Annex O
Policy on the Control of E. coli O157:H7/NM Contamination in Raw Beef Products

1. Effective date

This policy is effective on the date of publication.

2. Purpose

The purpose of this policy is:

• to provide clear guidance to industry and inspection staff on the measures required to control Escherichia coli O157:H7/NM in raw beef products; and
• to reflect the risk-based approach taken by the CFIA to address the risk posed by this pathogen.

3. Definitions

In the context of this policy, the following definitions apply.

3.1 Accredited laboratory

A laboratory that is formally recognized by the Standards Council of Canada (SCC), the Canadian Association for Laboratory Accreditation (CALA), or another accreditation body that is a signatory to the International Laboratory Accreditation Cooperation (ILAC) Mutual Recognition Agreement (MRA) as conforming to the requirements of ISO/IEC 17025:2005.

3.2 E. coli O157:H7/NM

Escherichia coli O157:H7/NM are enterohaemorrhagic, shiga toxin producing strains of Escherichia coli (E. coli).

Reference to E. coli O157 includes both H7 and non-motile (NM) bacteria.

The presumptive positive and confirmed positive results of E. coli O157 are defined as follows:

3.2.1 Presumptive positive

A sample that gives a positive test result for potential presence of E. coli serotype O157 from an enrichment broth is referred to as a presumptive positive (i.e. causes a positive reaction with a CFIA recognized screening test presented in Appendix 2 of this policy).

Operator may choose to treat presumptive positive results as confirmed positive results unless otherwise specified with the policy (see section 9.2 c andsection 12).

3.2.2 Confirmed positive

To be confirmed positive, a pure isolate from the original enrichment broth must exhibit typical appearance on selective agars and:

• is serologically and biochemically determined to be E. coli O157, and/or
• is genetically determined to be E. coli O157.

Additionally, to be a confirmed positive, one of the following criteria must be met:

• positive for Shiga toxin (ST) production; and/or
• positive for Shiga toxin gene(s) (stx).

3.3 Epidemiological evidence

Epidemiological evidence refers to data (descriptive or analytical in nature) demonstrating an association between a food product and human illness as determined by the study of the frequency, distribution and determinants of the particular outbreak.

3.4 Full lethality treatment

A beef product is considered to have received a full lethality treatment when the manufacturing process has been scientifically validated to achieve a 5 log reduction of E. coli O157.

3.5 Implicated product

Implicated product consists of, as a minimum, the lot of raw beef product that generated a positive result when tested for E. coli O157. Additional products may be implicated due to various circumstances as outlined in section 12.1.

3.6 Raw beef

Raw beef includes beef and veal, as well as their hearts, head meat, cheek meat, oesophagus, etc. Raw beef products include intact beef products, non-intact beef products and comminuted beef products.

Note:

Beef tails and tongues are excluded from the raw beef product definition since they are customarily fully cooked and have not been linked to human illnesses.

3.6.1 Intact raw beef

Intact raw beef is a piece of meat whose internal structure has not been modified. This category includes: dressed carcasses in whole/half or quarter format, primal and sub-primal cuts, steaks, roasts, briskets and stew beef, trimmings removed from the aforementioned parts, head meat, cheek meat, diaphragm, and intercostal muscle.

3.6.2 Non-intact raw beef

Non intact raw beef is beef that has been either:

• tenderized;
• injected;
• subjected to a process that incorporates a solution (e.g.: massaging, tumbling);
• comminuted.

Comminuted beef includes ground, finely textured, chopped, mechanically separated, flaked and minced beef.

3.6.3 Precursor material (PM)

Precursor material includes any raw beef products intended to be used for producing finished raw ground beef product (FRGBP). This includes, but is not limited to, trims, boneless beef, coarse ground beef, hearts, head meat, cheek meat, weasand meat. It includes primal cuts, such as chucks, if intended to be used for production of FRGBP.

3.6.4 Finished raw ground beef products (FRGBP)

Finished Raw Ground Beef Products include all raw ground beef products that will be sold to consumers in that state, as well as raw meat products that contain comminuted and formed beef (e.g., patties, burgers, steakettes, etc.). It includes pre-packaged product as well as bulk product that will be repackaged, either by processors or retailers, for consumer use. It does not include ground beef that will be used for further processing into different products (e.g., sausages).

4. Risk posed by E. coli O157

Based on currently available information, E. coli O157 contamination of raw beef is a health hazard likely to occur and control measures need to be implemented.

The following information must be included in the Hazard Analysis and Critical Control Points (HACCP) System:

• Food Safety Enhancement Program (FSEP) forms (or equivalent) on product identification and intended use are to be reviewed for completeness and accuracy with regards to the E. coli O157 hazard and controls;
• the E. coli O157 hazard shall be clearly identified on the FSEP form # 5 (or equivalent). This hazard must be passed through the decision tree (FSEP form # 8 or equivalent);
• the operator has to determine how the hazard will be addressed:
  • through CCP(s) at the establishment during production;
  • when receiving raw beef products;
  • through other validated measures that will prevent distribution of potentially contaminated products.

For abattoirs, particular attention must be paid to airflow, as well as employee movement between the clean and the relatively less clean processing areas as these factors have recently been strongly associated with the microbiological contamination levels of the carcasses. Segregation of incompatible activities, product flow, equipment sanitation as well as employee training (Good Manufacturing Practices [GMPs], hygienic handling, dressing procedures etc.) should also be evaluated. Finally, as adequate refrigeration plays an important role in preventing the growth of E. coli O157, operators should confirm that they have sufficient cooling capacity to handle their volume of production while meeting the carcass cooling performance requirements found in the Meat Hygiene Manual of Procedures, Chapter 17.

5. Control measures/interventions

Operators shall ensure that their HACCP system has been validated and is effective, such that the level of E. coli O157 in raw beef products is below the detectable level (i.e., no E. coli O157 detected in a sample when tested with one of the approved methodologies [refer to Appendix 2 of this policy for the relevant information]).

The following control measures are to be implemented:

5.1 Abattoirs

Cattle are the primary source of E. coli O157 that infects humans. The shedding of E. coli O157 in cattle feces is intermittent and increases during summer and fall season. Contamination of beef carcasses with E. coli O157 occurs during slaughtering and dressing procedures, especially during the de-hiding process. Subsequent use of any meat components derived from a contaminated carcass in FRGBP, would be considered a risk to human health.

5.1.1 All abattoirs

In the case of a slaughter establishment, the operator must:

• use at least one pathogen reduction intervention, such as steam pasteurizers, organic acid sprays, etc., validated according to section 8 of this policy, to reduce the E. coli O157 contamination to below detectable level. This intervention must be covered with a CCP. Where more than one intervention contribute to the reduction of E. coli O157 contamination to below detectable level, the earlier intervention steps that contribute to the reduction but cannot reduce E. coli O157 contamination to below detectable level may be managed by process controls. The final intervention step that results in the reduction of E. coli O157 contamination to below detectable level must be managed by a critical control point (CCP). (See FSEP Manual 3.2.7 - CCP and 3.2.8 - Process Control requirements)
• control air contamination, airflow (positive pressure) and employee movements between clean and relatively less clean processing areas;
• develop, implement and maintain a storage and transportation prerequisite program to ensure that conditions under which the carcasses are kept are satisfactory;
• implement a verification step to ensure the effectiveness of the interventions covered by the CCP. This procedure must include testing of the carcasses for the pathogen or a surrogate organism, as described in section 9.1 c of this policy, in the slaughter establishment or at receiving in the receiving establishment. In the latter case, the operator of the abattoir must have a CFIA approved integrated plan in place.

Additionally, operators may choose to:

• collaborate with producers and transporters on pre-harvest management practices aimed at reducing E. coli O157 load in live animals presented for slaughter;
• assess the cleanliness of animals at live animal receiving by mud scoring and use this information to establish appropriate procedures such as the sequencing of slaughter, reducing line speed, adding trimmers or enhancing monitoring of slaughter process;
• utilize a hide-on carcass wash to remove excess organic matter; and
• use additional process controls as part of a multi-hurdle approach to reduce the E. coli O157 contamination of raw meat product to below detectable level.

5.1.2 Abattoirs with boning/cutting operations

In addition to the requirements stated in section 5.1.1, the following requirements apply to abattoirs with boning/cutting operations:

• The storage prerequisite program must indicate that conditions under which the carcasses are kept until being boned/cut are satisfactory. This includes, but is not limited to, storage temperature as well as potential for cross-contamination;
• Conditions under which the carcasses are further processed (cutting/boning) must be satisfactory and meet the HACCP system specifications; and
• If any of the raw beef products obtained from the boning/cutting processes tests positive for E. coli O157 (at the abattoir with boning/cutting operations or at the receiving establishment), the investigation must also include the evaluation of all parameters of the slaughter process that have an impact on the presence and potential spread of E. coli O157.

5.2 Receiving establishment

The establishment receiving raw beef must choose either one of the following options:

• implement purchase specifications (please refer to section 6 of this policy) and determine that the controls provided by either the Receiving CCP or the Purchasing/Receiving/Shipping Prerequisite Program (B 2.1) receiving step are adequate to ensure that these purchase specifications are met. These purchase specifications must require that all suppliers have one or more validated CCPs in their production of raw beef shipped to the receiving establishment and that their controls are implemented and effective; or
• have a validated CCP(s) in place to control the hazard associated with E. coli O157 (e.g., all raw beef received is used in the production of fully cooked product or is subjected to a validated full lethality treatment).

The operators should also consider using additional process controls as part of a multi-hurdle approach to reduce the E. coli O157 contamination to below detectable level.

5.3 Mandated testing of PM that are used for the production of FRGBP

Products subject to mandated testing include trim (trim derived from primal and sub-primal cuts), head meat, cheek meat, weasand meat, hearts and finely textured beef because they are commonly used in production of FRGBP. As such, operators who generate these PM will be required to test these products. Unless demonstrated otherwise by the operators (through documented evidence and controls), the PM will be considered as potential input material for the production of FRGBP and must be tested as described below.

Additionally, if other less common raw beef components such as primal or sub-primal cuts (e.g. top round, sirloin cuts) are destined for use in the manufacture of particular types of ground beef (e.g. ground round or sirloin burger), these components must be tested.

The testing frequency for PM has been determined according to the volume of production (see sections 5.3.1 or 5.3.2) which may have an impact on the level of health risk to consumers. The tests used must satisfy the conditions set out in Appendix 2 of this policy document.

Operators must therefore establish their yearly production volume (in kg) for the trim, as well as for any of the other PM they produce, or a combination of PM that may be used in the production of FRGBP. When only a part of the production of a given product or a combination of products is intended for the production of FRGBP, operators only need to test the product that will be used for that purpose. This must be adequately documented in the operator's HACCP plan. For instance, records indicate that 30% of the hearts or a combination of hearts and cheek meat produced yearly are used for the production of FRGBP, then the hearts or the combination of hearts and cheek meat making up that 30% of total yearly production must be tested at the applicable frequency.

When the use of the PM cannot be verified by the CFIA (e.g. product is sold outside of the federally registered sector), it will be considered as a potential input for the production of FRGBP and must be included in the production that falls under the testing requirement. Exemptions from mandatory testing will only be considered by the CFIA if there are suitable controls (documented evidence in HACCP) in place to ensure that these components are not used to make FRGBP, and provided that the receiving facility is a federally registered establishment.

Operators must keep this information current, accurate and available to CFIA. In case of any increase in the production volume, operators must determine if it impacts the testing frequency of a given product.

5.3.1 Production of 25,000 kg or less per year of a given product

Considering the small volume of product this represents, operators need to test this product three (3) times per year in accordance with the N-60 sampling procedure. High prevalence season (from April to September), should be considered when determining when to test in the year.

5.3.2 Production of over 25,000 kg per year of a given product

Operators must implement a robust testing protocol for each production lot of any product that falls under this category.

A minimum of 60 sub-samples must be examined per lot.

A lot cannot exceed 5 combos and cannot weigh more than approximately 4,500 kg. An alternate unit to a combo may be defined and used by the operator (e.g., a pallet of boxes or a tote, buggy, vat, tub, etc.), provided the weight of the lot does not exceed approximately 4,500 kg and does not exceed an entire day's production.

All combos/units must be equally represented in the sample. For example, a minimum of 12 individual pieces would be taken from each combo of a 5 combo lot. For alternate units, a minimum of 60 equally distributed pieces must be collected across the lot (e.g. a 10 vat lot of trim could be sampled by collecting 6 pieces per vat, a 5 pallet lot could be sampled by collecting 12 pieces per pallet, etc.).

A minimum of 325 g of material from each lot shall be collected and submitted for testing. At least 65 g of material (12 pieces weighing 5 or 6 g each) would be collected from each combo in a 5-combo lot. For alternate units the amount of material collected from each unit would depend on the number of units in the lot but would still be made up of 5 or 6 g pieces collected equally from each unit, adding up to a minimum total of 325 g for the lot. The material collected for testing should represent the outside surface of the product (e.g. carcass surface for sampled trim, exposed surfaces of the heart muscle, external aspect of the diaphragm muscle, etc.). In other words, it must not be taken from inner meat tissue unless the normal production process has left only inner tissue to sample.

Notes:

1. In the N-60 sampling procedure, 60 represents the minimum number of sub-samples required, regardless of the size (number of combos/units) and weight of the lot. Whether the lot weighs 4,500 kg, 2,000 kg or 100 kg, 60 sub-samples must be collected from the lot.
2. Alternate parameters may be used to define robust testing provided they have been evaluated by the CFIA (the Meat Programs Division and Food Safety Division) and provide an equivalent or increased level of confidence.
3. For PM not amenable to excision sampling (e.g. finely textured beef), a minimum of 5 sample units of approximately 65 g each must be collected from a lot which cannot exceed approximately 4,500 kg. Sample units must be representative of the whole lot. Alternate parameters may be used to define robust testing according to note 2 above.

5.4 CFIA risk-based verification sampling of PM intended for use in production of FRGBP (M218)

This sampling plan reflects a risk-based approach based on factors such as seasonality (April to September), production volume and historical testing and inspection data. It has been designed for establishments that produce PM to verify the effectiveness of their control measures for E. coli O157.

Products targeted for sampling under this plan are PM, preferably trimmings and chucks that may be used as components of FRGBP. PM intended to be used in further processing into ready-to-eat products are not to be sampled under this plan. The inspector should sample only PM produced at the establishment. If the operator commingles PM from different suppliers, the sample is to be collected before the operator commingles PM.

On the basis of production volume, the establishments have been divided into four categories: extra large, large, medium and small. Current sampling frequency is outlined in the National Microbiological Sampling Plans and Assessment Criteria (available only to CFIA staff). Generally, all establishments will be sampled at a normal frequency. A compliance history including a positive E. coli O157 result from M218 testing of PM or M201 testing of the product downstream will be taken into account when placing an establishment on enhanced frequency of testing for the next 120 days. Such a decision will be taken by the area program specialist.

The samples are analysed for the presence of E. coli O157 and generic E. coli enumeration. Refer to the "National Microbiological Sampling Plans and Assessment Criteria" (available only to CFIA staff) for current information.

The CFIA strongly recommends holding the entire lot pending the result of analysis. In order to provide the establishment enough time to hold the lot, the CFIA inspector will notify the operator (24 hours notice) about the planned sampling. He will collect the sample using the N-60 method described in the table below after the lot has been assembled according to the operator's lot definition and section 10 of this policy.

For PM not amenable to excision sampling (e.g. finely textured beef), a minimum of 5 sample units of approximately 200 g each must be collected from a lot that has been assembled according to the operator's lot definition and section 10 of this policy. Sample units must be representative of the whole lot.

5.4.1 N-60 method of sample collection

If a specific production lot is composed of more than 5 containers of PM, the CFIA inspector will randomly select 5 containers for sampling. Number of pieces to be collected per combo (container) is as follows:

Using a sanitized knife, inspector will aseptically cut off a thin slice that is approximately 50 cm² for each of the 60 pieces [e.g. 10 cm (4") x 5 cm (2") x 0.3 cm (1/8")]. The priority is to collect samples from the outer surface of PM. Inspector will place the sampled slices into the sterile plastic sample bag and weigh the sample to ensure that approximately 1 kg (2 lb) was collected.

5.4.2 Follow-up on E. coli O157 positive results under sampling plan M218

When E. coli O157 is detected in a sample, the sampled lot is considered adulterated. The inspector will verify whether the establishment has held or shipped the positive lot. If the lot has been shipped outside of the federal system, the CFIA inspection staff will notify the Office of Food Safety and Recall (OFSR) immediately.

The CFIA will verify that the positive product is transported and disposed of in an appropriate manner and the investigation is performed by the operator in accordance with Section 12 of this policy.

The CFIA will also determine the number of follow-up samples to be collected at the establishment producing PM and supplying establishments based on the investigation.

Note:

In case the same lot is also sampled and tested by operators, the lot cannot be released into commerce on the basis of a negative result obtained under the operator's program if the result obtained under the CFIA program is positive for E. coli O157.

5.5 CFIA verification sampling of FRGBP (M201)

This sampling plan has been designed for all federally registered establishments that produce FRGBP. The samples are analysed for the presence of E. coli O157 and generic E. coli enumeration.

Before collecting the sample, the inspector must ensure that the lot meets operator's lot definition and the criteria set out in section 10 of this policy. When combo(s) of PM is selected for grinding, the whole combo(s) must be used for the sampled lot.

From a selected lot of FRGBP, the inspector will randomly select and collect sample, preferably at the grinding stage. Inspector will collect 5 sub-samples of 200 g each, such that they are representative of the entire lot. The inspector will prepare the sample for shipping according to CFIA standards using sufficient ice packs and insulated packaging material.

Note:

If any other ingredients are placed through the grinding stage along with the beef trims, inspector should consult with the laboratory as to the expected turn around time for results. It is known that ground beef/veal samples containing spices cannot be assessed using the rapid screening method and must be analyzed using the cultural method. As such these test results will not be available for 4-5 days and the operator will be required to hold the sampled lot for this period of time.

Refer to the "National Microbiological Sampling Plans and Assessment Criteria" (available only to CFIA staff) for current information.

5.5.1 Operator's follow-up to E. coli  O157 positives under CFIA verification sampling of FRGBP, sampling plan - M201

Operators must conduct follow up sampling at their own expense at the rates described below in response to a positive E. coli O157 result reported under this CFIA verification sampling plan. These samples are to be collected under CFIA supervision and to be tested in an accredited laboratory using a CFIA recognized testing method. Recognise methods can be found in Appendix 2 of the policy.

Following maximum and minimum sampling limits must be respected:

• a maximum of 2 follow-up samples per shift per day, separated in time, unless the establishment cannot continue to operate under that sampling frequency (e.g. because the establishment cannot fill orders and hold all sampled product);
• a minimum of 3 follow-up samples per week unless the establishment produces the product in question less than three times per week or unless the establishment cannot continue to operate under that sampling frequency.

From a selected lot, the operator will randomly select and aseptically collect 5 sample units of 200 g each, using sterile gloves, sterile plastic sample bag and sanitized equipment (knife, hook, tongs, etc.). The operator will prepare the sample for shipping according to CFIA standards using sufficient ice packs and insulated packaging material.

If there is a positive sample during the follow up testing, the operator is required to continue sampling until 8, 12 or 16 consecutive negative samples (see Table 2, above) have been collected.

Accredited laboratories must report results to the operator as well as the Food Safety Division, CFIA.

The operator needs to notify the IIC upon the receipt of results.

Product must be held when follow-up testing is conducted. For notification of results to CFIA and supplying establishments refer to section 12.3.2.

5.6 Raw beef used to manufacture fully cooked products or used in a process that includes a full lethality treatment

During the hazard analysis, the operator may determine that the hazard related to E. coli O157 is likely to occur, but that no new CCP(s) is/are required in the establishment because all products are subjected on site to a full lethality treatment or shipped directly to another federally registered establishment where similar steps are taken to control the hazard.

Where this occurs, the operator must provide details on the control measures in place to ensure that the product receives the full lethality treatment and that it is not diverted elsewhere (i.e., how inventory is managed, if boxes have special markings [e.g., "For Cooking"/"For Full Lethality Treatment"]). The measures need to be incorporated and evaluated for compliance and effectiveness within the operator's HACCP system.

If presented, the CFIA will evaluate the request of an operator for approval of a control program for movement of untested product to an intermediary registered establishment prior to being shipped to an establishment that will provide the full lethality treatment. The initial shipping establishment must have a tracking system that will allow the CFIA to verify that all untested material was given a full lethality treatment in the final federally registered establishment.

In the case of product sent for a full lethality treatment after it has been tested and found positive for E. coli O157, or product that the operator has chosen to treat as positive for the pathogen based on screening results, refer to section 12.2.1 of this policy.

5.7 Production of non-intact raw beef products

When manufacturing non intact raw beef products, the equipment used may transfer bacteria from the surface of a contaminated cut of meat to its interior, as well as cross contaminate subsequent portions processed by the same equipment.

Non-intact raw beef products may not pose the same level of health risk to consumers as PM. The certificate of analysis (COA) or verification testing is not mandatory for such products.

However, operators of federally registered establishments producing or handling non intact raw beef products must identify E. coli O157 as a hazard likely to occur in such products and demonstrate that their HACCP system effectively manages this hazard. The HACCP plan(s) must include control measures according to "industry best practices". These may include an agreement with supplier, assessment of sanitation process, verification of sanitation process by equipment sampling, use of antimicrobials and proper labelling with validated cooking instructions.

If such products are found to be associated with a positive E. coli O157 result, the operator must notify the CFIA immediately and take appropriate corrective and preventative measures. The CFIA will assess such incidents on a case-by-case basis.

5.7.1 Labeling requirements for mechanically tenderized beef

In addition to the basic labeling requirements, mechanically tenderized beef intended to be sold at retail or to hospitals, restaurants and institutions must be identified as "mechanically tenderized" on the label. Cooking instructions to cook mechanically tenderized beef to a safe internal temperature must be provided on the label. Health Canada currently recommends cooking mechanically tenderized beef at least to an internal temperature of 71°C (160°F).

5.8 Raw intact primal and sub-primal cuts

Intact primal and sub-primal cuts used for purposes other than the manufacture of FRGBP do not pose the same level of risk as FRGBP. In contrast to FRGBP, the interior of these intact raw beef products is considered free of pathogens. Consequently, customary cooking of these products is expected to inactivate any E. coli O157 that might be present on surfaces. An establishment producing and distributing pre-packaged intact steaks may conclude it does not need to change its HACCP system for these products.

Trims generated during the manufacture of primal and sub-primal cuts must be subjected to controls to ensure that the end products in which they will be used do not contain detectable levels of E. coli O157. As trims are commonly used for the production of FRGBP, they must be tested at the frequency prescribed under section 5.3 of this policy. Alternately, these trims may be used for the production of a product that will be cooked or subjected to a full lethality treatment in a federally registered establishment and under a HACCP system that addresses the potential E. coli O157 contamination hazard.

Should any part of beef cuts such as primal and sub primal cuts, including boneless chucks that are generally not destined for the production of ground beef, be used by an operator for that purpose, it must be subjected to robust testing applicable to PM. This testing may be done by the operator producing these products (at their buyer's request) in accordance with section 5.3 of this policy, or by the operator receiving these products in accordance with sections 5.3 and 9.2 c of this policy.

5.9 Beef products processed for raw consumption (BPPRC)

BPPRC are the products that are prepared in establishments and may be pre-packaged as products that are either intended to be consumed or are likely to be consumed raw. Examples are carpaccio and steak tartare.

Operators producing or intending to produce BPPRC must develop, implement and maintain a CFIA approved control program including N-60 testing of PM and/or finished product. Wherever possible PM should be tested according to section 5.3.

However, if testing of PM is not practical, a minimum of 60 sample units should be examined per lot of BPPRC. Sample units must be representative of the whole lot which cannot exceed approximately 4,500 kg. A total of 325 g per lot should be analyzed for E. coli O157. Approximately 6 g of material from the meat surface of each piece (in other words, not sterile inner tissue) should be examined. Alternative sampling protocols may be used, provided that they are as/or more rigorous and have been approved by the CFIA.

6. Purchase specifications

Operators receiving raw beef must include the following purchase specifications in their HACCP system:

1) A letter of guarantee (LOG) from supplying companies must be on file identifying the validated intervention(s) (CCP) as well as other measures used to reduce, prevent or eliminate the hazard associated with E. coli O157 for all beef products. The LOG must be dated and signed by the operator, or a designated person, of the supplying establishment. For establishment shipping product pending result, the LOG must include a statement to the effect that when the supplier obtains an unsatisfactory result, the operator(s) of the establishment(s) which have received implicated product as well as the supplier's local CFIA inspection staff will be notified.

Note:

1. Once notified, the operator(s) of the receiving establishment(s) must then inform their respective IIC.
2. A COA, although providing an increased level of confidence, cannot be used in lieu of a LOG. The latter provides confirmation that the process under which the product was manufactured is under control and that the purchase specifications are met. A COA provides additional information about testing results for a specific lot.

2) As a verification step for the receiving CCP or Purchasing/Receiving/Shipping Prerequisite Program B 2.1, the receiving operator must verify that the supplier is applying effective measures for the control of E. coli O157 in PM used in the manufacture of FRGBP and is doing all testing required in section 5.3 of this policy.

The receiving operator may achieve this by specifying the following requirements within purchase specifications for each of their suppliers:

2.1) For suppliers producing more than 25,000 kg per year of raw beef trim and ground meat components (per category):

• The supplier performs robust testing on each production lot of raw beef trim and ground meat components in accordance with section 5.3 of this document;
• Upon approval by the CFIA, an alternate method which is able to provide the same level of control as use of the COA may be considered;
• The supplier provides a COA for each lot of raw beef trim and ground meat precursors sent to the receiving establishment. The COA must indicate results of testing for the specific lot being sent. The result of the test must be negative (E. coli O157 not detected);
• The supplier is subjected to third-party audits and can demonstrate through the audit reports that the requirements in section 6 Purchase specifications are being implemented in a suitable manner. Third-party audits must be conducted by a firm, organization or individual who has a proven track record in performing this type of review. On an exceptional basis and with the agreement of the supplying establishment, an audit may be conducted by an employee of the receiving establishment. The CFIA reserves the right to reject an auditor who fails to suitably demonstrate that he or she meets the above criteria or who refuses to provide reports to the CFIA upon request.

Note:

If all of the above requirements are met, the receiving establishment would not need to perform verification testing of received lots of raw beef trim or ground beef components.

2.2) For suppliers producing 25 000 kg or less per year of raw beef trim and ground meat components (per category):

A letter to this effect is provided by the supplier and kept on file by the purchaser. The supplier also confirms on a yearly basis that the required tests have been performed.

7. Cross-contamination potential

In the case of an operator handling both raw beef products considered to be below detectable level for E. coli O157 and beef products potentially contaminated with E. coli O157, the operator must develop and implement a written segregation protocol to ensure that raw beef products received for cooking or other full lethality treatment are not used in the production of FRGBP and that cross-contamination is prevented.

The appropriate information must be reflected on FSEP forms (Potential cross-contamination). The segregation procedures must include monitoring, verification and deviation procedures as well as record keeping, and be auditable and effective. For example, a letter of guarantee regarding E. coli O157 is required only for products received for raw ground beef product manufacturing and not for those received for further cooking or other full lethality treatment. For audit purposes, these segregation procedures would be audited under the appropriate Prerequisite Program.

8. Validation of pathogen reduction step(s) - abattoirs

Validation of in-plant pathogen reduction steps (CCPs) must be performed according to the FSEP approach. FSEP defines validation as: obtaining confirmation that the elements of the HACCP system are complete and effective in controlling biological, chemical and physical hazards. This may include ingredient sampling, end product sampling, etc. In this case, this means more specifically doing all of these three steps:

Step 1

Gather published scientific information on the experimental reduction effect of the chosen pathogen reduction intervention and all relevant critical factors (e.g., intervention "Y" should result in a 2.0 log reduction considering parameters defined under experimental conditions [pressure, temperature, time, chemical concentration, etc.]).

Step 2

Demonstrate the effectiveness in reducing a suitable surrogate level by "X" logs under in-plant operational conditions (e.g., generic E. coli or Enterobacteriaceae as indicators are reduced by "X" logs by intervention "Y") for each of the interventions that the operator has chosen to implement. An appropriate surrogate is an organism that has a heat resistance, growth range, pH range, ability to grow on selective media, etc. that is similar to E. coli O157. This is normally done by comparing surrogate levels in a sample taken before and after the intervention. The operator shall choose a statistically significant sample size over a period of four (4) months to demonstrate that on-site intervention is achieving the log reduction targeted. E. coli O157 must not be introduced for experimental purpose in registered establishments. Please see Appendix 1 for guidance on sample size, required number of samples (Table 1) and statistical analysis, and refer to Annex T of the United States section of Chapter 11 of the Meat Hygiene Manual of Procedures for the complete information about the testing procedure.

Step 3

The sum of the log reductions from all interventions (CCPs) must result in a final product that is below the detectable level for E. coli O157. The demonstration should be based on a statistically significant number of finished product samples i.e., provide a 95% level of confidence that E. coli O157 is below detectable level. Because of the expected variability of E. coli O157, this validation sampling should be done by randomly taking the required number of samples over one (1) month of production as indicated in Appendix 1, Table 2. Based on the current prevalence of E. coli O157 in raw beef products, sampling plans used for validation of controls must use 0.1% expected prevalence. Sampling of carcasses should be conducted in accordance with Annex T of the United States section of Chapter 11 of the Meat Hygiene Manual of Procedures for the complete information about the testing procedure.

Note:

1. If the lot size (i.e., number of animals slaughtered) falls between two values appearing in column 1 of Tables 1 and 2, the highest number should be selected in order to determine the sample size.
2. When testing for surrogate organisms, it is the prerogative of each individual company to decide if they want to use an accredited laboratory or not. Acceptable methods can be found in the Health Canada Compendium of Analytical Methods. The "application" section of the method chosen must be appropriate for the intended purpose.
3. Testing for E. coli O157 for validation or verification purposes must be performed in accredited laboratories using methodology as described in Appendix 2 of this Annex.
4. Laboratory results (laboratory tests certificates) shall indicate which method has been used to analyze the samples.

Validation is to be completed initially to demonstrate that the interventions put in place by the operator are effective in producing safe meat products (i.e. that are below detection level for E. coli O157). Validation has to be conducted again when the operations have changed substantially and/or pathogen intervention(s) has (have) been added or modified in a novel manner. The plant management and VIC/IIC may consult with the area program specialist regarding the need to revalidate a pathogen reduction intervention.

9. Verification activities and testing

9.1 Abattoirs (with or without boning/cutting operations)

Pathogen reduction steps are expected to become part of a CCP when passed through the decision tree (FSEP form # 8). Verification must therefore be performed in accordance with the FSEP Program. In this case, this means more specifically that:

1. records are reviewed;
2. on-site reviews are conducted to ensure that the written program is implemented as planned; and
3. an appropriate level of random product testing is done on a routine basis for E. coli O157 or a surrogate organism. For the purpose of this testing, a lot may be defined as one carcass that is randomly selected for testing. The carcass, as well as one carcass preceding and following the tested carcass, should be held pending the receipt of results. In the event that the result for the tested carcass is positive, all three carcasses are considered as contaminated (unless tested as indicated in section 10 of this policy) by the pathogen and must be subjected to a full lethality treatment or denatured and condemned.

When operators decide to test for a surrogate organism, they will have to meet the following conditions:

• Justify the choice of the surrogate organism used for that purpose;
• Indicate what level of that organism will be considered as a non-compliance (critical limits) and provide the rationale for that decision;
• Include in the HACCP system the corrective actions that will be implemented when the concentration of the surrogate organism is above the critical limit;
• This option may be used after the CFIA has reviewed and accepted the operator's proposal for the use of this surrogate organism.

The following minimum testing frequencies have been established for the purpose of carcass sampling:

The operator is responsible for developing and maintaining the sampling procedure that will meet the sampling plan's objectives and specifications. Operators may refer to Annex T of the United States section of Chapter 11 of the Meat Hygiene Manual of Procedures for guidance on sampling procedure.

Acceptable laboratory methodologies are listed under Appendix 2 of this policy.

Note:

Abattoir operators that can provide test results from downstream E. coli O157 carcass testing performed by affiliated operators at receiving of their carcasses may be exempted from carcass testing in their own establishment as indicated under section 5.1.1 of this policy.

9.2 Receiving establishments

Operators receiving raw beef products must ensure that the products received meet their purchase specifications either by a Receiving CCP or the Purchasing/Receiving/Shipping Prerequisite Program (B 2.1). Verification must therefore be performed. In this case, this means more specifically that:

1. records are reviewed;
2. on-site reviews are conducted to ensure that the written program is implemented as planned;
3. the operator has verified that the PM used in the manufacture of FRGBP have all been tested as required under section 5.3 and meet all the requirements under purchase specifications in section 6 2) of this policy. In such cases, verification testing of PM at receiving is not required. When receiving untested raw beef products that will be used for the production of FRGBP, the receiving establishment must conduct verification testing at receiving. The following requirements apply:
   • product tested must be from a single supplier;
   • no co-mingling of lots can be done (each lot must be sampled on its own);
   • product must be tested under robust testing (or an alternate protocol that is equivalent or better); and
   • presumptive positives must be confirmed;
4. auditing of suppliers (optional) may also be performed as a verification activity.

It is the operator's responsibility to implement these activities at a frequency which will ensure that finished product will meet applicable requirements.

10. Lot considerations

The operator must define the lot in their written program for the purpose of sampling carcasses, PM and FRGBP for E. coli O157 using the following guidelines:

• In case of carcasses, the lot is defined as whole carcass or both sides of a single carcass. However, to address cross-contamination potential when a carcass is found positive for E. coli O157, the tested carcass as well as the one carcass preceding and following the tested one will be considered as positive product, unless a laboratory analysis can demonstrate that E. coli O157 is below detectable levels using a specific methodology listed in Appendix 2;
• In case of PM and FRGBP, a lot is defined as comprising all cartons, packages or containers either:
  1. produced under the same conditions at one establishment from one effective clean-up and sanitation to the next effective clean-up and sanitation; or
  2. determined by the operator when implementing a statistically based sampling program (robust testing or alternate sampling protocol accepted by the CFIA). The operator must have an acceptable rationale that supports the alternative lot definition. Such lots will be referred to as redefined lots;
• A lot cannot exceed a single day's production. Further, a redefined lot cannot exceed 5 combos and cannot weigh more than approximately 4,500 kg;
• Before taking a sample for E. coli O157 testing, the operator must isolate and clearly identify the lot according to their written program and to the satisfaction of the CFIA inspector. It is strongly recommended that the lot, and any raw product manufactured from the lot, be detained pending receipt of laboratory results. The operator must further identify the supplying establishment number (if product received from another establishment), the production date, production lot number and any other relevant data available about the lot;
• FRGBP processors may have more than one grinding line for the production of FRGBP in their facility. The following conditions must be met for processors to consider that their FRGBP production lines are segregated, thus limiting the size of the lot to a specific line:
  • The entire lot of any precursor material that is used for the production of FRGBP must be processed on a single line (in other words, a given lot of PM cannot be split between two grinding lines);
  • Each grinding line must be clearly defined with regard to its equipment and the product flow;
  • Process controls must be in place to prevent cross-contamination between the different production lines;
• For other raw beef products, if no satisfactory scientific basis is provided by the operator for lot definition, the default lot considered by the CFIA will be the product produced under the same conditions at one establishment from one effective clean-up and sanitation to the next effective clean-up and sanitation; and
• It should be noted that if an operator has an acceptable rationale that supports an alternate lot definition and regularly tests specific lots of product for E. coli O157, this information could possibly be a basis for determining whether one E. coli O157-positive lot will implicate other lots produced on the same day.

11. Transit of raw product within the federal system

When an establishment receiving raw beef is in turn supplying raw beef products to another establishment, the following guidelines apply:

• Both operators must have purchase specifications to prevent E. coli O157 from entering their facilities, which could be based on those described in section 6 of this policy, as part of their verification activities. In this case, the LOG provided by the establishment that receives and ships raw beef should state that the operator has a control in place to address the risk associated with E. coli O157 at receiving (CCP or pre-requisite program), and has LOGs from all of their suppliers on file;
• In addition to purchase specifications addressing E. coli O157 referred in section 5.2, receiving establishments must ensure that measures are in place to prevent E. coli O157 growth or contamination after product is received.

For additional information on the LOG, please refer to section 6 of this policy.

11.1 Shipping of tested product pending receipt of results

Operators must maintain control over tested products until results are obtained. This will prevent tested products of unknown status from being offered to the public until they have been found satisfactory.

Operators who elect to ship tested product prior to being informed of the laboratory results must therefore inform the recipient that the product cannot be used before they are notified that the results were satisfactory. When a product has been distributed prior to its testing results being known and the test results indicate a presumptive or confirmed E. coli O157 result, the CFIA must be immediately notified and the incident treated as a potential recall situation. The CFIA strongly advises against this procedure.

When tested product is shipped pending reception of results, the following conditions apply:

1. The shipping establishment must:
   • have a written protocol in place;
   • maintain complete records identifying the type of product, as well as the quantity, being shipped;
   • identify the product in an appropriate way;
   • control product while it is in transit - the use of company seals is mandatory; and
   • get confirmation from the receiver that the product (specifying the type and quantity) was received;
2. The receiving establishment must:
   • have a written protocol in place;
   • maintain complete and accurate records of all products received under these conditions; and
   • keep the product segregated until the laboratory results are obtained.

Note:

1. The above requirements also apply to tested product shipped outside of the federally registered system pending receipt of results.
2. Operators cannot export tested raw beef products pending receipt of results.

For shipping of positive or presumptive positive products to other federally registered establishments, refer to section 12.2.

12. Positive results for E. coli O157

Unless otherwise specified within the policy (e.g. see section 9.2 c), presumptive positive results may be considered as positive results by the operator. When this is the case, the measures that apply are the same as if the laboratory result had been a confirmed positive. When this presumptive positive result impacts on another establishment (e.g. product tested at the receiving step was a presumptive positive), the operator performing the test must have a prior agreement with the supplier as to whether a presumptive positive is accepted as a positive result, or a complete laboratory confirmation will be performed to determine either a positive or negative result (please refer to Appendix 2 of this policy). All registered establishments who purchase raw beef must document prior agreements in their HACCP systems. This ensures that product disposition and follow-up for both the supplier establishment and purchasing establishment can proceed expediently when a presumptive positive result is obtained.

When obtaining positive results for E. coli O157, whether confirmed or considered as positive, the operator must take immediate action, in accordance with sections 12.1 and 12.2 of this policy.

12.1 Scope of implicated product

A) When an operator is implementing a sampling protocol under a statistically based sampling program (robust testing/redefined lot approved by the CFIA):

PM: The redefined lot of PM that tested positive and any lot of PM, FRGBP or of a prepared meat product that contains a portion of the PM from the redefined lot that tested positive will be considered implicated. Additionally, CFIA expects operators to consider other potentially implicated combos (i.e. like-combos) based on factors such as the source of PM, time of production, production line and high event period (refer to section 14). The CFIA recognizes that the rationale and decision in which the operator determines potentially implicated products can be complex. However, the basic principles of this activity must be a part of the written program and be made available to CFIA staff upon request.

FRGBP: The redefined lot of FRGBP that tested positive as well as one or two lots produced immediately before and after the positive lot will be considered implicated. Any lot of FRGBP or of a prepared meat product that contains a portion of the redefined lot that tested positive will be considered implicated.

B) When an operator is not using a redefined lot, a sampling protocol is not in place or the operator is not implementing a robust sampling protocol according to this policy:

PM: All the PM produced under the same conditions as the tested product at one establishment from one effective clean-up and sanitation to the next effective clean-up and sanitation would be implicated.

Any lot of PM, FRGBP or of a prepared meat product that contains a portion of the implicated PM (as described above) will also be considered also implicated.

FRGBP: All the FRGBP produced under the same conditions as the tested product at one establishment from one effective clean-up and sanitation to the next effective clean-up and sanitation would be implicated.

Any lot of FRGBP or of a prepared meat product that contains a portion of the implicated FRGBP (as described above) will also be considered also implicated.

The scope of implicated products described in this section may be further expanded depending on the following indications in the overall context of the situation:

• epidemiological evidence;
• High Event Period (HEP) situation;
• out-of-control process; and
• inadequate lot identification and sampling protocols.

Note:

Intact raw beef linked to same source material as PM that has tested positive, may be implicated, on a case-by-case basis, depending on their end use (e.g. FRGBP, mechanically tenderized beef, BPPRC), when there is epidemiological evidence of illness, process deviations, or systemic HEP situation.

12.2 Product disposition

The following options for product disposition, which must be conducted under the CFIA's authorization and supervision, are available. Whichever option is selected, the traceability component must be covered in detail in the operator's HACCP system.

12.2.1 Product salvage - cooking

Salvage the product by further processing into a fully cooked finished product within the federally registered sector through a validated cooking process (full lethality treatment). If done in a different establishment, it must be transferred under company seal and the adulterated product must be sent directly to an establishment that provides the heat treatment to the product; appropriate records must be kept by all involved operators to ensure complete control over the product until the risk has been addressed. In all cases where product is salvaged, the following requirements apply:

• The cooking process must be validated and approved by the Area Program Specialist (meat processing) before being used;
• The operator must maintain a complete and up-to-date inventory of all products being salvaged because it was found positive for E. coli O157 or because the operator has chosen to treat it as positive for E. coli O157. The inventory shall include the following information: initial lot numbers and test results, type of product, weights, time/date of cooking and lot numbers of finished products.

When positive product is shipped to another federally registered establishment for cooking, the following conditions apply:

1. The shipping establishment must:
   • maintain complete records identifying the type of product, as well as the quantity, being shipped for cooking.
   • label the product with "For cooking only" (stamp or sticker). If the product is being stored in an off-site registered storage at the time that the E. coli O157 test results are known, the product can be labelled at the storage facility, with the consent of the CFIA inspector.
   • control product while it is in transit. The use of company seals is mandatory.
   • get confirmation from the receiving establishment that the product (specifying the type and quantity) was received. (Note: this should be stated in the operator's protocol.);
2. The receiving establishment must:
   • confirm the receipt of positive product with the shipping establishment.
   • maintain complete and accurate records of all positive products received for cooking process.
   • meet the above requirements applicable to this procedure (approved cooking process and inventory of product).

Note:

1. Operators may ask the CFIA to permit the storage of positive product prior to subjecting it to the full lethality treatment (cooking). Any such request must be presented to the Area Program Specialist for evaluation. If accepted, the HACCP systems of the shipping establishment, the storage establishment and the receiving establishment must address this situation. Appropriate controls must be in place and monitored (including but not limited to product inventory, segregation procedures, seals used for transit, etc.).
2. The operator may choose to salvage a potentially contaminated product (e.g. potentially implicated product which has not been confirmed to be below detectable level of E. coli O157 ), however, all above controls for positive product shall apply.

12.2.2 Denature and condemn

Denature and condemn the product under the direct supervision of the CFIA.

12.2.3 Reject

In the case of a product received from another registered establishment, the operator may reject the product and, providing the supplier has agreed in advance, return the product to the supplier under company seal for appropriate disposition. Records must be kept by both operators to ensure that the positive product is adequately controlled until it is subjected to either one of the two options described above.

12.3 Follow-up actions - operators

Any raw beef product that is positive or presumptive positive for E. coli O157 is considered adulterated by the CFIA and the operator must inform the CFIA immediately. For auditing purposes, the information must be presented to the CFIA in written form.

12.3.1 Positive results obtained from products produced at the establishment

Operators using a robust testing protocol as a process control will very likely get more positive results than operators who test only for verification purposes. The CFIA still expects these operators to comply with sections 12.2 and 12.3 of this policy.

Operators must summarize and analyze all laboratory results (CFIA testing and industry testing) on a daily basis (see section 13) and develop criteria that would point towards a situation where the number of positive results exceeds what is typically expected within a certain timeframe (see section 14). The operator must evaluate the impact of any positive result, investigate any potential link between positive results and take additional measures (e.g. pathogen reduction steps) to control the risk posed by E. coli O157, should the result of such investigation warrant.

The operator must take positive results as evidence that their HACCP system has been ineffective in producing a product where E. coli O157 that is below detectable levels. Consequently, the operator must immediately notify the inspector in charge (IIC) /veterinarian in charge (VIC) of the establishment, who will in turn notify their Inspection Manager and the Area Program Specialist. The operator must also take the following actions:

1. Ensure that any affected product is under control.
2. Investigate the cause of the deviation by evaluating, as appropriate:
   1. all applicable HACCP controls;
   2. sanitation procedures (Prerequisite Programs);
   3. any other pertinent documentation and procedures.

      Note:

      1. This also includes beef slaughter activities, whether these take place on site or at another registered establishment.
      2. While each positive must be investigated in an effort to identify the probable cause and to guide corrective measures, the scope and depth of an investigation may be adjusted according to current circumstances, taking into account such factors as the frequency and number of positives being found (process awareness), and the type(s) of product(s) affected.
3. Apply a corrective action to eliminate the cause of deviation.
4. Ensure that the corrective action has brought the CCP and/or Prerequisite program(s) under control.
5. Perform a food safety assessment on the affected product and determine if other products were implicated. Determine the appropriate disposition.
6. Implement preventative measures to prevent recurrence of the deviations. Serious consideration should be given to increase the efficacy of the pathogen reduction steps used in the HACCP system.
7. Verify the effectiveness of the preventative measures. When the pathogen reduction step has been substantially modified to increase its efficacy, the intervention must be re-validated. The plant management and IIC/VIC may consult with the area program specialist regarding the need to revalidate a pathogen reduction intervention.
8. For each corrective action and preventative measure, the designated employee must specify on the record: a target date for completion of corrective actions and preventative measures; the actual completion date for these corrective actions and preventative measures. Each entry includes the date, and is signed or initialled by the establishment employee making the entry.
9. Provide the inspector with the operator's action regarding the adulterated product explaining how the product will be handled, controlled, brought back into compliance or disposed of (denatured and condemned).

If no contamination source can be identified a report indicating this should be produced and made available to the CFIA, as well as the rationale used to reach that conclusion. The report should include:

• documents reviewed;
• procedure used to review;
• dates reviewed;
• evaluation of the impact of this positive result on their own testing results for this type of product; and
• appropriate signatures of reviewer(s).

In all cases, the CFIA inspector evaluates the investigation done by the operator in collaboration with the designated Area Program Specialist, Complex Supervisor and Regional Veterinary Officer. When the investigation's conclusion or the corrective actions taken are judged inadequate, a Corrective Action Request (CAR) is issued under Compliance Verification System.

12.3.2 Positive results obtained as a result of verification testing done on products received from a supplier

In the case where a positive result is obtained as a result of verification testing done on products received from a supplier, the following requirements apply:

• The operator having performed the test must immediately notify the Inspector in Charge and the supplying establishment of the unsatisfactory results. The supplying operator will then inform the IIC/VIC at their establishment of the unsatisfactory result. The IIC/VIC of the supplying establishment will inform their own Inspection Manager and Area Program Specialist of the situation;
• The receiving operator must dispose of affected product according to one of the three options presented under section 12.2 of this policy;
• The supplying establishment must treat this notification as evidence that their HACCP system may have been ineffective in producing a product where E. coli O157 is below the detectable level and must investigate the situation accordingly and take immediate follow-up actions according to section 12.3.1 above.

The Area Program Specialist will ensure that the necessary details are communicated to any other CFIA Area Staff who may need the information.

12.4 CFIA control of affected product

Any raw beef product that is positive or presumptive positive (positive screening test result) for E. coli O157 is considered adulterated by the CFIA. That is why the operator must inform the CFIA of this situation immediately. For auditing purposes, the information must be presented to the CFIA in the written form.

According to the Meat Inspection Regulations, 1990 (MIR):

20. (2) Where an adulterated meat product in a registered establishment can be made to conform to the standards prescribed by this Part for an edible meat product, the meat product shall be held by an inspector until it is made to conform to those standards by the operator.

130. (1) No person shall remove or alter an official seal or official tag applied by or under the authority of an inspector unless authorized to do so by an inspector.

(2) Any food animal, meat product or other thing being held on the instructions of an inspector shall not be handled or used in any way without the permission of an inspector.

The CFIA inspector will hold the product using held tag CFIA/ACIA 0093 until it is made to conform to the standards by the operators. The CFIA inspector will keep the appropriate information on file. The inspector must also indicate at the back of the held tag the sections of the MIR under which the product is detained.

The inspector must be provided with the operator's action with regards to the adulterated product explaining how the product will be handled, controlled, brought back into compliance or disposed of (denatured and condemned). Deviation records must be kept with all the needed information (e.g. product involved quantity, lot identification and laboratory result).

If an adulterated product is sent to another federally registered establishment for cooking, the HACCP plan of this (these) other establishment(s) must also address the needed controls. In such cases, the CFIA inspector of the other establishment must be notified on each occasion.

As per section 130 of the MIR, the inspector must either remove the held tag in person or authorize its removal. This must also be documented as it pertains to this specific issue in both the operator's HACCP plan and the inspector's files.

Note:

1. If the product was imported, the IIC through the Area Program Specialist will immediately notify in writing the National Specialist, Import Program, Meat Programs Division, in Ottawa. The National Specialist, Import Program will notify the authorities of the exporting country for a follow-up investigation.
2. If any product affected by the unsatisfactory result is in distribution, the inspection staff must inform the Office of Food Safety and Recall (OFSR).

13. Process awareness

Process awareness is intended to provide an in-depth understanding of the operator's overall effectiveness of control measures. It is achieved through examination of data in real-time and over time, for the purpose of detecting deviations or deviation patterns.

It can assist in identifying events where a potential loss of process control may occur or has occurred in day to day operations. Process awareness also involves the study of time and /or spatial patterns suggestive of unstable processes, contamination events and scoping of sources of contamination. It helps in forecasting activities, provides information on prevalence of E. coli O157 and its indicators. It is helpful in achieving an improved control over time and gaining experience in controlling E. coli O157.

Process awareness can be enhanced through routine recording of the chronology or specific times of certain operations, such as slaughter, fabrication, trim collection and grinding. This information can enable real-time or retroactive determination of relationships between source materials and between products, which can help in assessing the significance of positive test results.

Each slaughter and processing establishment handling raw beef must have a written program for process awareness which takes into account previous test results from operator's mandated and non-mandated testing, CFIA testing and any third party test results for this analysis. The conclusion drawn from analysis and the rationale must be made available to the CFIA. As a part of process awareness, the operator must establish criteria/limits to define periods when the analysis indicates a potential loss of control. The out-of-control process may include High Event period (HEP) situation described in section 14.

13.1 Requirements for process awareness in establishments producing FRGBP

Currently establishments producing FRGBP are not required to test for E. coli O157. Therefore, the operators must develop microbiological verification procedures for their specific operations, including the following characteristics:

• location: Operator may choose to conduct verification testing of incoming raw materials at receiving and/or finished product or may choose to utilize testing done by supplying establishments. But this testing must be in addition to the mandatory testing of PM for E. coli O157;
• target organism(s): Operator may select target organisms which provide the most useful information for determining control of the process; and
• sampling frequency: Operator may determine sampling frequency based on the specific operation.

The operator will be required to justify their choice of location, target organism and frequency in their written program. They will also have to determine and justify the criteria/limits that indicate a potential loss of control and follow up actions to be taken.

13.2 Operator's follow-up in case of out-of-control situations

Operators must notify CFIA inspection staff when the process is out-of-control. Additionally, they are required to perform root cause analysis and take appropriate corrective actions and preventative measures according to their written program. Even if no issue can be identified, a report indicating this fact and a rationale for this conclusion should be made available to the CFIA. Purchase specifications mentioned in section 6 of this annex may include an agreement or arrangement for the notification of suppliers and follow-up actions at their end.

13.3 CFIA verification activities

CFIA inspection staff will review and approve the establishment's written program. They will also verify that the testing, result assessment and follow-up actions are being implemented according to the operator's written program. They will perform a Compliance Verification System (CVS) task to verify establishment's corrective actions.

14. High event period (HEP) situation

HEP is a situation in which an establishment experiences a high number or rate of positive results for E. coli O157 in PM from production lots containing the same source material. That is, the PM was produced from one or more carcasses slaughtered and dressed consecutively or intermittently within a defined period of time (e.g., shift). A HEP situation may indicate systemic breakdown of the slaughter dressing operation that may implicate other parts of the beef carcass in addition to the PM that tested positive. It may indicate an unsanitary condition during the slaughter/dressing operation which may have resulted in widespread contamination across production lots.

As such, the negative tested trims obtained from same source materials as those which have tested positive may potentially be contaminated.

Supplying and receiving establishments conducting robust N-60 sampling and testing programs for beef trimmings must identify and document HEP criteria. This will allow for adequate identification of implicated and potentially implicated products beyond the products that were reported positive and this will minimize the chance of adulterated product being released into commerce.

For example, the demonstration of a temporal and spatial relationship between positive lots could indicate a potential loss of control over the process and a likely HEP situation. Here temporal means the time of processing based on recorded start and stop times of relevant operations and spatial refers to the area of carcass from which different types of trim and intact cuts are derived. Demonstrating a temporal and spatial relationship between positive and other lots could be used to determine the scope of implicated products.

Operators must distinctly identify whether the HEP situation is localized or systemic prior to determining scope of implicated product and the rationale for this decision must be presented to the CFIA.

A localized situation may affect only the product of one lot due to an isolated problem while a systemic situation may affect more products due to a broader problem.

The CFIA requires establishments to take action if sampling of PM produces a positive rate statistically significantly greater than 5%. While developing HEP numerical criteria, establishments may opt for 95%, 99% or 99.95 % confidence level. Refer to Appendix 3 for suggested HEP numerical criteria.

Establishments may develop alternate proposals for defining HEP criteria based on temporal and spatial relationships between positive lots. Such proposal must be presented to the CFIA for approval.

14.1 Expected actions in HEP situations

In addition to actions described in section 12.3.1, operators need to conduct further investigation to determine if the negative-tested lots of trimmings and intact primal and sub-primal products produced from the same source material as the positive- tested trimmings have E. coli O157 reduced to below detectable level.

During a HEP situation, primal and sub-primal cuts linked in time and space to positive lots of trim may be considered adulterated because they were prepared under unsanitary conditions and therefore might be expected to have higher than normal levels of contamination. Some of the factors considered in this determination will be:

• rate of PM E. coli O157 positives during the HEP;
• magnitude of process deviations during the HEP;
• microbial test results on primal and sub-primal cuts, if applicable, and other relevant test results;
• interventions, for example: antimicrobials applied to primal and sub-primal cuts;
• cross contamination controls; and
• end use of primal and sub-primal cuts.

The actions taken in response to a HEP situation will depend upon the findings of the investigation of the positive results. The operator may consider increased testing when it experiences HEP, however if the establishment is able to find the root cause of the HEP and takes corrective actions to prevent positive results from recurring, then an increase in testing will not be needed. This increase in testing can be achieved by either defining smaller lots of trimmings (one combo bin instead of 5 combo bins) or selecting additional samples from 5 combo lots (e.g. N-75 instead of N-60) and should continue until the establishment has a high degree of confidence that the corrective actions are effective. Additionally, operator may increase monitoring and verification of both slaughter and dressing procedures, implement additional antimicrobial interventions and test additional products.

15. CFIA's actions during extended non-compliance

The CFIA may perform HACCP System and Design Tasks in cases of extended non compliance indicated by, but not limited to recall, port-of-entry violation, systemic HEP and open CARs related to sampling, prerequisite programs and CCPs associated with E. coli O157.

Appendix 1: Validation Procedure – Pathogen Reduction Step for Abattoirs

Sampling Protocol

The sampling should encompass a four (4) month period. It is recommended to sample during the seasons having the highest prevalence level.

Carcasses should be randomly selected over the validation period. Days, and hours within a day, should be determined in advance to create a sampling plan. At collection time, carcasses should be selected in a blind manner (e.g. 5th carcass following a carcass purposely selected). Side A (right or left) should be sampled for Enterobacteriaceae (EB) evaluation before the intervention of interest. Side B (left or right) of the same carcass should be sampled after the intervention. For the next carcass identified in the sampling plan, side B should be sampled first (before) and then side A (after). As a principle, the selection of side A or B should be alternated from one selected carcass in the sampling plan to the next one for the EB evaluation before the intervention of interest.

Statistical Evaluation

To assess the efficacy of the reduction intervention beyond chance, it is recommended to use a paired t-Test. This test is available in Excel (as an Add-Ins) under Tools / Data Analysis / t-Test: Paired Two Samples for Means. The Input Variable 1 Range should correspond to the values of EB counts before the intervention. The Input Variable 2 Range should correspond to after EB counts. The Hypothesized Mean Difference should be set to 0, and the Level of Signification (Alpha) to 0.05. Successful interventions should result in a P value inferior to 0.05.

Confidence interval +/- 0.5 log on the difference of EB log counts with 95% Confidence Level

n = Number of beef carcasses to be sampled over a 4-month period according to the 4-month production volume (N)

N = Production volume over 4 months

Group 1 = Standardized intervention i.e. fully automated commercial equipment with monitoring devices (e.g. steam pasteurizers)

Group 2 = Intervention that is not fully automated i.e. equipment involving some manual intervention or not having monitoring devices for all parameters (steam vacuum, organic acid sprays, etc.)

Group 3 = New interventions and all other interventions not in groups 1 or 2

Appendix 2: Testing Considerations for E. coli O157

(as pertains to FRGBP, PM and carcass surface meat used to produce the PM)

A. Sample pick-up

The collected sample must be representative of the lot being tested and must meet the screening methodology specifications.

B. Screening methodology (optional - the lab may proceed directly to confirmation)

Testing with both screening and confirmation methodologies must be performed in a lab accredited by the Standards Council of Canada (SCC), the Canadian Association for Laboratory Accreditation (CALA), or another accreditation body that is a signatory to the International Laboratory Accreditation Cooperation (ILAC) Mutual Recognition Agreement (MRA) as conforming to the requirements of ISO/IEC 17025:2005 for specific tests.

A presumptive positive result will be considered definitive, i.e. to correspond to a confirmed positive test result for E. coli O157 unless the sample proceeds to the confirmation method as described in section C (Confirmation methodology).

For ground beef/pattie, five (5) sample units of 65 g each must be tested (65 g taken from each of the 5 sample units submitted), representing a total of 325 gof meat product. This may be tested as a composite of 325 ± 32.5 g, unless otherwise specified in the test method used.

For beef trim or trim component samples consisting of 60 intact pieces of trim, an analytical unit must be taken from each of the 60 pieces of trim to make up a total of 325 g of meat product for analysis. This may be tested as a composite of 325 ± 32.5 g, unless otherwise specified by the test method used.

The following screening methods or equivalent¹ are acceptable:

• MFLP-16 - Detection of Escherichia coli O157:H7 in Foods - Assurance GDS^TM for E. coli O157:H7 Gene Detection System
• MFLP-30* - Detection of Escherichia coli O157:H7 in select foods using the BAX^® System E. coli O157:H7 MP

  *If using MFLP-30, the enrichment broth mTSB-n must not be used with the BAX^® System E. coli O157:H7 MP for the detection of E. coli O157:H7 in raw processed seasoned (with spice) meat

• MFLP-12 - Identification of Escherichia coli O157:H7 and Verotoxin-producing Escherichia coli O157/NM by the ADIAFOOD Rapid Pathogen Detection System, a Real-time PCR System
• MFLP-81 - Detection of Enterohemorrhagic E. coli (EHEC) in Food Products and Food Ingredients by the Assurance EHEC Enzyme Immunoassay (EIA) Method
• MFLP-82 - Detection of E. coli O157 by the Merck Singlepath^® E. coli O157 Kit
• MFLP-76* - The DuPont Qualicon BAX^® System real-time method for the detection of E. coli O157:H7 in raw beef trim and raw ground beef

  *If using MFLP-76 for the analysis of ground beef, each 65 g analytical unit of ground beef must be analysed individually.

• MFLP-98 - Detection of E. coli O157:H7 in food products by the VIDAS^® E. coli O157 (including H7) method

The above listed methods can be found in the Health Canada Compendium of Analytical Methods site. Note that the most recent published version of the method should be used.

Note:

Additional requirements may apply when testing product for export to other countries. Operators must follow MOP Chapter 11 Guidelines for testing of meat products for export to other countries, such as the United States of America.

Methods listed in chapter 11 may have additional methods not listed in this appendix (e.g. MLG FSIS methods) or the same methods with alternate dilutions and/or sample sizes identified. These will be considered acceptable only for products that are exclusively exported to the United States, or other countries (as applicable). Note that not all methods listed in this appendix are accepted for exports purposes.

C. Confirmation methodology requirements

The following method or equivalent² is acceptable:

• MFLP-80 - Isolation of E. coli O157:H7 or NM in Foods

The above method can be found in the Health Canada Compendium of Analytical Methods site. Note that most recent published version of the method should be used.

In addition to the confirmation requirements outlined in sections 7.6.1 to 7.6.4 of MFLP-80, the accredited laboratory must be able to perform either test A or B (verotoxin testing) as described in section 7.6.5 of MFLP-80.

The confirmation test must be done from the same broth that was tested presumptive positive by the screening test.

The confirmation procedure must commence within 24 hours of the initial positive screening result.

Appendix 3: Suggested High Event Period (HEP) Numerical Criteria

The following tables are provided to help establishments derive parameters for determining whether they have experienced a HEP. The tables provide specified numbers of positive results (given in the first column) occurring within a specified number of samples (entries within the remaining columns), that would indicate that the true percent positive of E. coli O157 findings would be greater than or equal to the specified percent positive given in the column headings, for the following percent confidence intervals:

• with 95% confidence (Table 1);
• close to 99% confidence (Table 2); and
• close to 99.95% (Table 3).

In the tables below, note that the test result from one composite sample of multiple slices (e.g., N60 sample) is considered one positive or negative result.

Table 1: True positive percent of E. coli O157 findings is greater than corresponding lower bound percentage in column with 95% confidence, given the number of positive results (rows) within corresponding number of samples (interior table entries)

Based on Table 1, if there were 4 or more positive results within 28 samples, then there would be 95% confidence that the process positive percent was not less than 5%.

Table 2: True positive percent of E. coli O157 findings is greater than corresponding lower bound percentage in column with about 98.85% confidence, given the number of positive results (rows) within corresponding number of samples (interior table entries)

Based on Table 2, if there were 4 or more positive results within 18 samples, then there would be 98.85% confidence that the process positive percent was not less than 5%.

Table 3: True positive percent of E. coli O157 findings is greater than corresponding lower bound percentage in column with about 99.95% confidence, given the number of positive results (rows) within corresponding number of samples (interior table entries).

Based on Table 3, if there were 4 or more positive results within 8 samples, then there is about 99.95% confidence that the positive percent was not less than 5%.

As mentioned in section 14, the CFIA requires establishments to take action if sampling of PM produces a positive rate statistically significantly greater than 5%. They may wish to choose more stringent criteria For example, an establishment may decide that if its percent positive exceeds 3.5%, the establishment has experienced a high event period.

If Tables 1-3 above do not meet an establishment's needs for determining HEP the establishment may develop alternate proposals. Such proposal must be presented to CFIA for approval.

Acknowledgement: Above tables have been excerpted from the United States Department of Agriculture - Food Safety Inspection Service (FSIS) Compliance Guideline for Establishments Sampling Beef Trimmings for Shiga Toxin-Producing Escherichia coli (STEC) Organisms or Virulence Markers - PDF (241 kb) available.

---

¹"or equivalent" - contact the CFIA, Food Safety Science Directorate, Executive Director, for the CFIA requirements to demonstrate equivalency of methodology not listed as acceptable in this document.

² "or equivalent" - contact the CFIA, Food Safety Science Directorate, Executive Director, for the CFIA requirements to demonstrate equivalency of methodology not listed as acceptable in this document.

Next page: Annex P - CFIA Protocol for Evaluating Bulk Container Freezing Processes for Meat Products | Previous page: Annex K - Tracking Sheet - Option Used for the Control of E. coli O157:H7 in Dry and Semi-dry Fermented Sausage