Generic E. coli Testing for Sheep, Goats, Equines, Ducks, Geese, and Guineas

DEPARTMENT OF AGRICULTURE

Food Safety and Inspection Service

9 CFR Parts 310 and 381

[Docket No. 97-004F]

RIN 0583-AC32

Generic E. coli Testing for Sheep, Goats, Equines, Ducks, Geese,

and Guineas

AGENCY: Food Safety and Inspection Service, USDA.

ACTION: Final rule.

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

SUMMARY: The Food Safety and Inspection Service (FSIS) is requiring

establishments that slaughter sheep, goats, horses, mules, and other

equines, and establishments that slaughter ducks, geese, and guineas,

to sample and test carcasses for generic E. coli. This final rule

extends the sampling and testing requirements already applied to

establishments that slaughter cattle, swine, chickens, and turkeys.

Regular microbial testing by slaughter establishments is necessary to

verify the adequacy of the establishment's process controls for the

prevention and removal of fecal contamination and associated bacteria.

EFFECTIVE DATE: January 25, 2000.

FOR FURTHER INFORMATION CONTACT: Daniel L. Engeljohn, Ph.D., Director,

Regulation Development and Analysis Division, Office of Policy, Program

Development, and Evaluation, FSIS, Room 112 Annex Building, Washington,

DC 20250-3700; telephone (202) 720-5627.

SUPPLEMENTARY INFORMATION:

Background

On July 25, 1996, FSIS published a final rule, ``Pathogen

Reduction; Hazard Analysis and Critical Control Point (HACCP)

Systems,'' (61 FR 38806). The new regulations (1) require that each

establishment develop, implement, and maintain written sanitation

standard operating procedures (Sanitation SOP's); (2) require regular

microbial testing for generic E. coli by establishments that slaughter

cattle, swine, chicken, and turkey to verify the adequacy of each

establishment's process control for the prevention and removal of fecal

contamination and associated bacteria; (3) establish pathogen reduction

performance standards for Salmonella that slaughter establishments and

establishments producing raw ground products must meet; and (4) require

that all meat and poultry establishments develop and implement a system

of preventive controls designed to improve the safety of their

products, a Hazard Analysis and Critical Control Point (HACCP) system.

At present, all inspected establishments that slaughter cattle,

swine, chickens or turkeys must sample and test carcasses for generic

E. coli. These establishments have developed sampling plans and sample

at specified frequencies, locations, and sites. They maintain records

of results and evaluate the results using either the m/M criteria

developed in FSIS' baseline studies or, if m/M criteria are not

available, statistical process control techniques. Establishments

defined as ``very low volume'' may sample at an alternative frequency.

Also, establishments operating under HACCP may develop alternative

sampling frequencies if certain requirements are met. The Pathogen

Reduction/HACCP final rule and the ``Pathogen Reduction/HACCP;

Technical Corrections and Amendment'' final rule (62 FR 26211) provide

detailed information about the need for these requirements.

On November 3, 1997, FSIS published a proposed rule in the Federal

Register (62 FR 59305) proposing to extend the sampling and testing

requirements for generic E. coli to meat establishments that slaughter

sheep, goats, and equines and to poultry establishments that slaughter

ducks, geese, and guineas. FSIS believes that regular microbial testing

by all slaughter establishments is necessary to verify the adequacy of

the establishment's process controls for the prevention and removal of

fecal contamination and associated bacteria.

[[Page 66548]]

Comments

FSIS received five comments during the public comment period that

ended January 2, 1998. Two of the commenters are members of the meat

industry, and two commenters represent industry trade associations. The

fifth commenter is from the Ministry of Agriculture of a major U.S.

trading partner. None of the commenters expressed any opposition to

extending the generic E. coli sampling and testing requirements to

minor species, and one commenter found both the costs and hours

necessary for implementation to be reasonable.

The principal areas of disagreement with the proposed rule dealt

with sampling/testing rates and sites. One of the commenters proposed

that the sampling for sheep and goats be done less frequently than for

cattle and swine. Another commenter disagreed with the application of

the 1 per 300 sampling frequency and proposed an ovine sampling rate of

1 per 810 as being more appropriate for establishments that slaughter

large numbers of sheep and lambs.

The sampling/testing site was another area about which there was

disagreement with the proposal. One commenter expressed a preference

for a single sampling site rather than the three sites proposed by

FSIS. Another commenter disagreed with the proposed sampling sites and

recommended that alternative sites be described in the final rule to

accommodate different dressing systems.

The flank, brisket, and rump are the same sites that were used by

FSIS when conducting the baseline studies for cattle and swine. FSIS'

decision not to change sites is based on the fact that there are no

available data to demonstrate that one-site sampling will provide

results comparable to the baseline survey data.

Moreover, it is appropriate, under HACCP, that the same three sites

be used for sampling all livestock carcasses. The Agency's

understanding of the minor species' dressing practices is that these

three sites are the most likely places where contamination would be

found. Although species-specific data are not currently available,

researchers at Colorado State University (CSU) are conducting sampling

analysis at three locations on lamb carcasses: the leg (rump, for

beef), flank, and breast (brisket, for beef). The Agency will consider

amending the regulations if the CSU data indicate a persuasive need to

considering other sampling sites. With regard to frequency of sampling/

testing, FSIS proposed a maximum limit of 13 samples per day, as was

done with poultry.

One commenter recommended a smaller template for lamb carcasses,

such as an area between 3 x 3 inches and 5 x 5 inches. In response to

this recommendation, FSIS agrees that a smaller template is reasonable

for sheep and goats, smaller species than livestock species, and will

specify a 5 x 10 cm template in its sampling procedures.

The Final Rule

FSIS now is extending these sampling and testing requirements to

sheep, goats, horses, mules, and other equines, defined as livestock in

9 CFR 301.2 (qq). All establishments slaughtering sheep, goats, horses,

mules, or other equines now are required to meet the sampling and

testing requirements in 9 CFR 310.25. Similarly, establishments that

slaughter ducks, geese, and guineas now are required to meet the

sampling and testing requirements in 9 CFR 381.94. These establishments

will only be required to test sheep, goats, equines, ducks, geese, or

guineas if they primarily slaughter these types of livestock or

poultry. FSIS considers the livestock or poultry an establishment

slaughters in the largest number to be that establishment's primary

type of livestock or poultry slaughtered. Finally, this final rule also

corrects an inadvertent inconsistency in the headings of Secs. 310.25

and 381.94 that appeared in the proposed rule by including the phrase

``process control verification criteria and testing'' as well as

removing the phrase ``for Salmonella'' in Sec. 310.25.

Sampling Frequencies and Definitions for Very Low Volume

Establishments

For the Pathogen Reduction/HACCP final rule, FSIS used a

methodology to select sampling frequencies so that, in the subgroup of

establishments accounting for 99 percent of total production for each

type of livestock or poultry, the 5 percent of establishments with the

highest production volume would each conduct a minimum of 13 E. coli

tests, or one complete test window each day. Under these frequencies,

90 percent of all cattle, 94 percent of all swine, 99 percent of all

chickens, and 99 percent of all turkeys would be slaughtered in

establishments conducting a minimum of one complete E. coli sampling

window per day to provide a minimum, adequate basis for process control

verification.

FSIS developed alternative sampling frequencies for establishments

defined as ``very low volume.'' If there are published m/M criteria for

the type of livestock or poultry primarily slaughtered, the

establishment must sample that type of livestock or poultry at a

minimum frequency of once per week, starting the first full week of

operation after June 1 of each year until a series of 13 tests has met

those m/M criteria. If there are no m/M criteria for the type of

livestock or poultry primarily slaughtered, a very low volume

establishment must collect at least one sample per week, starting the

first full week of operation after June 1 of each year, and continue

sampling at a minimum of once each week that the establishment operates

until June 1 of the following year or until 13 samples have been

collected, whichever comes first. This provision will be eliminated

once m/M criteria are developed for the primary type of livestock or

poultry slaughtered.

FSIS permits very low volume establishments to test at this

frequency, in part, because of their relatively simple and stable

production environments. Also, FSIS assumes that the total risk of

exposure to enteric pathogens from products produced at such

establishments would be small and roughly proportional to the amount of

products produced. FSIS requires these establishments to begin testing

in June because it is most important for these establishments to

conduct testing during the summer months, when there is a seasonal peak

in the occurrence of foodborne diseases attributable to the major

bacterial pathogens.

The Pathogen Reduction/HACCP final rule (61 FR 38842) noted that

very low volume cattle and swine establishments account for only 1.5

percent and 1.3 percent of overall production, respectively. Very low

volume chicken and turkey establishments account for .05 percent and

.01 percent of overall production, respectively.

FSIS is now requiring establishments that slaughter sheep, goats,

and equines to sample at the same frequency now required for cattle,

one test per 300 carcasses. Similarly, FSIS is requiring establishments

that slaughter ducks, geese, and guineas to sample at the same

frequency now required for turkeys, one test per 3,000 carcasses. FSIS

also is requiring establishments that slaughter sheep, goats, equines,

ducks, geese, and guineas, except those defined as very low volume

establishments, to conduct sampling at a frequency of at least once per

week to provide a minimum, adequate basis for process control

verification.

FSIS is requiring ``very low volume'' establishments that slaughter

sheep, goats, equines, ducks, geese, and guineas to collect at least

one sample per week, starting the first full week of

[[Page 66549]]

operation after June 1 of each year, and continue sampling at a minimum

of once each week that the establishment operates until June 1 of the

following year or until 13 samples have been collected, whichever comes

first. At this time, baseline studies have not been conducted to

develop m/M criteria for sheep, goats, equines, ducks, geese, and

guineas. When m/M criteria are developed for any of these types of

livestock or poultry, FSIS intends to require sampling at a minimum

frequency of once per week starting the first full week of operation

after June 1 until a series of 13 tests has met those m/M criteria.

Sheep and Goats

In fiscal year 1993, 93 establishments slaughtered primarily sheep

and goats. FSIS is requiring that these establishments sample at a

frequency of one test per 300 carcasses or at least once a week,

whichever is greater, unless they are very low volume establishments.

At this sampling frequency, 85 percent of all sheep and goats will be

slaughtered in establishments conducting a minimum of 13 samplings each

day or one complete E. coli test window. A very low volume sheep or

goat slaughter establishment is one that annually will slaughter no

more than 6,000 head. Based on fiscal year 1993 data, 61 of the 93

establishments will be classified as very low volume and account for

1.9 percent of total sheep and goat production.

Equines

In fiscal year 1995, eight establishments slaughtered equines under

Federal inspection for human food. These eight establishments

slaughtered only equines. The Agency now is requiring that horse, mule,

or other equine slaughter establishments sample at a rate of one per

300 carcasses or at least once a week, whichever is greater, unless

they are very low volume establishments. Very low volume equine

establishments are those that will annually slaughter no more than

6,000 equines. Two of the equine establishments, slaughtering 5.6

percent of overall production, will be classified as very low volume.

Ducks, Geese, and Guineas

In fiscal year 1995, there were 12 establishments that slaughtered

primarily ducks and two establishments that slaughtered primarily

geese. FSIS is not aware of any federally inspected establishment

currently slaughtering guineas. FSIS now is requiring establishments

that slaughter ducks, geese, and guineas to sample at a frequency of

one test per 3,000 carcasses or at least once a week, whichever is

greater, unless they are very low volume establishments. At this

frequency, 96 percent of all ducks will be slaughtered in

establishments conducting a minimum of one E. coli test per day. Very

low volume duck, geese, or guinea establishments are those that will

slaughter no more than 60,000 ducks, geese, or guineas, respectively, a

year.

In FY 1995, 25 establishments slaughtered 19.2 million ducks. Only

12 establishments slaughtered primarily ducks. These establishments

produced 98.7 percent of all ducks slaughtered under Federal

inspection. One of the 12 establishments produces less than 0.2 percent

of ducks slaughtered and will be defined as a very low volume duck

establishment.

Eight establishments under Federal inspection slaughtered 159,000

geese in FY 1995. Only two establishments slaughtered primarily geese

and only one of these establishments slaughters more than 60,000 geese.

Alternative Sampling Frequencies

Establishments operating under a validated HACCP plan in accordance

with 9 CFR 417.2(b) will be permitted to substitute an alternative

frequency if the alternative is an integral part of the establishment's

verification procedures for its HACCP plan. Establishments will not be

allowed to use an alternative frequency if FSIS determines, and

notifies the establishment in writing, that the alternative frequency

is inadequate to verify the effectiveness of the establishment's

process controls.

Sampling Plans

Establishments that slaughter sheep, goats, equines, ducks, geese,

and guineas will prepare written specimen collection procedures. The

procedures will include the identification of employees designated to

collect samples, the location(s) of sampling, how sampling randomness

is achieved, and how samples are handled to ensure sample integrity.

The written procedures will be made available to FSIS upon request.

Sampling Locations

Establishments that slaughter sheep, goats, and equines will

collect samples from chilled carcasses. Carcasses boned before chilling

(hot boned) will be sampled after the final wash. For ducks, geese, and

guineas, samples will be taken from the end of the chilling process,

after the drip line. If the bird is boned before chilling, the sample

will be taken from the end of the slaughter line instead of the end of

the drip line.

Sampling Sites

Samples from sheep, goat, and equine carcasses will be taken by

sponging tissue from three sites: the flank, brisket, and rump. The

sponge is to be placed afterward in an amount of buffer to transfer any

E. coli to a solution, which then is analyzed for E. coli. Hide-on

carcasses will be sampled by sponging from inside the flank, inside the

brisket, and inside the rump. Samples from ducks, geese, and guineas

will be collected by taking whole birds from the end of the chilling

process after the drip line and rinsing them in an amount of buffer

appropriate to the type of bird being tested.

Recordkeeping

Establishments will enter test results onto a process control chart

or table and record the results in terms of colony forming units per

milliliter (cfu/ml) for poultry carcasses or per square centimeter

(cfu/cm2) for livestock carcasses. Establishments will use the records

to evaluate test results. These records will be maintained at the

establishment for 12 months and must be made available to inspection

program personnel upon request. Inspection program personnel will

review results over time to verify effective and consistent process

control.

Evaluation Criteria

Establishments will evaluate results using statistical process

control techniques until such time as m/M criteria are established for

these types of livestock and poultry. FSIS intends to give high

priority in its baseline plan to collect data that will support

establishing m/M criteria for ducks and geese, and sheep.

International Implementation

The Federal Meat Inspection Act and the Poultry Products Inspection

Act require that meat and poultry products imported into the United

States be produced under an inspection system that is equivalent to the

U.S. inspection system. In determining the equivalency of a foreign

country's eligibility to import meat or poultry products into the

United States, FSIS evaluates the laws, policies, and administration of

that country's inspection system. This assessment includes on-site

reviews of individual establishments, laboratories, and other

facilities within the foreign system. With this final rule, countries

that export products of sheep, goats, equines, ducks, geese, and

guineas to the United States must implement

[[Page 66550]]

equivalent generic E. coli sampling and testing programs as a condition

of maintaining eligibility for access to the U.S. market.

The burden for demonstrating equivalence rests with the exporting

country, and the United States is free to set any level of protection

it considers appropriate to control or eliminate a foodborne hazard.

Equivalent regulatory systems need not be identical. FSIS has

established a level of protection that domestic establishments must

achieve. Exporting countries may propose alternative methods of

achieving equivalent levels of protection and are advised to consult

with FSIS on any proposed alternatives that they believe will meet U.S.

requirements.

Executive Order 12866 and Regulatory Flexibility Act

This final rule has been determined to be not significant, and

therefore, has not been reviewed by the Office of Management and

Budget.

The Administrator made an initial determination that this rule will

not have a significant economic impact on a substantial number of small

entities, as defined by the Regulatory Flexibility Act (5 U.S.C. 601).

This final rule is an extension of the Pathogen Reduction/HACCP

rule, which is economically significant. Many aspects of that

economically significant rule, such as the public health risks

associated with pathogens present in fecal contamination and the

potential health benefits of pathogen reduction, are applicable to this

rule. In the Final Regulatory Impact Assessment (FRIA)(61 FR 38945,

July 25, 1996) for the Pathogen Reduction/HACCP rule, FSIS addressed

these areas in detail.

By extending the requirement for generic E. coli testing to

additional types of livestock and poultry, this final rule will

increase the effectiveness of pathogen reduction efforts and generate

proportional increases in predicted public health benefits. The

benefits assessment in the FRIA was based on a proportional reduction

assumption: that is, an assumption was made that a reduction in

pathogens at the manufacturing level leads to a proportional reduction

in foodborne illness. Public health benefits are quantified in terms of

reduced cost of foodborne illness.

This final rule will affect an estimated 101 federally inspected

livestock slaughter establishments and a smaller number of State

inspected livestock slaughter establishments. The total of 101

federally inspected establishments includes 11 establishments that

slaughter only sheep and goats, 82 establishments that slaughter cattle

and/or swine but slaughter sheep or goats as their primary type of

livestock, and eight establishments that slaughter equines. In

addition, there are 574 establishments that slaughter sheep or goats

but are not affected because they primarily slaughter cattle or swine.

This final rule also will affect an estimated 14 federally

inspected poultry slaughter establishments and possibly a few State

inspected poultry slaughter establishments. These include 12 federally

inspected establishments that slaughter primarily ducks and two

establishments that slaughter primarily geese. There are 14

establishments that also slaughter ducks and/or geese, but are not

affected because they slaughter primarily chickens or turkeys. There

are currently no establishments that slaughter guineas under Federal

inspection.

Executive Order 12898

Pursuant to Executive Order 12898, ``Federal Actions to Address

Environmental Justice in Minority Populations and Low-Income

Populations,'' FSIS has considered potential impacts of this final rule

on environmental and health conditions in low-income and minority

communities.

This final rule extends the sampling and testing requirements

already applied to establishments that slaughter cattle, swine,

chickens, and turkeys. As explained in the economic impact analysis

above, the Final Regulatory Impact Assessment (FRIA) for the Pathogen

Reduction/HACCP final rule addressed in detail the fact that this rule

will generally benefit FSIS, the regulated industries, and consumers.

The final rule will not require or compel meat or poultry

establishments to relocate or alter their operations in ways that could

adversely affect the public health or environment in low income and

minority communities. Further, this final rule will not exclude any

persons or populations from participation in FSIS programs, deny any

persons or populations the benefits of FSIS programs, or subject any

persons or populations to discrimination because of their race, color,

or national origin.

Cost Analysis

This analysis is based on the same estimates and assumptions that

were used to develop the FRIA for the Pathogen Reduction/HACCP rule and

include:

(1) An average cost of $24 for collecting and analyzing a sample

for generic E. coli.

(2) A cost of $640 for the preparation of a sampling plan. This

estimate is based on 25 hours at $25.60 per hour, the average wage of a

quality control manager.

(3) A cost of $403 per establishment for an estimated three out of

four establishments that would require extra training for aseptic

sampling.

(4) An estimate of 26 sample collections annually by very low

volume slaughter establishments. (The proposed requirement is a minimum

of 13.)

(5) An estimate of five minutes to record and review laboratory

results for each sample by an employee earning $13.42 per hour.

Sheep, Goats, and Equines

Unless otherwise specified, this cost analysis is based on data

from the Agency's Enhanced Economic Database. This database includes

Animal Disposition Retrieval System (ADRS) data from FY 1993. Sheep and

goat production were combined in the Enhanced Economic Database.

Although the proposed rule treats sheep and goats as two separate types

of livestock, the cost analysis is based on combined sheep and goat

production. This has a minimal impact on the accuracy of the cost

estimates.

There are 11 establishments that slaughter only sheep and goats.

This final rule will extend mandatory generic E. coli testing to these

establishments. Each of these 11 establishments will be required to

develop a sampling plan at a cost of $640 per establishment or $7,040

in total. This cost would include items such as preparing a written

plan, establishing sampling procedures, locating a laboratory and

arranging for necessary supplies, and developing the statistical

process control techniques to be used for analyzing results.

This analysis assumes that eight establishments (75 percent) would

require training in aseptic sampling at a cost of $3,224 (8 times

$403). Three of the eleven establishments will be very low volume

establishments and will analyze 26 samples per year for a recurring

cost of $1,872. Based on production data and a sampling rate of one in

300, the other eight establishments will analyze a total of 8,015

samples annually at a cost of $192,360. Recording and reviewing costs

for 8,015 samples will require 668 hours annually and cost $8,970. The

annual recording and reviewing costs for the three very low volume

establishments will be $87 (6.5 hours at $13.43 per hour).

As discussed above, there are 82 establishments that slaughter

cattle and/or swine but slaughter sheep or goats as their primary type

of livestock. There

[[Page 66551]]

will be no additional cost for 58 of these establishments because these

58 are now required to conduct sampling as very low volume cattle or

swine slaughter establishments and will be very low volume sheep or

goat establishments under this rule. The impact on these 58

establishments will be a shift in the type of livestock sampled. The

Agency is not aware of any basis to conclude that establishments could

not make this shift without additional costs for sampling plan

development.

The other 24 establishments within the 82 that slaughter cattle

and/or swine and sheep or goats are now required to test cattle or

swine. However, under this final rule, they will have to conduct

additional analyses based on their sheep or goat production. Their

sheep/goat production is greater than the larger of their cattle or

swine production. As they shift from cattle or swine to sheep or goats,

annual sampling will increase by 2,928 samples or $70,272 per year.

Annual recording and reviewing costs will be $3,277 (244 hours at

$13.43 per hour).

This final rule will also extend mandatory generic E. coli testing

to 8 establishments that slaughter equines for human food. Based on FY

1995 ADRS data, these eight establishments will be required to conduct

469 analyses per year. It is assumed they will all have to develop

sampling plans ($640 each) and that six will have to obtain training in

aseptic sampling ($403 per establishment). Two of the eight

establishments that slaughter equines will meet the definition for a

very low volume establishment. The total recurring cost for 469

analyses will be $11,256. Recording and reviewing costs will be $525

per year (39 hours at $13.43 per hour).

In conclusion, there are 43 federally inspected livestock slaughter

establishments that will experience increased costs under this final

rule. The one time up-front costs will total $17,802, $5,642 for

training in aseptic sampling and $12,160 for sampling plan development.

The total recurring cost for the 43 establishments will be $288,619,

$275,760 for sample collection and analysis and $12,859 for recording

and reviewing test results.

All the costs discussed above for establishments that slaughter

sheep, goats, and equines are summarized in Table 1.

Table 1.--Costs for Implementing Generic E. coli Sampling Programs in Sheep, Goat, and Equine Establishments

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

Sample

Number of Training for Sampling collection Recording

Production category establishments aseptic plan and analysis and review

sampling development (recurring) (recurring)

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

Exclusively Sheep or Goats with Annual 8 $2,418 $5,120 $192,360 $8,970

Production over 6,000..................

Exclusively sheep or Goats with Annual 3 806 1,920 1,872 87

Production under 6,000.................

Primarily Sheep or Goats with Annual 24 ............ ............ 70,272 3,277

Production over 6,000..................

Equine.................................. 8 2,418 5,120 11,256 525

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

Total............................... 43 5,642 12,160 275,760 12,859

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

Ducks, Geese, and Guineas

The ADRS data show that 28 federally inspected establishments

slaughtered ducks and/or geese in FY 1995. FSIS is not aware of any

establishment slaughtering guineas. Six establishments slaughtered only

the types of poultry covered by this final rule. This final rule

extends mandatory generic E. coli testing to six federally inspected

poultry establishments that are not currently required to test. There

are eight poultry slaughter establishments that currently test chickens

or turkeys, but slaughter more ducks or geese and, therefore, will

shift their testing program to ducks or geese. Seven of these

establishments will have to conduct more testing because they will not

be very low volume establishments based on their duck or goose

production. The eighth establishment will shift from a very low volume

establishment that slaughters chickens to a very low volume

establishment that slaughters ducks, and it and will not incur any

additional costs.

In summary, under this final rule, 14 establishments will test

ducks or geese. Two of these establishments will be very low volume

establishments. All 14 poultry slaughter establishments affected by

this final rule were included in the Pathogen Reduction/HACCP FRIA as

very low volume poultry slaughter establishments, that is, annual

chicken slaughter under 440,000 and annual turkey slaughter under

60,000. The methodology applied in the FRIA started with all 306

poultry slaughter establishments (FY 1993 ADRS data) in the Agency's

Enhanced Economic Database. FSIS calculated the costs for 208

establishments processing more than 440,000 chickens annually and the

costs for 48 establishments processing more than 60,000 turkeys

annually. FSIS treated the remaining 50 poultry slaughter

establishments as very low volume establishments.

This methodology most likely overestimated costs; more recent FY

1995 ADRS data include six poultry slaughter establishments processing

ducks and/or geese exclusively. This cost analysis separates the costs

already addressed and the incremental costs of basing sampling

frequency upon duck and geese production. The costs already addressed

that are actually costs of this final rule include the cost of six

sampling plans at $640 per plan or $3,840; training in aseptic sampling

for five establishments at $403 per establishment or $2,015; sample

collection and analysis costs for 156 (6 multiplied by 26) samples per

year at a cost of $24 per sample or $3,744; and recording and record

review costs of $175. Using duck and geese production levels from FY

1995, five of the six establishments slaughtering only ducks and geese

will have to collect and analyze an additional 2,281 samples per year

at an annual cost of $54,744. The recording and reviewing costs for

2,281 samples will be $2,553 annually. The other establishment is a

very low volume establishment.

As discussed above, there are seven establishments that are

currently required to test for chickens or turkeys as very low volume

establishments, but will have to conduct more analyses under this final

rule because they will not be very low volume establishments based on

their duck or goose production. These seven establishments will have to

collect and analyze an additional 3,769 samples annually at a cost of

$90,456. Recording and review costs at five minutes per sample will

total $4,218 per year.

[[Page 66552]]

The total cost for extending testing to ducks, geese, and guineas

includes a one-time cost of $5,855 for training and sampling plan

development and an annual recurring sampling and recording cost of

$155,890. The cost for requiring generic E. coli sampling in

establishments that slaughter ducks, geese, and guineas are summarized

in Table 2.

Table 2: Costs for Implementing Generic E. coli Sampling Programs for Duck, Goose, and Guinea Establishments

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

Sample

Number of Training for Sampling collection Recording

Production category establishments aseptic plan and analysis and review

sampling development (recurring) (recurring)

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

Ducks and Geese only-Costs Included in 6 $2,015 $3,840 $3,744 $175

FRIA...................................

Ducks and Geese only-Costs not \1\ 5 ............ ............ 54,744 2,553

previously accounted for...............

Establishments Currently Required to 7 ............ ............ 90,456 4,218

test Chickens or Turkeys...............

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

Total............................... 13 2,015 3,840 148,944 6,946

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

\1\ Included in the 6 above.

Overall Summary of Cost Analysis

This final rule will extend mandatory generic E. coli sampling

requirements to 25 federally inspected establishments, 11 that

slaughter sheep and goats exclusively, eight that slaughter equines,

and six that currently slaughter only ducks and/or geese. The

nonrecurring up-front cost for these establishments will total $23,657.

The annual recurring cost for collecting and analyzing 10,999 samples

and recording and reviewing results for these 25 establishments will be

$276,286. There are 31 establishments that currently test cattle,

swine, chickens or turkeys that will have to increase their testing

programs by 6,697 samples. The increase in annual 28 recurring costs

for these 31 establishments will be $168,223.

The costs summarized in Tables 1 and 2 are maximum costs because

the final rule will allow for establishments operating under a

validated HACCP system to use sampling frequencies other than those

specified in this final rule if the alternative sampling frequency is

an integral part of the establishment's HACCP plan. The cost estimates

in Tables 1 and 2 do not account for possible reductions in sampling

frequencies.

Executive Order 12988

This final rule has been reviewed under Executive Order 12988,

Civil Justice Reform. Because this rule is final, (1) all state and

local laws and regulations that are inconsistent with this rule will be

preempted; (2) no retroactive effect will be given to this rule; and

(3) administrative proceedings will not be required before parties may

file suit in court challenging this rule.

Paperwork Requirements

Abstract: As part of microbiological testing, each slaughter

establishment will be required to develop written procedures outlining

specimen collection and handling. The slaughter establishments will be

responsible for entering the results into a statistical process control

chart or table. The data and chart will be available for review by FSIS

upon request.

Estimate of Burden: Agency subject matter experts estimate that it

will take 25 hours for establishments to develop a microbial sampling

and analysis plan. It will take an estimated 17.5 minutes to collect

samples and 5 minutes per sample to enter data into charts, and then

review and file the information.

This final rule will require 25 federally inspected establishments

to develop sampling plans. FSIS estimates that each plan will require

25 hours to develop. Plan development for 25 establishments will

require 625 burden hours. Fifty-six establishments will be required to

collect samples and to record new or additional test results. These 56

establishments will be required to collect and record and review the

results of 17,696 analyses, annually. To collect samples at 17.5

minutes per sample, 5,161 burden hours will be required. It will take

1,475 burden hours at 5 minutes per result to record and review

results.

Respondents: Livestock and poultry product establishments.

Estimated Number of Respondents: 56.

Estimated Number of Responses per Respondents: 18,402.

Estimated Total Annual Burden on Respondents: 7,261 hours.

Copies of this information collection assessment can be obtained

from Lee Puricelli, Paperwork Specialist, Food Safety and Inspection

Service, USDA, 300 12th Street, SW, Room 109, Washington, DC 20250-

3700.

List of Subjects

9 CFR Part 310

Meat Inspection, Microbial testing.

9 CFR Part 381

Poultry and poultry products, Microbial testing.

For the reasons set forth in this preamble, 9 CFR chapter III is

amended as follows:

PART 310--POST MORTEM INSPECTION

1. The authority citation for part 310 continues to read as

follows:

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

2. Section 310.25 is amended by revising the section heading, the

first sentence of paragraph (a)(1) introductory text, paragraphs

(a)(2)(ii), (a)(2)(iii), and the first sentence of (a)(2)(v)(A) to read

as follows:

Sec. 310.25 Contamination with microorganisms; process control

verification criteria and testing; pathogen reduction standards.

(a) * * *

(1) Each official establishment that slaughters livestock must test

for Escherichia coli Biotype 1 (E. coli) * * *.

* * * * *

(2) * * *

(ii) Sample collection. The establishment must collect samples from

all chilled livestock carcasses, except those boned before chilling

(hot-boned), which must be sampled after the final wash. Samples must

be collected in the following manner:

(A) For cattle, establishments must sponge or excise tissue from

the flank, brisket and rump, except for hide-on calves, in which case

establishments must take samples by sponging from

[[Page 66553]]

inside the flank, inside the brisket, and inside the rump.

(B) For sheep, goat, horse, mule, or other equine carcasses,

establishments must sponge from the flank, brisket and rump, except for

hide-on carcasses, in which case establishments must take samples by

sponging from inside the flank, inside the brisket, and inside the

rump.

(C) For swine carcasses, establishments must sponge or excise

tissue from the ham, belly and jowl areas.

(iii) Sampling frequency. Slaughter establishments, except very low

volume establishments as defined in paragraph (a)(2)(v) of this

section, must take samples at a frequency proportional to the volume of

production at the following rates:

(A) Cattle, sheep, goats, horses, mules, and other equines: 1 test

per 300 carcasses, but a minimum of one sample during each week of

operation.

(B) Swine: 1 test per 1,000 carcasses, but a minimum of one sample

during each week of operation.

* * * * *

(v) * * *

(A) Very low volume establishments annually slaughter no more than

6,000 cattle, 6,000 sheep, 6,000 goats, 6,000 horses, mules, or other

equines, 20,000 swine, or a combination of livestock not exceeding

6,000 cattle and 20,000 total of all livestock. * * *

* * * * *

PART 381--POULTRY PRODUCTS INSPECTION REGULATIONS

3. The authority citation for part 381 continues to read as

follows:

Authority: 7 U.S.C. 138f, 450; 21 U.S.C. 451-470, 7 CFR 2.18,

2.53.

Subpart K--Post Mortem Inspection; Disposition of Carcasses and

Parts

4. Section 381.94 is amended by revising paragraph (a)(2)(iii), the

first and second sentence of paragraph (a)(2)(v)(A), and table 1 in

paragraph (a)(5)(i) as follows:

Sec. 381.94 Contamination with microorganisms; process control

verification criteria and testing; pathogen reduction standards.

(a) * * *

(2) * * *

(iii) Sampling frequency. Slaughter establishments, except very low

volume establishments as defined in paragraph (a)(2)(v) of this

section, must take samples at a frequency proportional to the

establishment's volume of production at the following rates:

(A) Chickens: 1 sample per 22,000 carcasses, but a minimum of one

sample during each week of operation.

(B) Turkeys, Ducks, Geese, and Guineas: 1 sample per 3,000

carcasses, but a minimum of one sample during each week of operation.

* * * * *

(v) * * *

(A) Very low volume establishments annually slaughter no more than

440,000 chickens or 60,000 turkeys, 60,000 ducks, 60,000 geese, 60,000

guineas or a combination of all types of poultry not exceeding 60,000

turkeys and 440,000 birds total. Very low volume establishments that

slaughter turkeys, ducks, geese, or guineas in the largest number must

collect at least one sample during each week of operation after June 1

of each year, and continue sampling at a minimum of once each week the

establishment operates until June 1 of the following year or until 13

samples have been collected, whichever comes first.

* * * * *

(5) * * *

(i) * * *

Table 1.--Evaluation of E. coli Test Results

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

Maximum number

Lower limit of Upper limit of Number of permitted in

Types of poultry marginal range marginal range sample tested marginal range

(m) (M) (n) (c)

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

Chickens.................................... \1\ 100 \1\ 1,000 13 3

Turkeys..................................... * NA *NA * NA * NA

Ducks....................................... * NA * NA * NA * NA

Geese....................................... * NA * NA * NA * NA

Guineas..................................... * NA * NA * NA * NA

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

\1\ CFU/ml.

* Values will be added upon completion of data collection programs.

* * * * *

Done at Washington, DC, on: November 18, 1999.

Thomas J. Billy,

Administrator.

[FR Doc. 99-30602 Filed 11-26-99; 8:45 am]

BILLING CODE 3410-DM-P