed was based on the results 
of its 1996 survey and the data that were submitted to FSIS by 
industry. FSIS believes that this calcium limit can be consistently 
achieved by industry and represents a more appropriate level than that 
in the 1994 rule.
    Regarding the comment about different calcium levels for beef and 
pork, FSIS considered data for different species that were submitted by 
industry groups as well as the data gathered by FSIS in the 1996 
survey. A summary of the data is presented in the technical addendum, 
which is available in the Docket Room and on the FSIS web page. The 
data show that average calcium levels for AMR pork and beef products 
are approximately 100 mg/100 g. FSIS believes that these data suggest 
that with regard to bone solids, there would not be any significant 
difference between pork and beef. Therefore, the required calcium 
targets for pork and beef AMR products are the same in this interim 
final rule.
    As mentioned above, in 1994, FSIS believed that the performance 
standards it established regarding calcium as a measure of bone solids 
content, and the physical conformation of the bones exiting the system 
were sufficient to ensure that the AMR production process was in 
control, and that the characteristics and composition of the resulting 
AMR product would be comparable to those of meat. However, based on the 
results of the 1996 AMR survey, FSIS concluded that the established 
performance standards, even if met, were not sufficient to ensure that 
AMR product would be comparable to meat and as a consequence proposed 
different standards in 1998. In particular, regarding compositional 
parameters, the 1996 results showed that the AMR products produced at 
the time were not comparable to hand-deboned product with respect to a 
number of measures, even when the calcium limit designed to measure 
bone solids content was met.
    The 1998 proposed rule identified a calcium limit of 130 mg/100 g 
product. This level was premised on a target average level of 
approximately 100 mg/100 g product but did not specify whether the 130 
mg/100 g was an average or an absolute level. Data collected by the 
Agency and submitted by industry indicated that the average calcium 
level obtained for AMR pork and beef products is approximately 100 mg/
100 g, but that there was wide variation in individual establishment 
results. Furthermore, the average of the calcium results in the 2002 
Beef AMR Survey was below 100 mg/100 g, but again, there was wide 
variation in individual results.
    FSIS is clarifying in this interim final rule that no analysis can 
exceed the regulatory maximum of 130 mg/100 g sample. This level of 
calcium in the product does not affect the appearance, texture, or 
other quality aspects of the product and is a small amount of calcium 
when compared to the calcium content generally contained in 
MS(Species).
    In deciding on a calcium level, FSIS understands that it is 
virtually impossible for calcium levels in AMR product to be equal to 
those of hand-deboned product, which is essentially 0 mg/100 g. The 
presence of small amounts of calcium does not affect the qualitative 
characteristics of the product and only trivially affect its 
compositional aspects. Thus the standard will ensure that AMR product 
is ``meat.'' In addition, this standard creates a clear distinction 
between AMR product and MS(Species) product, which generally has more 
than triple the calcium of AMR. At the same time, FSIS has tried not to 
set such a low level for calcium that it would not be economically 
feasible to produce AMR product.
    Comment: A commenter thought that calcium samples should be taken 
at the intermediate stage of the AMR process, because at this stage the 
calcium samples would indicate whether bones are being broken or 
crushed.
    Response: FSIS is only concerned about the levels of calcium in the 
final AMR product as a means of ensuring that an excess amount of bone 
solids is not introduced into the product. It is not using a calcium 
measurement level to determine if bones are broken or crushed. Thus, 
FSIS is not including a standard to measure calcium at an intermediate 
stage in the AMR process in this interim final rule.

Bone Marrow

    Comment: Commenters stated that the methodology and data used to 
derive the iron criterion that was proposed as a measure for 
noncomplying product were incorrect, and that, therefore, the proposed 
values were not appropriate. Specifically, it was pointed out that the 
analytical procedures used in the FSIS 1996 survey were based on 
procedures that understated iron values. Further, a commenter disagreed 
with the Agency's

[[Page 1879]]

approach of correlating histological data and the bone marrow cell 
assessment, with iron content. The commenter claimed that the 
correlation was not high, and thus was not accurate.
    A commenter agreed that a measurement of total iron is a good 
indicator of the presence of marrow in meat and further claimed that 
the amount of iron in beef is well established. However, there were 
many comments that questioned both using excess iron as a measure of 
bone marrow and the methodology used to establish the limit in the 
standard. A commenter suggested not using protein at all in adjusting 
the iron requirement but, rather, using a straight iron value level. A 
commenter suggested that FSIS needs to account for the fact that AMR 
procedures remove connective tissue that contains little or no iron, 
and that muscle adjacent to the bone is higher in iron than is hand-
deboned muscle. Therefore, even if marrow components were absent, iron-
to-protein ratios (IPRs) would be higher in AMR products than those in 
hand-deboned meat.
    Another commenter claimed that the use of iron as proposed by the 
Agency would be biased against low fat, high protein products and 
suggested a simple IPR. Some commenters said that the iron levels 
established were too high and urged FSIS to make the target levels more 
consistent with hand-deboned product. These commenters suggested a 5 to 
10 percent variation in the IPR between AMR and hand-deboned meat. 
Commenters also suggested that establishments should not be permitted 
to determine their own IPR values, as was proposed.
    Response: FSIS will first address the measurement and methodology 
issue and then provide a justification for the excess iron measure it 
proposed. In the course of doing so, it will provide an explanation for 
the procedures that it used for deriving the iron performance standard 
contained in this interim final rule.
    Excess iron is the iron in excess of that which would be expected 
given the protein value if the product was meat. The measure for excess 
iron for the 2002 survey was: excFe=Fe-kP, where P is the protein (%), 
Fe is the iron (mg per 100 g), and k is a constant equal to 1.1 times 
0.138. The 0.138 is the assumed IPR for the corresponding hand-deboned 
meat product, and the 1.1 is an adjustment factor.
    Measurement and methodology. While the measurement used by FSIS was 
accurate, the Agency agrees that the methodology and measurement 
procedures used in developing the standards for iron in the 1998 
proposed rule were not consistent with common laboratory analyses for 
iron measurement. FSIS used a hydrochloric acid wet-ash digestion 
procedure to measure the iron levels of samples collected in the 1996 
survey because this methodology was considered faster and less labor 
intensive than traditional dry-ash procedures (i.e., dry-ash procedure 
for digestion). The wet-ash procedure predictably underestimates the 
true level of iron. In contrast, the method used by ARS scientists, 
which is based on a dry-ash procedure for digestion, dries the samples 
and obtains iron results approximately double those obtained by the 
FSIS procedure. Further, the results obtained by the ARS dry-ash 
procedure are more consistent with levels previously reported for hand-
deboned product in Agricultural Handbook 8 (now called USDA Nutrient 
Database for Standard Reference, Release 12).
    ARS analyzed split samples from the 1996 survey for FSIS, and FSIS 
used the ARS results along with more current FSIS data for deriving the 
standards for iron in this interim final rule. For samples in which 
there were no dry-ash procedure results, the FSIS wet-ash procedure 
results were multiplied by 2.11, which is the average ratio of the 
results from the dry-ash procedure to those that FSIS found using the 
hydrochloric acid wet-ash procedure (See the technical addendum for 
additional information in the FSIS docket room and on the web site).\5\
    FSIS agrees with the commenter's concern about FSIS'' approach of 
correlating histological data and bone marrow cells with iron content 
and thus is not including a standard for bone marrow cells in this 
interim final rule. Although bone marrow cells are unique to bone 
marrow, they have been found in hand-deboned product probably as a 
consequence of contamination of the muscle tissue during the carcass 
splitting process during slaughter.
    FSIS justification for using excess iron as a measure of bone 
marrow. FSIS has determined that there is no practical methodology to 
measure bone marrow using commercial practices. Bone marrow contains 
many of the same components as muscle tissue and blood. Therefore, FSIS 
sought to establish in the 1998 proposal a practical methodology that 
would predict whether the known composition of hand-deboned meat was 
sufficiently different from AMR as a consequence of the incorporation 
of bone content (other than calcium) in AMR. FSIS deemed this 
additional bone content to be an indication of the presence of bone 
marrow. Consequently, iron, which is contained in marrow and in blood 
tissue, was chosen as a practical surrogate for bone marrow.
    To determine whether there were excess iron levels in AMR, and thus 
bone marrow in this product, the Agency proposed using an adjustment 
based on the protein value because an analysis of the data from a prior 
survey demonstrated that there was a correlation between iron and 
protein results. Protein levels will change with iron levels, 
everything else being equal. If bone marrow, which has a higher IPR 
value than meat, is added to product, the measured IPR value would be 
greater than the IPR for corresponding hand-deboned product without 
bone marrow. Accounting for measurement error, if this difference is 
large enough, it can then be concluded that bone marrow at more than a 
negligible amount is in the product.
    One of the commenters pointed out that a problem with the above 
model is that the AMR process removes connective tissue that contains 
little or no iron. The Agency believes that the effect of this removal 
is not large and would not change the basic premise of the model 
presented above. From the 1996 FSIS survey, the Agency determined that 
the average difference in protein between pre- and post-desinewed AMR 
product was about 0.5 percent, based on a post-desinewed product 
average protein of about 16.5 percent. Therefore, as a percentage of 
protein, the amount of protein associated with connective tissue 
removed during the desinewing step averaged only about 3 percent and 
does not represent a large proportion of the protein that is in the 
final product.
    In addition, it is possible that, during AMR processing, some 
unbound water is removed which would result in the removal of some 
water-soluble protein and dissolved solids.\6\
    FSIS recognizes that these two factors, removal of connective 
tissue with low iron and protein and removal of unbound water, may 
result in an increase in the IPRs of AMR product. However, FSIS does 
not believe that such a possible increase renders the use of an excess 
iron measurement inaccurate for assessing AMR process control. Although 
FSIS does not believe that the effects of these factors would be 
substantial, it has taken them into consideration in this interim final 
rule and is using a 10 percent factor for adjusting the protein levels 
used for calculating levels of excess iron in AMR product.
    Another issue raised by the commenters regarding the 
appropriateness of the excess iron

[[Page 1880]]

measurement was that meat close to the bone has higher IPRs than meat 
farther from the bone. FSIS agrees with the commenter. However, the 
IPRs would be expected to be higher in AMR product than in hand-deboned 
product, even though no bone marrow would be introduced.
    FSIS has decided to allow alternative IPRs to be used in this 
interim final rule to reflect the inherent differences that exist among 
starting products.
    Regarding the comment made that the use of the excessive iron 
measure as proposed would be biased against high protein and low fat 
products, FSIS believes that for practical purposes, the difference 
between the excessive iron and the IPR calculations is not great.
    In this interim final rule, however, FSIS is adopting a different 
excess iron limit measurement than the one proposed in 1998. This new 
limit is based on a more current examination of excess iron 
measurements for hand-deboned product from the 2002 survey of AMR 
product. See footnote 1 in new Sec.  318.24(c)(1)(ii) for a detailed 
explanation of the formula derived for the excess iron value 
measurement.
    An assumption used by FSIS in the derivation of the excess iron 
value measurement for this interim final rule was that there would be 
duplicate measurements of iron and protein taken by establishments on 
an individual sample. Performing duplicate measurements on an 
individual sample is recommended because, on a few occasions in the 
2002 survey, large differences for samples were found when duplicate 
measurements were made. Thus, to ensure that AMR product is consistent 
with meat, FSIS is adopting a measured 3.5 mg/100 g excess iron limit 
based on duplicate analyses of samples of AMR product.

Related Comments

    Comment: Several commenters alleged that FSIS has singled out AMR 
technology for scrutiny while products derived from a low temperature 
rendering process (LTRP) were approved by FSIS for the school lunch 
program without any scientific basis or public input. The suggestion 
was made that FSIS withdraw the proposed rule on AMR products until 
comparable rules to regulate LTRP products have been developed and 
implemented.
    Response: The Agency has focused on meat produced by AMR systems 
because it is the main product not produced by hand-deboning, and is a 
product in which constituents not expected in boneless meat can be 
incorporated as a result of the process used for its production. Other 
technologies, such as LTRP, generally involve the removal of components 
such as fat and muscle. The Agency intends to further evaluate how it 
regulates other types of operations that are used to manufacture meat 
and poultry trimmings from various starting materials. The Agency seeks 
more specific comment and data on the compositional characteristics of 
LTRP and similar products derived from non-AMR systems.
    Comment: A commenter said the proposal was based on an antiquated 
regulatory foundation because the definition of meat is obsolete and 
is, in effect, an anatomical description. In addition, the commenter 
maintained that the proposal was an attempt to relate a chemical 
constituent of AMR-derived product to the former USDA Handbook 8 
references for regulatory purposes and conflicted with Agency policies 
regarding constituents of other meat products.
    Response: Meat is defined in anatomical terms, and not chemically, 
because it is directly obtained from livestock and not chemically 
derived from other elements. Therefore, the regulatory definition of 
meat refers to the parts of livestock that are edible (as opposed to 
inedible parts/organs). The former Handbook 8 details the composition 
of foods but does not represent a formula for making ``meat.'' FSIS is 
not relating a constituent of AMR product to former Handbook 8 data on 
the composition of meat. AMR product is meat unless it includes 
constituents such as spinal cord and DRG that are not expected 
constituents of boneless meat. In addition, FSIS has determined that 
AMR product is meat unless the process by which it is produced 
incorporates expected constituents, such as calcium and iron, at 
excessive levels.
    Comment: A commenter asked about FSIS' response to the report on 
AMR technology and on worker safety issues related to AMR systems.\7\
    Response: Regarding the report, which was produced by the 
Georgetown University Center for Food and Nutritional Policy, FSIS 
generally agrees with the historical and technical aspects of the 
report on AMR systems. The report addressed the disagreements that have 
characterized the regulated introduction of mechanical deboning in this 
country, and how these initiatives have attracted the attention of 
consumer advocacy groups. The 1999 report states that the presence of 
CNS tissue in meats of any kind should be avoided and cited FSIS' 
prohibition against spinal cord in AMR meat since 1997.
    The report discussed the reduction in worker-related injuries as 
perhaps the greatest societal advantage of AMR systems. FSIS agrees 
that manual deboning and the use of motorized knives are dangerous 
because they are associated with direct injuries and cumulative trauma 
disorders (CTDs). The report noted that some studies have demonstrated 
a 38 percent increase in CTDs as a consequence of working in deboning 
operations.
    FSIS agrees with the statements in the report about the efficiency 
of AMR systems that makes meat processing operations more safe and 
profitable. However, for the reasons presented in this interim final 
rule, the Agency disagrees with the Sparks report's assertion that 
further rulemaking to refine the 1994 final rule is unwarranted.
    Comment: A commenter asked whether FSIS agreed with the cost 
estimates in the Sparks Companies, Inc., report, which provided an 
economic analysis of the 1998 proposed AMR rule.\8\
    Response: FSIS does not agree with some of the conclusions in the 
Sparks report. For example, FSIS believes that it is unlikely that all 
AMR systems will be removed and replaced with tertiary hand-deboning 
procedures, as the report suggests. Not all of the AMR systems are used 
to process split vertebral columns with exposed and extruding bone 
marrow tissue. Some systems are used to process only brisket or sternum 
and rib bones. The expected continued use of non-vertebral bones in AMR 
systems would considerably reduce the capital cost loss of $40 million 
estimated in the report.
    The report's discussion of capital costs also fails to take into 
account depreciation of the AMR systems since 1994, which would 
considerably reduce the capital cost loss. In addition, the cost of 
auto-knives may be somewhat over-estimated because the report assumes 
that the knives depreciate within a year. FSIS would suggest that the 
authors of the report should have used only the flow of services of the 
knives, not the depreciation of the entire capital stock of the knives 
within a year.
    However, the report was helpful and provided the Agency with 
important data to gauge volume and yield data, for example, and to gain 
a greater understanding of the extent of the AMR beef and pork industry 
in this country.
    These comments and all of the other public comments submitted in 
response to the 1998 proposal are available for review in the FSIS 
Docket Room and at the FSIS Web site.

[[Page 1881]]

Consumer Group Petition

    Because of its concerns about the presence of spinal cord and DRG 
in AMR product, in 2001, a consumer group, the Center for Science in 
the Public Interest (CSPI) on behalf of other consumer and public 
health associations, petitioned USDA to institute regulatory actions to 
prohibit spinal cord and DRG in AMR beef products.\9\ In addition, a 
consortium of 14 animal welfare, farmer, environmental, and public 
health groups voiced similar concerns and urged USDA and the FDA to 
take immediate regulatory action.\10\

2002 Survey of AMR Products

    In order to assess the current industry practices associated with 
AMR systems, the petition submitted by CSPI, and the need for further 
Agency action with regard to AMR, the Agency determined that it needed 
to conduct a survey of AMR systems (i.e., the 2002 Survey of AMR 
Products). Another purpose of this survey was to characterize the 
recovered product of AMR systems regarding texture and appearance, look 
at current production practices (e.g., pressure settings and type of 
source materials) and yield data, and determine how those practices 
influence the calcium and iron levels of the final product.
    In January 2002, FSIS began collecting random samples from the 42 
piston-driven AMR systems in production at 34 establishments harvesting 
AMR product derived from beef vertebrae or beef vertebrae mixed with 
other types of beef bones. Several establishments had more than one 
operating AMR system processing beef vertebrae.
    Over a 7-month period, samples from each AMR system that uses beef 
vertebrae as source material were randomly collected. An FSIS 
laboratory tested the products for the presence of spinal cord and DRG. 
At random times over the 7-month period, FSIS collected final (after 
the desinewer) product samples and intermediate (before the desinewer) 
samples from each of the active machines. In addition, the AMR system 
model and identification number, type of starter (input) product, and 
the maximum pressure applied and pressure hold or dwell time (at the 
maximum pressure) of the systems were noted. Most of the samples also 
were tested for the food chemistry constituents calcium, iron, and 
protein.
    Although some of the establishments (4 of 34 or 12 percent) were 
able to produce final AMR product with no spinal cord or DRG on a 
consistent basis (based on all (six or more) samples being negative), 
other establishments consistently produced samples that tested positive 
for spinal cord and DRG. For the survey, approximately 35 percent of 
the final AMR product samples tested positive for spinal cord or DRG: 
29 percent for spinal cord and 10 percent for DRG.
    The occurrence of spinal cord and DRG was not considered to be 
significantly correlated; that is, the presence of one of these tissues 
in a sample did not significantly affect the likelihood of the presence 
of the other. This lack of significant correlation suggests that there 
may be different factors that determine the presence of these tissues 
in AMR product. On the other hand, estimated values of excess iron and 
calcium were positively correlated, suggesting that there is a common 
set of factors that influence their levels. See the final report on the 
2002 survey results in the FSIS Docket Room or at the FSIS web site for 
additional details.\11\

FSIS Directive 7160.3

    In August 2003, FSIS issued Directive 7160.3, Revision 1, to 
provide instructions to inspection program personnel for sampling 
boneless comminuted beef products from AMR systems in which vertebral 
columns are used and on actions to take if the product contains spinal 
cord.\12\ The directive did not address the presence of DRG tissue in 
AMR product because the Agency had not included DRG in the 1998 
proposed rule.
    After doing follow-up verification sampling, the Agency was 
especially concerned that some establishments were not adequately 
addressing the problem of spinal cord in AMR product. The directive 
defined the range of follow-up actions available to the Agency when 
product from an AMR system is found to contain spinal cord tissue. FSIS 
withheld label approval for those establishments whose AMR system 
repeatedly failed to produce product that was free of spinal cord. 
Thus, these establishments effectively were not allowed to produce AMR 
meat from beef vertebrae.

Overview of This Interim Final Rule and Request for Comments

    FSIS is amending the meat inspection regulations in Parts 301, 318, 
and 320 of the Code of Federal Regulations by modifying the definition 
of ``meat;'' adding or modifying non-compliance criteria for bone 
solids, bone marrow, brain, trigeminal ganglia, spinal cord, and DRG; 
requiring the development, implementation, and maintenance of a written 
program, including documentation and recordkeeping requirements, for 
ensuring process control; and declaring inedible the skulls and 
vertebral column bones from cattle that are 30 months of age and older. 
As indicated in a new Section 310.22, which is adopted in another 
interim final rule issued today (see Docket 03-025IF in this 
issue of the Federal Register), skulls and vertebral column bones from 
cattle 30 months of age and older are inedible and cannot be used for 
human food. Therefore, if skulls or vertebral column bones from cattle 
30 months of age and older are used in AMR systems, the product exiting 
the AMR system is adulterated, and the product and the spent bone 
materials are inedible and cannot be use used for human food. For AMR 
product derived from the bones of cattle younger than 30 months, the 
presence of CNS-type tissues will render the product misbranded. 
Similarly, for AMR product derived from the bones of livestock other 
than cattle, the presence of CNS-type tissues will result in 
misbranding. For AMR product derived from the bones of all livestock, 
the restrictions associated with bone solids and bone marrow also 
relate to misbranding.
    FSIS is amending Sec.  301.2(b), the definition of ``meat'' to make 
it clear that boneless meat may not include significant portions of 
bone or related components, such as bone marrow, or any amount of CNS-
type tissues. Therefore, product produced using an AMR system must not 
include significant amounts of bone or related components. It also must 
not include any brain, trigeminal ganglia, spinal cord, or DRG.
    Section 318.24(a) provides that skulls and vertebral column bones 
of cattle 30 months of age and older, as provided for in a new section 
310.22 which is adopted in another interim final rule issued today (See 
Docket 03-025IF in this issue of the Federal Register), cannot 
be used in AMR systems. In addition, the recovered meat product exiting 
the AMR system must not significantly incorporate bone solids or bone 
marrow, as measured by the presence of calcium and excess iron, and 
cannot contain any brain, trigeminal ganglia, spinal cord, or DRG.
    Section 318.24(b) provides that establishments operating AMR 
systems are required to develop, implement, and maintain procedures 
that ensure that their production process is in control. The 
establishment must incorporate its production process procedures in a 
written program that is designed to ensure the ongoing effectiveness of 
the process control program. Because of the food safety concerns 
presented by SRMs, for establishments that process

[[Page 1882]]

cattle, the written program must be in the establishment's Hazard 
Analysis and Critical Control Point (HACCP) plan, or in its Sanitation 
Standard Operating Procedure (Sanitation SOP) or other prerequisite 
program.
    By declaring SRMs inedible and prohibiting their use for human 
food, FSIS will ensure that materials that could present a significant 
risk to human health, but whose infectivity status cannot be readily 
ascertained, are excluded from the human food supply.
    Because BSE was recently confirmed in a cow in the United States, 
FSIS has determined that the SRMs, adopted in another interim final 
rule issued today (see Docket 03-025IF in this issue of the 
Federal Register), are unfit for human food. Thus, the status of these 
materials has changed from edible to inedible. Such a change is likely 
to affect the underlying hazard analysis that must be conducted as 
prescribed by 9 CFR 417.4(a)(3). Therefore, in response to this change, 
FSIS expects that establishments that slaughter cattle or process 
carcasses or parts of cattle will reassess their HACCP plans in 
accordance with 9 CFR 417.4(a)(3) to address SRMs.
    Under Sec.  318.24(b), the written program must include the 
observation of bones entering the AMR system and the testing of the 
product exiting the AMR system. The establishment shall maintain 
records on a daily basis sufficient to document the implementation and 
verification of its production process. The establishment shall make 
the documentation available to inspection program personnel.
    Section 318.24(b) makes clear that establishments will be expected 
to determine how and when they will test product for calcium, iron, 
spinal cord, and DRG. Based on the supporting documentation provided by 
the establishment, and FSIS's own verification, FSIS will make a 
determination whether the product is misbranded or adulterated. FSIS 
expects that the establishment will ensure that each production lot is 
in compliance with the provisions of this regulation.
    Regarding the testing methodology for spinal cord and DRG, FSIS 
will continue to use its validated histological procedures. However, 
FSIS is aware that establishments have access to methodology that is 
not as specific or sensitive as the FSIS methodology and that is 
considerably less expensive to perform. FSIS encourages establishments 
to use any methodology that is effective. FSIS cautions establishments, 
however, that if the establishment's methodology is not adequate to 
discern complying product from non-complying product, FSIS will ensure 
that non-complying product is not allowed to enter commerce.
    Because of the expense and time associated with highly sensitive 
and specific tests, such as the methodology used by FSIS, researchers 
have been working on quicker and less costly tests. One such research 
effort has employed ELISA technology. For the 2002 AMR beef survey, an 
ELISA procedure was examined by FSIS, but FSIS concluded that the test 
was not sufficiently specific or sensitive. Not only were there many 
false positive and negative results (when compared to the FSIS 
histological results), the rates of false positive and negative results 
were establishment dependent. This latter finding could imply that 
there was some other component in the product interfering with the 
test.
    FSIS is aware that there are a number of research efforts underway 
to improve the sensitivity and specificity of the rapid tests that can 
be used in lieu of the normative histological tests for evaluating the 
presence of spinal cord and DRG. FSIS does not want to preclude the use 
of such tests by establishments. Therefore, FSIS is soliciting 
information during the comment period on alternative test methods and 
performance specificity and sensitivity. FSIS is interested in 
identifying a test for use by establishments that is as sensitive to 
the presence of spinal cord and DRG in product as the histological test 
employed by FSIS, but that is less expensive and less time consuming.
    The production process is not in control if the skulls of livestock 
entering the AMR system contain any brain or trigeminal ganglia tissue, 
or the vertebral column entering the AMR system has any spinal cord. In 
addition, the process is not in control if the recovered product 
contains unacceptable levels of bone solids or bone marrow, or any 
level of spinal cord or DRG, as provided for in Sec. 318.24(c). In 
addition, the production process is not in control if the product is 
not properly labeled or spent bone materials are not properly handled.
    Section 318.24(c)(1) describes the five criteria that define when 
recovered AMR product may not be used and labeled as ``meat.'' They 
include a measure for excess bone solids (calcium content above the 
stated level); a measure for excess bone marrow (iron in relation to 
protein above the stated level); the presence of brain or trigeminal 
ganglia; the presence of spinal cord; and the presence of DRG.
    In Sec. 318.24(c)(2), if the recovered product derived from any 
livestock fails under any of these criteria, it cannot be labeled as 
``meat.'' In addition, product derived from beef skulls or vertebral 
column bones from cattle younger than 30 months containing CNS-type 
tissues cannot be used as an ingredient of a meat food product. For 
example, this product, if it contained spinal cord, cannot be labeled 
as ``Beef with Spinal Cord'' or ``Beef with Spinal Cord Meat Food 
Product'' because detached spinal cord is prohibited from use in the 
preparation of edible product other than for edible rendering (9 CFR 
318.6(b)(4)). It also cannot be labeled as MS(Beef) because FSIS has 
determined MS(Beef) to be inedible and prohibited its use as human food 
(see Docket 03-025IF in this issue of the Federal Register. 
Such product can be rendered to produce products identified as beef 
stock, beef extract, and beef flavoring without any identification of 
the source materials other than ``beef'' because the source materials 
are edible, not inedible. FSIS has determined that it is appropriate to 
now prohibit product that contains CNS-type tissues derived from cattle 
younger than 30 months of age for use in a meat food product, except 
for the sale of brain or the use of brain in which its presence is 
required to be reflected prominently and conspicuously in labeling. 
FSIS has established precedent for not allowing detached spinal cord 
for use in meat food products, but does allow its use for edible 
rendering. FSIS requests comment on whether product derived from the 
bones of cattle younger than 30 months (as well as product from 
livestock other than cattle) that may contain CNS-type tissues should 
continue to be allowed in edible rendering, or whether such product 
should be inedible and not allowed in edible rendering or allowed in 
descriptively labeled meat food product. FSIS requests comment on 
whether edible rendered products derived from bones of livestock in 
which the bones may contain CNS-type tissues should be required to bear 
a common or usual name that reflects the potential presence of CNS-
tissue (e.g., ``beef stock derived from materials that may contain 
spinal cord''). FSIS will be working with FDA on this issue.
    As discussed above, skulls or vertebral column bones from cattle 30 
months of age and older may not be used at all in AMR systems. Product 
derived from bones of cattle other than skulls or vertebral column 
bones may bear a name that is not false or misleading but cannot bear 
the name ``Mechanically Separated (Beef).'' In another interim final 
rule issued today (see Docket 03-025IF in this issue of

[[Page 1883]]

the Federal Register), FSIS has determined that MS(Beef) is inedible 
and prohibited its use as human food. Such product would not contain 
CNS-type tissues because only the skulls and vertebral column bones 
contain CNS-type tissues.
    For purposes of this rule, bone marrow from cattle is not 
identified as an SRM. The scientific evidence to establish that cattle 
bone marrow is a tissue that demonstrates infectivity is inconclusive 
at this time (see Docket No. 03-025IF, also published in this issue of 
the Federal Register for additional information about bone marrow). 
Therefore, product from cattle of any age (e.g., through the use of AMR 
systems using long bones rather than vertebral column bones) that fails 
to meet the bone marrow standard is misbranded. FSIS seeks comment on 
this issue.
    Section 318.24(c)(3) provides that spent skulls and vertebral 
column bone materials from cattle eligible to enter an AMR system 
(i.e., from cattle younger than 30 months of age) are eligible for 
edible rendering, as is the product derived from these bones that 
contains CNS-type tissues (see Sec. 318.24 (c)(2)(i) or (ii).
    Although some non-complying AMR product derived from the vertebral 
column of pork and livestock other than cattle may be diverted to use 
as MS(Species), such a practice has not been customary in the past 
because MS(Species) rarely, if ever, is produced in the United States. 
FSIS is considering rulemaking on MS(Species) from species other than 
cattle regarding the presence of CNS-type tissue in this product and is 
seeking comment on this issue.
    Section 320.1 is amended to extend the recordkeeping requirements 
to the entire AMR process control system. The current regulation 
applies only to the calcium criteria. This change is necessary to 
ensure that establishments maintain appropriate records documenting 
that they are controlling the entire process, including the appropriate 
identification and segregation of cattle and their derived products. 
The establishment may determine to incorporate the control procedures 
and recordkeeping into their HACCP plan or into their Sanitation SOP or 
other prerequisite program. Such control procedures may be based on the 
guidance prepared by the Canadian government for their industry.

Request for Comments

    FSIS requests comments on the measures contained in this interim 
final rule, and specifically on whether the Agency has chosen measures 
that are most appropriate for preventing human exposure to the BSE 
agent in the United States.

Emergency Action

    Given the fact that a cow in Washington State tested positive for 
BSE on December 23, 2003, it is necessary to issue this rule on an 
emergency basis. BSE infectivity has been confirmed in the brain, eyes, 
trigeminal ganglia, tonsils, spinal cord, DRG, and distal ileum. 
Furthermore, most of these tissues have demonstrated infectivity before 
experimentally infected animals developed clinical signs of disease. 
Thus, BSE infectivity in these tissues is not readily ascertainable. 
Therefore, FSIS has determined that it must take immediate action to 
ensure that materials that could present a significant risk to human 
health in beef derived from AMR systems and the spent bone materials 
derived from AMR systems are excluded from the human food supply.
    Under these circumstances, the FSIS Administrator has determined 
that prior notice and opportunity for public comment are contrary to 
the public interest, and that there is good cause under 5 U.S.C. 553 
for making this rule effective less than 30 days after publication in 
the Federal Register. FSIS will consider comments received during the 
comment period for this interim rule (see DATES above). After the 
comment period closes, the Agency will publish another document in the 
Federal Register. The document will include a discussion of any 
comments received in response to this interim rule and any amendments 
made as a result of those comments.
    In an effort to ensure that establishments comply with this interim 
final rule upon publication in the Federal Register, FSIS will provide 
guidance to inspection program personnel regarding the implementation 
strategy. At a minimum, FSIS inspection program personnel will be 
directed to meet with management of each affected establishment to 
discuss how and when the establishment expects to complete its 
reassessment of its HAACP plan to ensure that SRMs and MS(Beef) do not 
adulterate product.

Executive Order 12866 and the Regulatory Flexibility Act

    This rule has been reviewed under Executive Order 12866. It has 
been determined to be economically significant for purposes of E.O. 
12866.
    The emergency situation surrounding this rulemaking makes timely 
compliance with Executive Order 12866 and the Regulatory Flexibility 
Act (5. U.S.C. 601 et seq.) impracticable.
    FSIS is currently assessing the potential economic effects of this 
action. When this work is complete, the Agency will publish a notice of 
availability in the Federal Register and will provide an opportunity 
for public comment.

Executive Order 12988

    This interim final rule has been reviewed under Executive Order 
12988, Civil Justice Reform. This rule: (1) Preempts State and local 
laws and regulations that are inconsistent with this rule; (2) has no 
retroactive effect; and (3) does not require administrative proceedings 
before parties may file suit in court challenging this rule. However, 
the administrative procedures specified in 9 CFR 306.5. must be 
exhausted before any judicial challenge of the application of the 
provisions of this interim final rule, if the challenge involves any 
decision of an FSIS employee relating to inspection services provided 
under the FMIA or PPIA.

Paperwork Reduction Act

    In accordance with section 3507(j) of the Paperwork Reduction Act 
of 1995 (44 U.S.C. 3501 et seq.), the information collection and 
recordkeeping requirements included in this interim final rule have 
been submitted for emergency approval to the Office of Management and 
Budget (OMB). OMB has assigned control number 0583-XXXX to the 
information and recordkeeping requirements.
    Title: Advanced Meat Recovery Systems.
    Type of collection: New.
    Abstract: FSIS has reviewed the paperwork and recordkeeping 
requirements in this interim final rule in accordance with the 
Paperwork Reduction Act. Under this interim final rule, FSIS is 
requiring a new information collection activity. FSIS is requiring 
establishments that produce meat from AMR systems to ensure that bones 
used for AMR systems do not contain brain, trigeminal ganglia, or 
spinal cord, to test for calcium (at a different level than previously 
required), iron, protein, spinal cord, and DRG, to document their 
testing protocols, to assess the age of cattle product used in the AMR 
system, and to document their procedures for handling product from 
cattle of any age in a manner that does not cause product to be 
misbranded or adulterated, and to maintain records of their 
documentation and test results.
    Estimate of burden: FSIS estimates that it will take establishments 
on a daily basis 30 minutes to collect the

[[Page 1884]]

information such as for calcium and iron and 30 minutes to sample for 
spinal cord and DRG. The Agency estimates that it will take 2 minutes 
to do recordkeeping of test results. FSIS also estimates that it will 
take establishments 2 hours to develop their testing protocols.
    Respondents: Establishments that produce livestock product (e.g., 
beef and pork) from AMR systems.
    Estimated Number of Respondents: 56.
    Estimated Number of Responses per Respondent: 1,201.
    Estimated Total Annual Burden on Respondents: 18,088 hours.
    Copies of this information collection assessment can be obtained 
from John O'Connell, Paperwork Reduction Act Coordinator, FSIS, USDA, 
112 Annex, 300 12th Street, SW., Washington, DC 20250-3700.

Additional Public Notification

    Public involvement in all segments of rulemaking and policy 
development is important. Consequently, in an effort to better ensure 
that minorities, women, and persons with disabilities are aware of this 
interim final rule and informed about the mechanism for providing their 
comments, FSIS will announce it and make copies of this Federal 
Register publication through the FSIS Constituent Update, which is 
communicated via Listserv, a free e-mail subscription service. In 
addition, the update is available online through the FSIS web page 
located at http://www.fsis.usda.gov. The update is used to provide 
information regarding FSIS policies, procedures, regulations, Federal 
Register notices, FSIS public meetings, recalls, and any other types of 
information that could affect or would be of interest to our 
constituents and stakeholders. The constituent Listserv consists of 
industry, trade, and farm groups, consumer interest groups, allied 
health professionals, scientific professionals, and other persons who 
have requested to be included. Through the Listserv and web page, FSIS 
is able to provide information to a much broader, more diverse 
audience.
    For more information, contact the Congressional and Public Affairs 
Office, at (202) 720-9113. To be added to the free e-mail subscription 
service (Listserv) go to the ``Constituent Update'' page on the FSIS 
Web site at http://www.fsis.usda.gov/oa/update.htm. Click on the 
``Subscribe to the Constituent Update Listserv'' link, then fill out 
and submit the form.

Footnotes

    The following sources are referred to in this document and are 
available for review in the FSIS Docket Room (See ADDRESSES above) 
between 8:30 a.m. and 4 p.m., Monday through Friday.
    1. Harvard Center for Risk Analysis, Harvard School of Public 
Health, and Center for Computations Epidemiology, College of 
Veterinary Medicine, Tuskegee University, November 2001. Evaluation 
of the Potential for Bovine Spongiform Encephalopathy in the United 
States.
    2. Summary of Calendar Year 2003 AMR Testing, FSIS.
    3. Hasiak, R.J. and H. Marks, The ``Advanced Meat Recovery 
System'' Survey Project Final Report, February 21, 1997.
    4. FSIS Directive 7160.2, ``Meat'' Prepared Using Advanced 
Mechanical Meat/Bone Separation Machinery and Meat Recovery Systems, 
April 14, 1997.
    5. FSIS technical paper, Derivation of excess iron limits for 
meat products produced by Advanced Recovery Systems, July 21, 1999.
    6. Wyndom, W.R. and R.A. Field, Effect of method of analysis on 
iron content of beef from advanced meat recovery systems, May 2000.
    7. Georgetown University Center for Food & Nutritional Policy, 
Advanced Meat Recovery Systems, 1999.
    8. Sparks Companies, Inc., Advanced Meat Recovery Systems--An 
Economic Analysis of Proposed USDA Regulations, July 1999.
    9. Letter to FDA and USDA, submitted by Public Citizen, and 
signed by the Animal Welfare Institute. Cancer Prevention Coalition, 
Center for Food Safety, Community Nutrition Institute, Family Farm 
Defenders, Farm Sanctuary, Global Resource Action Center for the 
Environment, Government Accountability Project, Project Humane 
Farming Association, Institute for Agriculture and Trade Policy, 
National Family Farm Coalition, Organic Consumers Association, 
Public Citizen, and the U.S. Public Interest Research Group, April 
13, 2001.
    10. Petition for Regulatory Action to Bar the Use of Spinal Cord 
and Columns and Other Potentially Infectious Tissue from Beef in the 
Human Food Supply, submitted by the Center for Science in the Public 
Interest, on behalf of the American Public Health Association, 
Consumer Federation of America, Government Accountability Project, 
National Consumers League, and Safe Tables Our Priority, August 9, 
2001.
    11. Analysis of 2002 FSIS Bovine AMR Survey Results, prepared by 
the USDA, FSIS, February 2003.
    12. FSIS Directive 7160.3, Revision 1, Advanced Meat Recovery 
Using Beef Vertebral Raw Materials, August 25, 2003.

List of Subjects

9 CFR Part 301

    Meat and meat products.

9 CFR Part 318

    Meat inspection, Records.

9 CFR Part 320

    Meat inspection, Records.

0
For the reasons set forth above, FSIS is amending 9 CFR, chapter III, 
as follows:

PART 301--TERMINOLOGY

0
1. The authority citation for part 301 continues to read as follows:

    Authority: 7 U.S.C. 450, 1901-1906; 21 U.S.C. 601-695; 7 CFR 
2.18, 2.53.


0
2. In Sec.  301.2, the definition of ``Meat'' is revised to read as 
follows:


Sec.  301.2  Definitions.

* * * * *
    Meat. (1) The part of the muscle of any cattle, sheep, swine, or 
goats which is skeletal or which is found in the tongue, diaphragm, 
heart, or esophagus, with or without the accompanying and overlying 
fat, and the portions of bone (in bone-in product such as T-bone or 
porterhouse steak), skin, sinew, nerve, and blood vessels which 
normally accompany the muscle tissue and that are not separated from it 
in the process of dressing. As applied to products of equines, this 
term has a comparable meaning.
    (i) Meat does not include the muscle found in the lips, snout, or 
ears.
    (ii) Meat may not include significant portions of bone, including 
hard bone and related components, such as bone marrow, or any amount of 
brain, trigeminal ganglia, spinal cord, or dorsal root ganglia (DRG).
* * * * *

PART 318--ENTRY INTO OFFICIAL ESTABLISHMENTS; REINSPECTION AND 
PREPARATION OF PRODUCTS

0
3. The authority citation for part 318 continues to read as follows:

    Authority: 7 U.S.C. 138f, 450, 1901-1906; 21 U.S.C. 601-695; 7 
CFR 2.7, 2.18, and 2.53.


0
4. Section 318.24 is revised to read as follows:


Sec.  318.24  Product prepared using advanced meat/bone separation 
machinery; process control.

    (a) General. Meat, as defined in Sec.  301.2 of this subchapter, 
may be derived by mechanically separating skeletal muscle tissue from 
the bones of livestock, other than skulls or vertebral column bones of 
cattle 30 months of age and older as provided in Sec.  310.22 of this 
subchapter, using advances in mechanical meat/bone separation machinery 
(i.e., AMR systems) that, in accordance with this section, recover 
meat--
    (1) Without significant incorporation of bone solids or bone marrow 
as measured by the presence of calcium and iron in excess of the 
requirements in this section, and
    (2) Without the presence of any brain, trigeminal ganglia, spinal 
cord, or dorsal root ganglia (DRG).

[[Page 1885]]

    (b) Process control. As a prerequisite to labeling or using product 
as meat derived by the mechanical separation of skeletal muscle tissue 
from livestock bones, the operator of an establishment must develop, 
implement, and maintain procedures that ensure that the establishment's 
production process is in control.
    (1) The production process is not in control if the skulls entering 
the AMR system contain any brain or trigeminal ganglia tissue, if the 
vertebral column bones entering the AMR system contain any spinal cord, 
if the recovered product fails otherwise under any provision of 
paragraph (c)(1), if the product is not properly labeled under the 
provisions of paragraph (c)(2), or if the spent bone materials are not 
properly handled under the provisions of paragraph (c)(3) of this 
section.
    (2) The establishment must document its production process controls 
in writing. The program must be designed to ensure the on-going 
effectiveness of the process controls. If the establishment processes 
cattle, the program must be in its HACCP plan, its Sanitation SOP, or 
other prerequisite program. The program shall describe the on-going 
verification activities that will be performed, including the 
observation of the bones entering the AMR system for brain, trigeminal 
ganglia, and spinal cord; the testing of the product exiting the AMR 
system for bone solids, bone marrow, spinal cord, and DRG as prescribed 
in paragraph (c)(1) of this section; the use of the product and spent 
bone materials exiting the AMR system; and the frequency with which 
these activities will be performed.
    (3) The establishment shall maintain records on a daily basis 
sufficient to document the implementation and verification of its 
production process.
    (4) The establishment shall make available to inspection program 
personnel the documentation described in paragraphs (b)(2) and (b)(3) 
of this section and any other data generated using these procedures.
    (c) Noncomplying product. (1) Notwithstanding any other provision 
of this section, product that is recovered using advanced meat/bone 
separation machinery is not meat under any one or more of the following 
circumstances:
    (i) Bone solids. The product's calcium content, measured by 
individual samples and rounded to the nearest 10th, is more than 130.0 
mg per 100 g.
    (ii) Bone marrow. The product's added iron content, measured by 
duplicate analyses on individual samples and rounded to the nearest 
10th, is more than 3.5 mg per 100 g.\1\
---------------------------------------------------------------------------

    \1\ The excessive iron (ExcFe) measurement for an analyzed 
sample is equal to the obtained iron (Fe) result expressed in mg/100 
g measured and rounded to the nearest 100th or more for that sample, 
minus the product of three factors: (1) The iron to protein ratio 
(IPR) factor associated with corresponding hand-deboned product; (2) 
the obtained protein (P) result (%) for that sample; and (3) a 
constant factor of 1.10. In formula, this can be written as: ExcFe = 
mFe - IPR x Protein x 1.10, where ExcFe represents the excess iron, 
expressed in units of mg/100 g; mFe represents the measured level of 
iron (Fe, mg/100 g), IPR is the iron to protein ratio for the 
appropriate hand-deboned product, and ``Protein'' is the measured 
level of protein rounded to the nearest 100th and expressed as a 
percentage of the total weight of the sample. In lieu of data 
demonstrating otherwise, the values of IPR to be used in the above 
formula are as follows: For beef products the value of IPR is equal 
to 0.104, except for any combination of bones that include any beef 
neckbone product, for which the value of 0.138 is to be used; for 
pork product, the IPR value is 0.052. Other IPR values can be used 
provided that the operator of an establishment has verified and 
documented the ratio of iron content to protein content in the 
skeletal muscle tissue attached to bones prior to their entering the 
AMR system, based on analyses of hand-deboned samples, and the 
documented value is to be substituted for the IPR value (as 
applicable) in the above formula with respect to product that the 
establishment mechanically separates from those bones.
---------------------------------------------------------------------------

    (iii) Brain or trigeminal ganglia. Skulls that enter the AMR system 
have tissues of brain or trigeminal ganglia.
    (iv) Spinal cord. Vertebral column bones that enter the AMR system 
have tissues of spinal cord, or the product that exits the AMR system 
contains spinal cord.
    (v) DRG. The product that exits the AMR system contains DRG.
    (2) If product that may not be labeled or used as ``meat'' under 
this section meets the requirements of Sec.  319.5 of this subchapter, 
it may bear the name ``Mechanically Separated (Species)'' except as 
follows:
    (i) If skulls or vertebral column bones of cattle younger than 30 
months of age that enter the AMR system have tissues of brain, 
trigeminal ganglia, or spinal cord, the product that exits the AMR 
system shall not be used as an ingredient of a meat food product.
    (ii) If product that exits the AMR system contains spinal cord or 
DRG from bones of cattle younger than 30 months of age, it shall not be 
used as an ingredient of a meat food product.
    (iii) If product derived from any bones of cattle of any age does 
not comply with (c)(1)(i) or (ii), it may bear a common or usual name 
that is not false or misleading, except that the product may not bear 
the name ``Mechanically Separated (Beef).''
    (3) Spent skulls or vertebral column bone materials from cattle 
younger than 30 months of age that exit the AMR system shall not be 
used as an ingredient of a meat food product.

PART 320--RECORDS, REGISTRATION AND REPORTING

0
5. The authority citation for part 320 continues to read as follows:

    Authority: 21 U.S.C. 601-695; 7 CFR 2.7, 2.18, and 2.53.


Sec.  320.1  [Amended]

0
6. Section 320.1, paragraph (b)(10), is amended by removing ``of 
calcium content in meat derived from'' and adding, in its place, 
``documenting the development, implementation, and maintenance of 
procedures for the control of the production process using.''

    Done in Washington, DC, on: January 7, 2004.
Garry L. McKee,
Administrator.
[FR Doc. 04-626 Filed 1-8-04; 1:43 pm]
BILLING CODE 3410-DM-P