Protocol for piloting new technology and procedures for poultry slaughter

On this page

1. Introduction

The generic protocol is a tool to facilitate incorporation of licence holder's proposals on new meat hygiene technology or procedures. It specifies the basic process, tests and the resulting data which are required to allow an evaluation of an alternate procedure to achieve the similar or better outcome based on science. The generic protocol may need to be customized for specific situations and should meet the Safe Food for Canadians Regulations (SFCR) and its intent at all times.

2. Definitions

New technology
A new technology is defined as a novel technology never used by the licence holder (for example, material, substance or equipment etc.).
A protocol is defined as a detailed plan of the implementation of a totally new technology through an on-site trial in a federally registered establishment.
A procedure is a series of steps that licence holder will follow in order to achieve the objectives.

Some examples of procedures which have been previously examined by CFIA under this protocol are:

  • on-line or off-line reprocessing,
  • reuse of chiller overflow water,
  • less than daily sanitation of chillers,
  • alternate chilling of poultry carcasses, parts and giblets
  • alternate sampling procedures based upon ISO standards
  • reconfiguration of evisceration line(s) potentially affecting defect detection or carcass bacteria counts etc.

3. Prerequisites

The proposed test and control operation(s) must operate under a PCP (Preventive control Plan) under the SFCR and demonstrate continuous process control. The pilot study conducted must meet the SFCR requirements at all times.

4. Scope

  1. The study design and implementation should ensure that following are not compromised:
    • defect detection
    • defect removal
    • rejection process
    • pathogen reduction standards
    • animal welfare standards
    • food safety standards (for example, does not present a risk of contamination of food)
    • health and safety of the CFIA staff and licence holder personnel
  2. Additionally pilot study criteria:
    • it should not lead to hazards (biological, chemical and physical) that present a risk of contamination of a food
    • it should not lead to increased pathogen counts
    • it should not be against published criteria in other guidance documents (unless the criteria is part of the study)
    • it does not compromise CFIA ability to perform inspectional procedures
    • at all times the pilot should ensure there is zero tolerance for fecal contamination

Proposals may be submitted for any step in the slaughter process in an establishment for example, live hanging, scalding/defeathering, evisceration, chilling, cut-up/boning, packaging, etc.

If the licence holder is proposing changes to the existing regulatory requirements, an established legal regulatory amendment process must be used.

5. Appeal mechanism

Whenever there is doubt as to whether a proposed change requires evaluation under this protocol, license holder should inform local CFIA in advance as CFIA inspection staff may need to refer the proposal for further guidance and consultation with specialists.

6. Application procedure to CFIA

When CFIA services are requested, a submission containing the following items are to be provided to the CFIA veterinarian with supervisory authority:

  • a detailed protocol fully describing the proposed operation(s) including a complete experimental design for a pilot project including information referenced in the section titled Experimental Design
  • amended blueprints or process flow diagrams with facilities and equipment
  • description of how process control will be demonstrated based on visual observations and the charting of microbiological test results(refer to section titled Pass/Fail Criteria)
  • assurances that the licence holder understands it must perform or pay for all laboratory tests as well as the statistical analysis of the data
  • assurances that the licence holder will supply trained, competent personnel (refer to section titled Employee Competence)
  • specify the laboratory to be used and give assurances that the bacteriological test methods are listed as an Official Method of Analysis by Health Canada (HC) or by the Association of Official Analytical Chemists (AOAC), International. Indicate if the lab is accredited, and if not, include assurances that the company will provide full access to the CFIA to the lab to monitor applicable procedures and test results (refer to the section titled Laboratory Accreditation).


Control tests may commence before receipt of the letter of authorization provided there is agreement with local CFIA to monitor the procedures. However, additional tests or procedures may be required to complete the experimental protocol as amended by the licence holder.

7. New pilot protocol development

Generally, three (3) - five (5) replications may be adequate to provide sufficient information for related new technology and procedures including any required control procedures and associated standards or operational guidelines.

Published reports and information submitted by licence holder industry will be considered by CFIA when determining the number of pilot projects required prior to amending the regulatory policy.

A draft of the proposed policy may be circulated for comments to all stakeholders.

Confidential projects will be treated as information which is licence holder specific (that is, not to be published) until it becomes evident that several establishments have conducted similar pilot projects at which time regulatory policy formulation will be initiated.

8. Experimental design

The effect of the proposed change must be demonstrated by collecting control and treatment samples according to an experimental design and conducted as a pilot project under the CFIA oversight.

The following are suggested options:

  • complete all control sampling, and then collect all the treatment samples
  • split each test lot in half, collect the control samples, affect the change, and then collect the treatment samples (repeat for the required number of lots)
  • randomly split carcasses and/or parts and/or giblets from the same lot between two similar processing (for example, evisceration) lines (one line changed to function as treatment line) operating simultaneously in the same establishment

The proposed experimental design must include the following information:

  • current and proposed flow chart of affected, related activities
  • sampling locations
  • description of how control and treatment phases will be accomplished (see listed options above)

9. Employee competence

Applicable employees must be trained to facilitate the proposed change. Personnel must be accredited where required for specified functions for example, defect detection, defect removal, process control monitoring, rejection, reprocessing, quality control plans etc.

A written training program and employee training records must be on-site and readily accessible for CFIA examination.

10. Sampling location

The experimental protocol must define a precise location for collection of the sample(s) for each test.

The licence holder should determine sampling location in consultation with local CFIA for:

  • prevalence testing for carcass, viscera and cavity defects
  • prevalence testing for rejection process
  • prevalence testing for carcass defects
  • reconfiguration of the slaughter, evisceration and chilling lines

For pathological and processing defects, the carcasses will be selected:

  • downstream from team of establishment carcass/cavity/viscera detectors
  • before or after establishment helper/trimmer
  • before viscera is harvested (or discarded) or the carcass is trimmed (other than by the helper/trimmer)
  • before the carcass is vacuumed

For microbiology tests, carcasses will be collected as specified in the United States Department of Agriculture's (USDA) Pathogen Reduction/Hazard Analysis Critical Control Point (HACCP) regulations that is, minimum of 1/22,000 chickens, 1/3,000 turkeys, ducks, geese. guineas although the sampling location may be changed to suit the needs of the experiment.


  • Sampling may be conducted such that the test results can also be credited towards fulfilling export requirements.
  • Refer to Chapter 11, export requirements for the United States, Annex T for full details on sample selection and processing for bacteriology (Generic coli) testing.

11. Microbiology tests

Samples must be randomly selected and handled using sterile technique.

Bacteria counts may be determined using the carcass rinse technique (Butterfield's phosphate diluent (BPD), 400 millilitres (ml) for chickens, 600 ml for turkeys or by other procedures mandated by the CFIA (for example, swabbing for turkey carcasses). Rinsing with the diluent may occur in a compatible area of the establishment or alternatively, the carcass may be transported to the lab for the rinsing. Sterile sampling solutions, Butterfield's phosphate diluent (BPD), can be stored at room temperature. However, at least on the day prior to sample collection, check solutions for cloudiness. Do not use solutions that are cloudy, turbid or contain particulate matter. Place the number of containers of sampling solution (BPD) that will be needed for the next day's sampling in the refrigerator. Ensure that shipping containers, coolant packs and shipping documents are prepared as required. Place the number of containers of sampling solution (BPD) that will be needed for the next day's sampling in the refrigerator. Sterile Buffered Peptone Water (BPW) can be substituted for BPD. It must be noted however that there may be a slight rise in bacteria counts following this change (¼ log increase). The sampling material should not have antimicrobial properties which might reduce bacterial counts.

Microbiology tests must commence within 24 hours of sample collection and with approximately the same time interval between collection and laboratory processing for all samples.

Samples (carcasses or rinse fluid) must be refrigerated to 4°Celsius (C) or lower (but not frozen) until analyzed (on-site) or packaged for shipping. Shipped samples must be packed in insulated containers containing ice packs so as to maintain a carcass surface temperature of between 0 and 7°C during (overnight) transport to the lab.

Total E. coli count per ml or cm2 must be determined to serve as an indicator of faecal contamination.

Demonstrate pathogen reduction controls based on Salmonella spp. and Campylobacter spp.

Total Plate Count (TPC) must also be determined for each carcass to serve as a confirmatory test for the effect of the proposed change on process hygiene and to provide an indicator of shelf life.


Presence or absence for specified pathogens must be performed whenever it is determined by CFIA, in consultation with technical experts (including HC), that the proposed procedure may favour the growth of, or selective survival of, particular pathogen(s), or when deemed necessary to facilitate international acceptance of new or novel inspection methods, processes, or technology.

12. Laboratory accreditation

Pilot projects for processes/procedures which have been published in peer reviewed journals will usually not require an accredited laboratory as determined by CFIA. However, new, unpublished procedures may require the use of an accredited lab if required to facilitate international acceptance (and favourable export markets) as determined by CFIA or if requested by HC or the CFIA to resolve food safety concerns, and any required microbiological tests for foodborne pathogens.

Laboratories accredited by a federal, provincial or US government agency for the specific bacteriology test(s) or by an internationally recognized registrar for example, Standards Council of Canada (SCC), will be considered as accredited for this protocol. Laboratories of the federal or provincial governments and Universities will be recognized as having equivalent to accredited status for this protocol. Establishments wishing recognition of their in-establishment laboratory require a Quality Management System (QMS) for the lab, equivalent to that required for government (HC/CFIA) accreditation.

Non-accredited laboratories must be included within the establishments' prerequisite programs, as part of their PCP system, and be accessible to CFIA staff (for auditing the applicable test procedures, records and equipment) to qualify for use under this protocol. If remotely located from the establishment, the company must provide assurances of unrestricted access to CFIA staff and to pay for CFIA audit expenses on a fully cost recovered basis.


Upon completion of the pilot project under this protocol, an accredited lab is no longer required. Ongoing microbiological testing may be performed in the laboratory specified by the license holder's PCP system.

13. Sample size


This section describes test requirements for pilot projects which consistently remain in compliance with all regulatory requirements, standards and process controls (evisceration standards (ES), presentation standards (PS), defect detection standards (DDS), carcass dressing standards (CDS), and poultry rejection process (PRP)).

Whenever a Canadian national standard does not exist, then an establishment specific standard must first be established by collecting samples during a control period conducted over the same period of time as planned for the treatment phase of the pilot project.

Various studies have found that the greatest source of experimental variation is commonly between carcasses from the same lot processed under what appears to be similar conditions. However, a significant source of variation may be between the producers or lots. Therefore, sampling for tests should be spread over at least 30 lots and 20 producers for control samples and again for treatment samples or over a minimum of 20 lots or producers or shifts if collected as paired samples. Testing for defects must continue throughout the control and treatment phases of the pilot project.

a. Microbiology tests; (Total E. coli counts)

  • Option 1: One hundred control and one hundred treatment carcasses if collected as before and after samples
  • Option 2: Fifty paired samples if control and fifty treatment samples collected from same lot in same time period
  • Option 3: As required by CFIA when experimental design and process controls is considered valid by consulted technical experts

Using a 95% confidence limit and 80% power, and assuming a pooled variance of 0.58, a sample size of 100 carcasses was calculated as sufficient to detect a 0.3 log10 mean E.coli count difference between the control and treatment groups.

b. Defect detection and removal

Pre-chill testing using Defect Detection Standards and Carcass Dressing Standards

c. Defect non-detection rate

Monitoring tests for internal cavity defects are described Post-mortem Examination Program.

If a process is not covered by this protocol or published guidance, sample selection and frequency may be specified by CFIA.

d. Prevalence of defective carcasses requiring veterinary judgment

Prevalence of defective carcasses requiring veterinary judgment is required when proposed changes which have potential to affect evidence of generalized disease of public health significance in carcasses prior to post-mortem detection/inspection.

Licence holder will assist CFIA inspection staff when required tests are to be performed by the CFIA. The Veterinarian with Supervisory Authority will ensure that post-mortem detection/inspection is not compromised. Unusual situations which might affect veterinary disposition are to be referred to the veterinarian with supervisory authority throughout the prevalence testing.

In general, for an establishment with two similar evisceration lines and permitting paired sampling (for example, licensed holder's employee rotates to the other line every "X" min. and spends approximately the same amount of time on each line for each lot), a total of 14,000 carcasses should be evaluated that is, 700 carcasses/day × 20 days and 7,000  carcasses/line. However, if the testing can be replicated in a second establishment, (also with two lines) to facilitate statistical analysis, then 7,000 carcasses should be collected from each establishment over 10 working days that is, 700 carcasses/day × 10 days and 3,500 carcasses/line × 4 lines, again for a total of 14,000 carcasses.

Using a 95% confidence level, an 80% power, a one (sided) tailed test, and an estimated national prevalence rate of 1.2%, a sample size of 14,224 has been previously calculated as sufficient to detect a 0.3% difference (drop) in the number of carcasses with visible evidence of disease or conditions (at the post-mortem detection/inspection stations) and which require removal from the line for veterinary disposition.

14. Pass/Fail criteria:

a. Testing while in compliance:

i. Microbiology tests

If an establishment specific standard is required, the results from control samples must be analyzed to calculate required values for the control chart as selected for use during the treatment phase. Bacteriology counts for treatment carcasses must be plotted on a control chart for example, Shewart, Cusum or as illustrated in the USDA's Pathogen Reduction/HACCP regulations as outlined in Chapter 11, Exports, and United States of America.

Treated carcasses may be sold as "edible" if the establishment provides evidence satisfactory to the CFIA that product complies with all regulatory requirements. Treatment carcasses will be deemed to be in compliance whenever test results comply with the predetermined acceptance/rejection limits as specified in the application for the pilot project, for example:

  • if using the Shewart Control Chart, then no bacteria count from a treatment carcass may exceed the Upper Control Limit (UCL)
  • if using the chart from the USDA's Pathogen Reduction/HACCP regulations (1996), a moving window of the last 13 tested carcasses must indicate that no more than 3 exceeded the marginal value (little 'm') and none exceeded fail value (large 'M')

ii. Defect detection and removal

Defect detection and removal will be deemed to be in compliance whenever:

  • product rework is not required for example, under Carcass Dressing Standards or equivalent pre-chill tests
  • line speed reductions are not required as a result of the tests performed to monitor the detection and handling of defective carcasses by licence holder's employees

b. Corrective action

Whenever a bacteria count exceeds the fail/rejection limit and/or treated carcasses must be reworked under the CDS program and/or the line speed must be reduced due to ineffective carcass detection by establishment employees, then the cause must be determined, corrective action taken (including possibly amending the treatment protocol or implementing a critical control point (CCP) to improve process control). The treatment period counter must then be reset to zero i.e. start all over again. The CFIA must be informed of all corrective actions before implementing them.

The licence holder must terminate a pilot project for repeated failures that is, 2 or more resets to zero.

c. Data acceptance

Licence holder's statistical analysis of the data must indicate that the proposed change (treatment) either exceeds or at least maintains the then current CFIA standard(s) for the following items:

  • defect detection related to post-mortem judgment
  • defect removal
  • rejection process
  • process controls
  • compliance with "zero tolerance" for visible evidence of faecal contamination
  • pathogen counts and pathogen reduction standards

15. Records, ATIP and auditing

License holder's data and associated statistics analysis will be considered as confidential and will be released under ATIP (Access to Information Procedures) in a summarized generic format (for example, plant A, plant B, etc.) such that test results cannot be identified to a specific originating establishment.

Upon completion of analysis and review by CFIA and technical specialists, if requested all copies of raw data will be returned to licence holder through local CFIA office.

The following records must be stored on-site by establishment management and be readily accessible for review by the CFIA and government auditors:

  • completed submission and corresponding letter of authorization from CFIA
  • all pilot study data (for example, completed forms, test results)
  • records of ongoing monitoring and verification tests and other procedures for the retention period specified in the SFCR

Additional Guidance:

  1. The Food Safety Enhancement Program approach to a preventive control plan
  2. Evidence showing a control measure is effective