Animals > Animal Diseases > Bovine Spongiform Encephalopathy > Safeguards  

Canadian Food Inspection Agency
Feed Ban Review

March 2, 2005

Outline

Definitions

Flushing:  To clean mixers and other equipment by moving a non-prohibited ingredient, usually bulk grain or oilseed products through the system to purge the equipment of any residues of prohibited material that might be present.

HACCP:  The Hazard Analysis and Critical Control Point (HACCP) system is a system for controlling the safety and quality of food at the plant level.  It encompasses seven HACCP principles relating to, ultimately, preventing hazards from contaminating food.  Under HACCP, plants must identify all critical points in the manufacturing process that could effect the safety and quality of food products.  Then the manufacturer must establish procedures (e.g., cooking times and temperature for each critical control point.  HACCP places responsibility for ensuring food safety on the food manufacturer or distributor.  

Major non-compliance: Includes situations that could expose ruminants to prohibited material.  These situations include a failure to properly clean out equipment, and a failure to include the required statement "Do not feed to cattle, sheep, deer or other ruminants" on labels for feeds containing prohibited materials.

Minor non-compliance: includes situations where the procedures, records or labels may not meet one or more of the required, less critical elements (referred to as standards in this document) specified in the regulation.  Examples include a failure to indicate either the manufacturing date on a production log, or a customer's name and address on an invoice for a "cash sale".

Prohibited Material (also referred to as Prohibited Meat and Bone Meal - MBM): Protein, or any material that contains proteins, that originated from a mammal, other than pure porcine (pig) or equine (horse).  Prohibited material does not include milk, blood, gelatine, rendered animal fat or their products.

Sequencing: To predetermine the order of manufacturing different feed products so that any residual prohibited material is flushed into a feed product intended for non-ruminant species.

Specified Risk Material (SRM):  The skull, brain, trigeminal ganglia, eyes, tonsils, spinal cord and dorsal root ganglia (DRG) of cattle aged 30 months or older, and the distal ileum of cattle of all ages.

Ruminant: refers to any animals that chew a cud such as cows, deer, sheep, goats, etc. They are also referred to as "ruminant mammals".

Mammalian-to-ruminant feed ban: Refers to the 1997 Health of Animals Regulations that prohibits feeding protein derived from mammals to ruminant animals such as cows, sheep, deer, etc., with the exception of porcine or equine derived protein.

EXECUTIVE SUMMARY

Feed Ban Review

Bovine spongiform encephalopathy (BSE) is spread when cattle consume feed contaminated with the BSE agent. To limit this potential occurrence, Canada implemented a feed ban in 1997. The ban was one of the measures in place to mitigate BSE risks.  It was introduced proactively, years before BSE was detected in Canada's domestic cattle herd, and met the recommendations of the World Health Organization (WHO).

Canada's feed ban prohibits feeding most mammalian proteins to ruminant animals, such as cattle, sheep and goats. The ban requires rendering facilities, feed manufacturers, feed retailers and livestock producers to follow and document production and feeding procedures to prevent the inclusion of prohibited materials in feed and feed ingredients intended for ruminant animals, such as cattle, sheep and goats.

After the two latest cases of BSE had been confirmed in January 2005, the Canadian Food Inspection Agency (CFIA) undertook a review of the feed ban to determine whether this important safeguard has been providing an appropriate level of animal health protection since its inception.

The month-long review that concluded on February 18, 2005, examined the design of the feed ban, the CFIA's approach to implementing and enforcing the ban's requirements, and compliance levels at rendering facilities and feed mills. The review included examining inspection reports, regulatory documents and risk assessments; and conducting interviews and on-site inspections of feed mills and rendering facilities.

The ban was designed in accordance with international guidelines and drew from science and the most current understanding of BSE available at the time. Canadian officials consulted domestic and international stakeholders to develop regulations that could be implemented, and would be enforceable and effective. The United States' feed measures, also introduced in 1997, are almost identical to Canada's ban.

In preparing to introduce the feed ban, government officials met extensively with the feed mills, and rendering and livestock industries as the ban was taking shape in 1996. Once the new feed regulations had been introduced, the CFIA immediately developed an inspection program.  Communication with industry broadened to educate and inform the feed industry and farms about the regulatory requirements. The inspection program was introduced in phases.  From 1997-2000, inspection activities focussed on integrating the feed ban's requirements into standard industry practices. From 2000 onward, inspections increased in frequency.

Adherence to the feed ban requirements at feed mills and rendering facilities, particularly in recent years, has been high.  Instances of non-compliance have been largely administrative in nature, such as minor documentation and recording-keeping issues.   These do not pose BSE-related risks to animal health. On average, 95% of feed mills and 93% of renderers inspected over the past three years were either fully compliant or reported only minor compliance issues. During the on-site inspections conducted as part of this review, one significant case of non-compliance was observed. Corrective action was immediately taken. The CFIA is considering additional enforcement action.

The review team also note that since 2002, the time that feed mills and rendering facilities had taken to address issues of non-compliance has decreased. 

The feed and rendering industries are moving increasingly toward using dedicated production lines and facilities to handle either prohibited or non-prohibited materials. Of Canada's 29 rendering facilities, only six handle both types of material. Of this smaller group, all but two use dedicated production lines. Similarly, of Canada's 550 feed mills, only 94 both handle prohibited material, and manufacture ruminant feed. This shift toward using dedicated production lines is reducing potential opportunities for the cross-contamination of ruminant feeds with prohibited materials. 

In December 2004, the CFIA proposed new regulations requiring the removal of specified risk material (SRM) from all animal feeds. SRM are tissues that, in infected animals, contain the BSE agent. Removing this material from the feed-manufacturing process would both eliminate the risk of cross-contamination, and eradicate BSE from the national cattle herd more quickly.

Canada's feed ban has been developed according to sound science-based principles. The Government of Canada has taken significant steps to increase industry's awareness and understanding of the ban. Furthermore, through regular inspections, the government has been continuously verifying that the industry is appropriately implementing the feed ban requirements. Notwithstanding the need for continuous improvement and evolution, Canada's feed ban-as designed, implemented and currently applied-is providing an effective barrier that is contributing to reducing the risk of BSE, and which will lead to the eventual eradication of BSE from the national cattle herd.

INTRODUCTION

The Canadian government introduced an amendment to the Health of Animals Regulations in 1997 as part of a series of preventive measures to mitigate the spread of BSE in the Canadian herd.  This amendment, commonly referred to as the "feed ban" or "mammalian-to-ruminant feed ban," was aimed at preventing and controlling the spread of BSE.  It was modeled on measures adopted by the United Kingdom (UK) and other European countries that were experiencing high levels of BSE infectivity within their cattle populations. Up to that point, the only diagnosed case of BSE in North America had been discovered, in 1993, in a cow imported to Western Canada from the UK.

BSE likely entered North America during the 1980s when Canada and the United States imported a limited number of cattle from the United Kingdom. Given the long incubation period of BSE, some of these cattle could have been infected with BSE despite appearing healthy when entering either country. Canada imported 191 such animals during this period. The UK import, in 1993, that tested positive for BSE was diverted from the food and feed systems.

With the exception of 68 animals that had already died or been slaughtered, all cattle imported during the 1980s and remaining in Canada were identified, euthanized and tested negative for BSE. Of the 68 animals, 10 are considered to have posed the greatest risk because they originated from farms in the United Kingdom that later reported cases of BSE.

As noted above, the 1993 case was diverted from the human food and animal feed systems and, as a result, the risk of BSE infectivity had been assessed as extremely low. However, world-wide concern about this disease was increasing.  This concern prompted the Canadian and United States governments to take the proactive step of adopting a feed ban before there was any evidence that the disease was present in the North American herd. In 1996, the World Health Organization (WHO) recommended that all countries should adopt a ruminant-to-ruminant feed ban. Canada and the U.S. went somewhat further and adopted a partial mammalian-to-ruminant feed ban, which extends the ban beyond ruminants to include other species.

On January 11, 2005, after Canada had confirmed its third case of BSE in a domestic cow,  the Minister of Agriculture  and Agri-Food, the Honourable Andy Mitchell (who is also responsible for the Canadian Food Inspection Agency), tasked the CFIA with carrying out an overall assessment of the effectiveness of Canada's feed ban. He also invited other countries to participate in this review. The United States government sent a team of officials from the United States Department of Agriculture and the Food and Drug Administration to Canada to carry out its own independent review. 

In carrying out the review, the Canadian and American teams collaborated in terms of exchanging information and data on the Canadian feed ban.  However, each team conducted its own analysis and prepared a separate report. The Canadian team focussed on the regulatory process and implementation of the regulations, including how Canada's feed industry had adopted the requirements of the regulations, while the US team focussed on a comparison of the US and Canadian feed bans and how the Canadian ban would affect risks to the US cattle industry.    

Objectives of CFIA's Review

The review of the effectiveness of Canada's feed ban had three objectives:

• to determine the extent to which the design of the 1997 feed ban regulations was appropriate;

• to assess CFIA's approach to implementing and enforcing the feed ban; and

• to assess the extent to which the rendering and feed mill industries have complied with the 1997 feed ban regulations.

Scope and Approach of the review

In order to address the objective of the review, the review team examined:

• the process that CFIA used to design the 1997 feed ban regulations;

• the approach that CFIA used to support the implementation of the feed ban across the feed mill and rendering industries in Canada;

• the inspection strategy which CFIA uses to ensure that the feed mill and rendering industries comply with the feed ban regulations; and

• the implementation of the feed ban requirements on the part of industries, including the level of their compliance with these requirements.

The review was carried out from January 14, 2005 until February 18, 2005 and included a review of pertinent documents (including -- but not limited to --  regulatory impact analysis statements, historical inspection records, risk assessments and panel reports and training documents); interviews with key CFIA staff members at the program and operational level; and on-site inspections of a random sample of 31 commercial feed mills and seven rendering facilities to confirm the levels of compliance.  The feed ban regulations apply to feed retailers, livestock producers and on-farm feed mills as well. However the review focussed on commercial feed mills and rendering plants because they represent a higher risk in terms of potential contamination of non-prohibited material or  ruminant feed with prohibited material.

Inspection records were reviewed for the fiscal years, 2002-03, 2003-04 and 2004-05 up to January 14, 2005.  Reliable compliance data have been available only for the past three years from CFIA's electronic database (Multi-Commodity Activities Program - MCAP), which was launched in 2002.

For the on-site inspections, the CFIA ranked the 550 feed mills and 29 rendering facilities into high-risk or low-risk categories.  94 feed mills and six rendering plants fall into the high-risk category.  From the 550 feed mills, 27 were randomly selected from the high-risk category (those that handle prohibited material and manufacture ruminant feed) and 4 from the low-risk category (those that either do not handle prohibited material, or do not manufacture ruminant feeds).  From the 29 rendering plants, six are from the high-risk category (plants which handles both prohibited and non-prohibited material) and one plant that had recently converted to the low risk category were inspected as part of this review. 

Appendix 1 provides further details on the scope and approach of the review.

Using this report

This report is divided into three parts.  Part One presents background on BSE, the level of risk for BSE in Canada, the 1997 feed ban, and Canada's Feed Industry.  Part Two presents the results of the review of the 1997 feed ban.  More specifically, Part Two looks at the feed ban in terms of the three lines of enquiry:  the ban's design, its implementation and its effectiveness.  Part Three contains appendices to the various sections of the report.

Part 1: BACKGROUND

1.1  What is BSE?

BSE is a progressive, fatal infection of the central nervous system of cattle. It is one disease within a group of diseases collectively known as transmissible spongiform encephalopathies (TSEs). Although the exact cause of BSE is unconfirmed, it is associated with the presence of an abnormal protein called a prion. No treatment or vaccine is currently available for the disease.  It is believed to be spread amongst cattle by consuming-prion infected meat and bone meal.

1.2 Level of risk for BSE in Canada

It is probable that one or a few potentially infected cattle imported from the UK into Canada before 1990 would have entered the animal feed system. Rendered meat and bone meal from such animals would have been included in cattle feed, which was permitted at the time, and led to the development of additional cases of BSE. The same scenario would have been equally possible in the United States, where similar feed practices were followed. In 1997, acting on the recommendations of the World Health Organization (WHO), Canada and the United States introduced pre-emptive feed bans. At that time, this recycling of infectivity would have been dramatically reduced, and the level of BSE present in Canada would have crested and, through normal attrition due to slaughter and death, begun to decline.

1.3 Canada's feed industry

The feed ban regulations apply to Canada's feed and livestock industry, which consists of rendering plants, commercial feed mills (that manufacture half of Canada's production), and feed retailers.  The regulations also apply to on-farm feed mills and farms that feed cattle and other ruminants.

The feed sector is Canada's 16th largest manufacturing industry, with annual production of 15 million metric tonnes.  It employs about 9,000 people.  The value of production from this sector is $3.5 billion annually.  In addition to the feed sold in Canada, feed is exported (primarily to the US), and feed from the US is also distributed and sold in Canada.  Swine, dairy and poultry feed accounts for 85% of all feed produced and sold by Canadian feed manufacturers. 

The chart below depicts a typical feed cycle showing the feed inputs from farms, abattoirs and dead stock that enter the feed manufacturing cycle.  The product enters through the rendering process and is passed through to the feed mills and distributors.

[D]

Canada's rendering industry

It is important to note that only six of the 29 rendering facilities have a permit from CFIA that authorizes them to handle both prohibited and non-prohibited material. 

Two thirds of the 29 rendering facilities in Canada are owned and operated by large corporations.  These corporations include international vertically integrated food companies.  Seven rendering facilities are attached to federally registered slaughter plants, which helps to ensure that these companies have very tight control over the rendering and disposition of raw material from their operations.

In 2003, Canadian renderers processed approximately 2.2 million tonnes of inedible animal by-products.  More than 50% (556,000 tonnes) of the products manufactured by rendering were protein meals (including blood meal).  The remaining products were animal fats and fatty acids.  In 2003, Canada produced approximately 478,000 tonnes of MBM (including cattle, pork, poultry and fish protein meals). In 2003, approximately 535,000 tonnes of animal fats, fatty acids and oils, and 77,600 tonnes of blood and feather protein meals were also produced.

All of Canada's rendering plants are members of the Animal Protein Producers Industry Association (APPI), and this association has sponsored the adoption of HACCP-based Quality Assurance Programs.  A number of large rendering facilities representing 74 % of the production have implemented these process controls which include an audit by a third party to ensure compliance.

Commercial feed mills

Some 550 commercial feed manufacturers operate in Canada.  The geographic distribution of those feed mills is shown below.  Feed mills can receive feed ingredients from animal by-products or vegetable by-products.  Like the rendering industry, the Canadian feed industry is highly segregated; only about 94 of the 550 mills handle products that could result in cross-contamination.  The number of mills that handle both prohibited and non-prohibited materials has been declining over the past few years.  Many of Canada's feed mills are members of the Animal Nutrition Association of Canada (ANAC). This association, as in the case of the rendering association, has sponsored HACCP type process control programming.  The implications of this will be discussed later in the report.  Some 176 feed mills are HACCP certified.

	[D]
click on image for larger view

1.4 Canada's 1997 feed ban

As a precautionary measure, Canada imposed a partial mammalian-to-ruminant feed ban on August 4, 1997.  The ban was imposed in light of the outbreak of BSE in the UK and in Europe, and further to the World Health Organization's recommendations of 1996. The experience in the UK shows that a similar type of feed ban had a significant effect on reducing the incidence of BSE.  The objective of the ban was to minimize any potential risk of introducing or spreading BSE into Canadian herds. This ban, along with the other precautions, i.e. border controls and surveillance, would also provide an additional barrier to the entry and spread of BSE.

The regulations to ban prohibited material in feed for ruminants were published in Canada Gazette Part II on August 4, 1997 and came into force on the same day. A full description of the regulations can be found at http://laws.justice.gc.ca

The United States introduced similar regulations at the same time in an effort to institute a North American strategy to prevent BSE from gaining a foothold in the US and Canada. The Canadian and US regulations were harmonized. The provisions were similar in content and the timing of their application in order to maximize the integration of the North American feed system. A notable difference in Canadian regulation was the exclusion of poultry litter and plate waste from feeds.

When the ban was introduced, a decision was reached not to recall the feed that was currently in the system, given the perceived low risk. All retailers were given a grace period until September 3, 1997 to use or distribute feed already produced.  Feed manufacturers received a grace period until October 3, 1997 to comply with labelling requirements.  Livestock producers were given a grace period until October 3, 1997 to use the feed manufactured prior to the feed ban. The US also adopted these measures, given the similar risk profiles of the two countries. 

Part 2: REVIEW OF CANADA'S FEED BAN

This part of the report focuses on three lines of enquiry:  the design of the feed ban, the implementation of the feed ban inspection program and the industry's compliance with the requirements of the feed ban. 

2.1 The design of the 1997 feed ban regulations

This first line of enquiry assessed whether the processes and approach that the CFIA followed in developing the amendments to the Health of Animals Regulations relating to the feed ban met reasonable standards.  The review looked at the Agency's process for developing these amendments in terms of generally accepted criteria set out by the CFIA and the Central Agencies (Privy Council Office and the Treasury Board Secretariat) for assessing this aspect of the regulatory process.

For the design of the 1997 feed ban regulations, the review criteria represent a set of expectations that federal organizations should meet when developing regulations. Accordingly, the review team expected that in developing the feed ban regulations the CFIA would have:

• followed a proactive, anticipatory approach in developing the feed ban regulations;

• carried out appropriate analyses and gathered comprehensive, relevant information;

• communicated and consulted effectively with stakeholders.

For this part of the review, the primary sources of information were documentation and interviews with key participants in the feed program.

Canada followed a responsive and precautionary approach with respect to developing the 1997 feed ban regulations

Canada, along with international partners, had devoted considerable effort to controlling BSE over the decade leading up to the 1997 feed ban. The following is a chronology of events relating to BSE and the control measures Canada put in place to mitigate the spread of the disease:

• 1978 - Canada banned meat and bone meal (MBM) for livestock feed imported from the United Kingdom (as well as other countries) due to Foot and Mouth Disease.
• 1982 (to 1990) - Canada imported cattle from the UK.
• 1986 - BSE was first documented in the UK.
• 1988 - Importation of meat meal, bone meal, and blood meal were officially banned from all countries except the US
• 1990 - BSE was designated a reportable disease under the Health of Animals Act.
• 1990 - Canada prohibited further importation of cattle from the UK (a total of 191 animals had been imported from the U.K. during the period 1982 - 1990).
• 1990 - Canada places UK origin cattle under an animal health monitoring program.
• 1992 - The National BSE surveillance program was implemented.
• 1993 - A cow imported from the U.K. that was in the monitoring program initiated in 1990 was confirmed infected with BSE.
• 1994 - All remaining UK cattle imports were either returned to the UK, or were euthanized.  All tested negative for BSE.
• 1996 - World Health Organization recommended that all countries implement feed bans.
• 1997 - Canada and the United States implemented their respective mammalian protein feed bans (with some exceptions) as a precautionary measure in response to WHO recommendations.  CFIA amended its  National Feed Inspection Program to include the mammalian-to-ruminant feed ban regulations, which targeted renderers, feed manufactures, feed retailers and farms in order to verify compliance with the regulations.
• 2000 - CFIA suspended the importation of all rendered animal protein products, of any species, from any country that Canada did not recognize as free of BSE.
• 2001 - The creation of a Canadian Cattle Identification Program was instituted for cattle and bison, enhancing our ability to trace individual animals from the herd of origin to slaughter.
• 2003 - In May, Canada detected the first case of BSE in a cow born and reared in this country.
• 2003 - In July, Canada amended the Health of Animals Regulations and the Food and Drug Regulations to remove Specified Risk Material (SRM) from the human food supply.
• 2003 - In December, BSE was confirmed in a cow in Washington State that was imported from Canada.
• 2004 - In December, Canada proposed strengthening BSE precautionary measures by requiring all Specified Risk Material (SRM) to be removed from all animal feed.
• 2005 - On January 2 and 11, respectively, two additional cases of BSE were confirmed in indigenous Canadian cattle.

During the 1990s, Canada carried out a number of BSE-related studies, and Canadian officials were active in scientific and regulatory bodies here and internationally. Much debate took place on the implications of BSE and the options for preventing and controlling the disease.

In 1996, the World Health Organization issued a report on BSE that confirmed the link between the disease in animals and variant Creutzfeldt-Jakob Disease in humans. At that point, the CFIA moved quickly to act on the report's findings and to advise stakeholders of the amendments that it planned to make to the existing regulations. It then began to work with the Central Agencies (Privy Council Office and the Treasury Board Secretariat) and other federal departments to draft the feed ban regulations which were promulgated without delay. The CFIA continued to communicate frequently and openly with industry, producers and others about its intentions to amend the regulations and impose a ban on feeding prohibited material to cattle and other ruminants. 

It was evident from discussions with the CFIA officials and from the material reviewed that CFIA had considered BSE to be of critical importance. Equally clear was that CFIA officials had been actively considering the best course of action, taking into account their rapidly changing understanding of risks, and the costs and benefits of optional regulatory approaches.

The feed ban regulations were based on sound analyses

The CFIA developed the amendments according to analyses of risk and in response to a demonstrated problem.  The 1996 WHO recommendations provided a solid justification for the feed ban amendments to the regulations. In addition, the CFIA analyzed its regulatory options, and their associated benefits, costs and regulatory burden. This analytical work is documented in the Regulatory Impact Analysis Statement included in the March 29, 1997 Canada Gazette, Part 1.  Options that were assessed included taking no regulatory action, taking a path that differed from the one to be taken by the United States, and taking action that would match that of the United States. The third option, which also met the intent of the WHO recommendation, was chosen.

CFIA communicated and consulted with stakeholders extensively and continuously

The federal government is responsible for regulating the feed industry. However, the regulations relating to the feed ban considered the roles of the parties involved and exploited opportunities to promote co-ordination among governments. Consultations were carried out with provincial counterparts and with the CFIA's counterparts at USFDA .

As noted above, industry and other stakeholders were advised that the government intended to amend the Health of Animal Regulations, and dialogue continued throughout the development process.  A discussion paper was prepared and presented to industry associations, as was a draft of the Regulatory Impact Analysis Statement.  Feedback, in some cases disagreeing with the need for the feed ban, was received.  International counterparts were consulted throughout this period.  The documentation of the regulations showed that the CFIA had thoroughly analyzed the options and their implications, along with the proposed amendments.

The review team has concluded that the procedures which CFIA followed in developing the 1997 amendments to the Health of Animals Regulations met the review criteria.

2.2 CFIA's approach to implementing the feed ban regulations

This second line of enquiry assessed whether the processes and approach that CFIA had followed in implementing the amendments to the Health of Animals Regulations relating to the feed ban met reasonable standards. 

For this line of enquiry, the review team expected to find that in implementing the feed ban regulations that the Agency had:

• developed both a sound program framework and standards for assessing compliance, and that it had properly documented and communicated them to staff and industry; and
• developed appropriate inspection tools to assess compliance, which were  available for use by trained and professional inspection staff.

For this part of the review, the primary sources of information were documentation, and interviews with key participants in the feed program.

CFIA developed and implemented standards and a program framework without delay

The 1997 amendments to the Health of Animals Regulations imposed new requirements on commercial and on-farm feed mills, rendering facilities, and farms and feed retailers that process or handle both prohibited and non-prohibited material.  After the promulgation of the regulations, CFIA immediately developed and implemented an inspection program and inspection guides to assess industry's compliance with the regulations. This program was added to the existing feed program inspection activities.  At the same time, CFIA provided educational material to the feed sector about the regulatory requirements and their responsibilities in meeting them. 

The Agency targeted the top end of the feed production chain as a strategy for managing the risks associated with renderers and feed manufacturers supplying to thousands of farms.  Renderers were subject to an annual inspection, and  they received an annual permit to operate if they were found to be compliant.

The new feed ban regulations focussed largely on the processes for manufacturing feeds, thereby preventing ruminants from consuming feed containing prohibited animal proteins.  These regulations added a new dimension to the feed program, which had historically focussed on sampling products to verify safety and efficacy.

This change in focus had implications for both the regulated parties and CFIA inspectors. The feed ban regulations resulted in a change in inspection forms, and standards for compliance with the new regulatory requirements were developed and communicated to the regulated industries.  Initially the regulations were translated into standards designed to verify that prohibited material had been kept entirely separate from ruminant feed throughout the rendering, feed preparation and distribution process.  The inspection guidelines included standards for feed mills and for rendering facilities.  Additional standards were developed for other segments of the feed industry such as retailers and farms that feed ruminant animals. The design of the inspection program is based on the International Standards Organization's audit principles.

In 2000, the inspection activities were clarified to provide more guidance and remove subjectivity. The standards covered process controls and operating procedures relating to the separation of material and equipment clean-out (e.g. sequencing and flushing).  They also covered record-keeping requirements and packaging and labelling requirements to reduce the risk of contamination or misuse through the feed distribution chain. 

Within the first six months, nearly all renderers and feed mills had been inspected using the new inspection guidelines.  Thereafter, the feed mill inspection frequency returned to its three-year inspection cycle.  Rendering plants continued to be inspected and be granted operating permits.

Frequency of inspections increased when the new requirements were introduced 

The year 2001 saw new scientific developments and the continued spread of BSE in a number of countries.  Given the critical nature of the feed ban in controlling the spread of BSE, it was decided to increase compliance inspections at commercial feed mills to an annual cycle.

By 2002, CFIA had introduced a comprehensive "systems approach" to inspection including new inspection forms, which included a computerized inspection database to capture inspection results.  By this point in the development of the inspection protocol, a total of 13 feed ban-related inspection tasks with 86 standards for the commercial feed mills and a total of 18 feed ban-related inspection tasks with 62 standards for the rendering industry had been developed.  This approach to implementation has resulted in a solid program, with properly documented standards based on process-control principles. 

Inspection staff had appropriate inspection tools and training

Prior to the 1997 feed ban, the Feed Program Inspection Manual was the main instructional tool for training CFIA inspectors in inspecting feed mills and rendering facilities. Immediately after the 1997 feed ban had been instituted, the Manual was supplemented by newly developed guidelines, standards and procedures to assist inspectors in carrying out the new feed ban inspection tasks.  Workshops were held in all regions to get inspectors "up to speed" quickly with respect to the regulatory requirements and related inspection procedures. 

The initial workshops were practical and based on a "hands-on approach".  Senior inspectors and supervisors attended these sessions and, in turn, trained inspection staff in the new requirements flowing from the  feed ban regulations. Workshops focussed on the key points for inspectors to concentrate on when auditing plants to ensure that these facilities were properly implementing the requirements of the feed ban.  The workshop training also covered the controls and processes that facilities had to implement in order to comply with the feed ban's requirements.  In this interval, inspection staff also received HACCP and/or ISO training, which supported them in evaluating process-control systems.

By 2002, formal Training Modules had been developed to complement the inspector forms and to help standardized the approach to inspecting facilities, including planning, conducting inspections, reporting the results and follow-up.

The modules will form the basis for training feed inspectors leading to certification.  Modules are now being finalized.

A risk-based approach to non-compliance was also adopted to improve both the efficiency of inspections and compliance -- particularly for high-risk inspection items. This approach provides inspectors with better guidance on priorities for following up on non-compliance.

2.3 Industry compliance with feed ban requirements

The review team's third line of enquiry assessed whether the industry (rendering plants and feed mills) complied with the 1997 feed ban requirements.

For this line of enquiry the following criteria were used:

• Feed mills and rendering plants have complied with the  feed ban regulations, as evidenced by compliance with inspection requirements.
• Facilities have taken steps to address cases of non-compliance.
• Incidents of non-compliance are resolved in a timely manner.

For this part of the review, the primary sources of information were documentation (specifically a review of historical inspection records for the fiscal years spanning 2002 to up to January 14, 2005), data retrieved from the CFIA's Multi-Commodity Activities Program (MCAP), interviews with key participants in the feed program, and on-site inspections of 31 commercial feed mills and seven rendering facilities.

Industry introduced measures to comply with the feed ban

With the amendments to the Health of Animals Regulations, the feed industry was required to introduce measures to reduce the risk of cross-contamination of feed for cattle with prohibited material.   It is important to note, however, that while the feed ban requirements apply to the entire spectrum of the feed industry, only a small portion of the rendering and commercial feed mill industry is considered as being in the high-risk category. 

Historically, only a part of the rendering industry (13 out of 29 facilities) was considered to be high risk.    Most facilities have introduced separate production lines for prohibited and non-prohibited material to further reduce the risk of cross-contamination.  As of 2005, only two facilities have not yet introduced a separate production line approach.  Because rendering plants that mix production on the same line represent the greatest risk of cross-contamination, the practice of separating production lines greatly contributes to the regulatory goals of reducing the risk of introducing infected material into feed.  With respect to feed mills, in 2002-03, 120 of the 550 mills fit into the high-risk category.  By 2003-04, the number had decreased to 94 mills.   

Many of Canada's feed mills are members of the Animal Nutrition Association of Canada (ANAC).  In 1998, ANAC launched a national voluntary Feed Safety Program which was developed to assist feed manufacturers in implementing Good Manufacturing Practices (GMPs) for example, production and distribution records and written procedures for preventing cross-contamination.  Some 176 of the estimated 550 feed mills are HACCP-certified.  The certification audit is conducted by a third party consistent with HACCP requirements of the United Nations Codex Alimentarius Commission.  The audit also incorporates key elements of the CFIA's Food Safety Enhancement Program (FSEP), the USFDA HACCP program and the European Union's HACCP protocol.

As with the feed mill industry, renderers have taken steps to adapt their process controls which will enable the industry to comply with the feed ban.    For example, many renderers have adopted industry-generated, HACCP-based Quality Assurance Programs.  In 2001, 12 large Canadian rendering plants, responsible for 74% of the production, were audited by a third party to verify  compliance with their HACCP programs.

Industry compliance - trends and current status

The review team evaluated compliance data for the BSE-related inspection tasks at commercial feed mills for fiscal years 2002-03, 2003-04 and 2004-05, up until the January 14, 2005 cut-off date (partial year).  As part of this review, inspection teams consisting of CFIA feed program inspectors and program specialists and, in one third of the inspections, USDA and FDA officials carried out on-site inspections at a sample of 31 commercial feed mills and 7 rendering plants. The on-site inspections took place from January 31, 2005 to February 11, 2005. 

The MCAP database was found to be valid and reliable

CFIA's inspection database "the Multi Commodity Activities Program (MCAP)",  contains the results of the annual inspections of commercial feed mills and rendering plants.  MCAP became operational on April 1, 2002.   The information in this database was analyzed to determine trends in compliance rates.

The review team tested the reliability of the information in MCAP that was used to generate trends in compliance.  In order to do so, the team collected the original inspection reports for the last three years for the 31 feed mills and 7 rendering plants in the sample and analyzed the three-year trends for these facilities. 

The review team compared 1,065 specific task ratings that inspectors had assessed.  The team also compared the hard-copy inspection reports obtained directly from CFIA's field offices with the corresponding data in the MCAP electronic file.  Reviewers found 32 data transcription errors, which translates to an error rate of about 3 percent.  However, they noted that errors tended to cancel each other out to a large extent, resulting in a lower error rate. The review concluded that the database contents were reliable and, therefore provided a valid basis for assessing inspection results.

The scope of this analysis includes all inspection data available in the MCAP database for the period from April 1st 2002 to January 14th 2005, with the exception of inspections carried out in fiscal year 2002-03 using an older version of the inspection report.  These data have been excluded as there is no direct mapping of the feed ban-related inspection tasks from the older inspection report to the new inspection form used for most of the analysis period. In other words, the feed ban-related inspection tasks from the older version of the form included evaluations for medicated feeds and Good Manufacturing Practices (GMP) in addition evaluating tasks specific to the feed ban.

Trend analysis for compliance with BSE-related tasks at commercial feed mills (April 1, 2002 - January 14, 2005)

The team used extracts from the MCAP database to identify trends in compliance data for the BSE-related inspection tasks at commercial feed mills and rendering plants for fiscal years 2002-03, 2003-04 and for 2004-05, up until the January 14, 2005 cut-off date. 

Compliance rates for commercial feed mills and rendering plants for April 1, 2002 to January 14, 2005 are shown in Figures 1 and 2.  Every inspection involves assessing a number of feed-ban related inspection tasks, for example, labelling requirements, clean-up procedures and written documentation (see Appendix 2 for details). Each task, in itself, contains a number of standards which must be met in order to obtain a "satisfactory" rating for the task (see box).  For feed mills, the 13 tasks assessed contain 86 standards.  For renderers, the 18 inspection tasks contain 62 standards.  Appendix 3 includes more trend analysis by individual inspection tasks and discusses the effect on compliance rates.

Figure 1: Compliance on a Facility by Facility Basis

	[D]		[D]
click on images for larger view

Figure 1 shows that industry compliance, on a plant-by-plant basis, has been in the 92-97% range for feed mills and 90-97% range for renderers over the last three fiscal years. Figure 1 also shows major non-compliance items and indicates the low rate of such non-compliance.  (Minor items have not been included.)  The breakdown between major and minor non-compliance items is presented in Appendix 3.  Figure 3 in Appendix 3 indicates how including "minor" non-compliance items affects overall compliance rates for feed mills and renderers.  As noted above, major items can potentially result in cross-contamination whereas minor non-compliance items relate to administrative items (i.e. paperwork).

Figure 2:  Compliance on a Task by Task Basis

	[D]		[D]
click on images for larger view

Figure 2 illustrates the compliance trends at feed mills and rendering plants for this period for all BSE-related tasks. Overall compliance has been high and it has been improving during the period examined.

The data indicate clearly that the feed mill and rendering industries have a high and improving rate of compliance with the feed ban.

Instances of non-compliance are dealt with promptly

Another measure of compliance is the time that facilities take to deal with instances of non-compliance identified by CFIA's inspections. 

The number of days that an industry takes to resolve non-compliance issues as identified by inspectors is presented in Figure 3.  For both industries (particularly for the rendering industry), the average time to resolve an issue has decreased. This number reflects the time it takes for plants to remedy an unsatisfactory inspection item, and for inspectors to certify that a facility has taken the appropriate action.

Figure 3:  Time for resolving Non-Compliance Issues

	[D]		[D]
click on images for larger view

Commercial feed mills and rendering plants are currently inspected annually.  An on-site inspection is usually conducted in one to two days. Immediately after the inspection, the management of the facility receives the results of the inspection, and deficiencies (if any) are confirmed.  If corrective action is required, management submits an action plan to the inspector for approval, including the time to remedy non-compliance.  This period could be up to 60 days depending on the severity of the non-compliance issue; priority is given to resolving major non-compliance issues.  Inspection staff follow up to ensure that management has dealt with it appropriately.  The data show that the time taken to resolve instances of non-compliance is decreasing. 

Findings arising from the on-site selected inspections of the feed mills and rendering plants

For the on-site inspections, the team used a randomly generated weighted sample of feed mills.  A total of 31 feed mills were included in the sample - 27 from the high-risk category and 4 from the low-risk category. High-risk feed mills include those that receive prohibited material and manufacture feeds for ruminants. A sample from the remainder of the mills from the low-risk category, i.e. facilities in which there is no risk of cross-contamination, was also selected for inspection. The review team included low-risk facilities because they may have modified their practices since they were last inspected in 2003-2004.  These changes may have introduced prohibited material into their operation.

The review team ensured that high-risk rendering facilities were included in the sample.  All six rendering plants that had received Permits to Operate (effective April 1, 2004), allowing them to manufacture both prohibited and non-prohibited protein meals, were selected for analysis.  One rendering plant from the low-risk category was also included in the on-site inspections.   Between April 1, 2002 and January 14, 2005 a qualitative assessment of compliance of BSE-related inspection tasks at selected mills and renderers was also carried out.

Figure 4:  Compliance Trends for Selected Feed Mills and Renderers

	[D]		[D]
click on images for larger view

The results of the on-site inspections were similar to the historical trends in compliance for the entire population of commercial feed mills and rendering plants.  Figure 4 shows compliance rates for on-site inspections at feed mills and rendering plants over the period from April 1, 2002 to January 14, 2005. The high compliance noted in 2003-04 coincided with the discovery of the first domestic case of BSE. 

The findings from the on-site inspections, therefore, corroborate the results of the three-year compliance trends as recorded in the MCAP database.

Of the 31 feed mills randomly selected for an on-site inspection, the review team found one major non-compliance requiring an immediate response.  At this mill, the review team noted that the  daily production records of the batch mixer did not indicate that the mixer had been flushed between a production run of poultry feed, containing prohibited material, and a production run of ruminant feed.

Given this finding, the review team examined a further five days of batch mixer production records.  All records were found to be in compliance.  A corrective action plan required the mill to immediately institute new procedures.  Staff at this facility were trained on these procedural changes. Further enforcement action may be pursued.

SUMMARY

Canada's Feed Control Strategy

The design of the 1997 feed ban was based on sound analysis and procedures.  The regulatory content reflected up-to-date risk assessments and scientific information that emerged as the BSE situation changed.

When benchmarked against other countries whose indigenous cattle had been infected with BSE, Canada compares favourably.  For example, Canada implemented the feed ban well before any BSE cases had been detected in native cattle.

The 1997 feed ban was one of a suite of measures that Canada has adopted, over time, with respect to BSE. After BSE was discovered in the UK in 1986, Canada followed a highly responsive approach to preventing the emergence of the disease in this country through various means, including import controls, surveillance measures and monitoring of UK imports.

In 1996, the WHO recommended steps to protect animal and human health, Canada responded quickly and decisively to mitigate BSE-related risks. In 1997, Canada imposed the mammalian-to-ruminant feed ban. It was a strategic approach intended to protect animal health by closing one more potential pathway for BSE exposure.

According to the 1996 WHO recommendations on BSE, a feed ban constitutes a sound BSE risk-reduction strategy. The recommendations noted that in the UK, the BSE epidemic appears to have been due mainly to "recycling" material from cattle infected with the BSE agent back to cattle through contaminated feed. Given the UK experience, the 1997 feed ban, along with Canada's other measures, represents an essential step toward reducing BSE risks.

Implementation of the Feed Ban

In implementing the feed ban inspection program, the Agency launched a series of communication and educational initiatives.  They were aimed at the regulated feed industry and CFIA inspectors.  The CFIA ensured that inspectors had the tools and knowledge needed to appropriately inspect facilities and assess the level of compliance with the feed ban.

The feed industry has clearly taken steps to meet the requirements of the feed ban regulations.  For example, industry has developed manufacturing and other procedures specifically designed to avoid cross-contamination between feeds produced for cattle and other species.  Avoiding cross-contamination greatly reduces the risk of transmitting BSE to cattle.

Industry Compliance with Feed Ban Regulations

Industry's compliance rates with standards and regulations relating to areas of significant risk are high.  The actual compliance rates have improved since 2002. 

The review has demonstrated that overall, Canada's feed ban is effective.  It was designed according to clearly defined risks.  Industry's compliance with the feed ban is already high, and compliance rates are increasing.  As well, CFIA inspectors have worked closely with the feed mills and rendering plants to achieve high and increasing levels of compliance.

Canada's feed ban has contributed to reducing the risk of BSE in Canada to an extremely low level and will lead to the elimination of the disease over time. The results of Canada's surveillance program to date bear this out, based on the small number of cases found and the age of the affected animals.  Proposed enhancements to the ban would serve to further shorten the time required to achieve complete elimination of BSE from Canada.

PART THREE

Appendix 1 - About the Review

Objectives

The review of the feed ban has three objectives:

• to determine the extent to which the design of the 1997 feed ban regulations was appropriate;
• to assess CFIA's approach to implementing and enforcing the feed ban program; and
• to assess the extent to which the rendering and feed mill industries have complied with the 1997 feed ban regulations.

Scope and Approach

The review took place from January 14, 2005 until February 18, 2005 and included a review of pertinent documents (including but not limited to regulatory impact analysis statements, historical inspection records, risk assessments and panel reports, training documents); interviews with key CFIA staff members at the program and operational level; and on-site inspections of selected sample of 31 feed mills and 7 rendering facilities.  The feed ban regulations apply to feed retailers, livestock producers and on-farm feed mills as well.  However the review focussed on commercial feed mills and rendering plants, as they represent a higher risk in terms of potential contamination of non-prohibited material or ruminant feed with prohibited material.

For the on-site inspections, the CFIA stratified the 550 feed mills and 29 rendering facilities with respect to potential risk of contamination of ruminant feed with prohibited material (high risk).  94 feed mills and six rendering plants fall into the high-risk category.  From the 94 feed mills, 27 were randomly selected from the high-risk category and 4 from the low-risk category (those that do not handle prohibited material or do not manufacture ruminant feeds).  From the high-risk category for rendering plants, all six  rendering plants and one rendering facility recently converted to the low-risk category were inspected as part of this review.

Inspection records were reviewed for the fiscal years 2002-03, 2003-04 and 2004-05 up to January 14, 2005.  The review team examined three years of data from  CFIA's electronic database (Multi-Commodity Activities Program - MCAP).  Reliable data has been available only for the past three years from MCAP, which was launched in 2002

Three lines of enquiry were identified:  the design of the 1997 feed ban regulations, the implementation of the 1997 feed ban regulations, and the application of the regulations by the feed mill and rendering industry.

Criteria

The review criteria were drawn from generally accepted criteria for the design of regulatory programs, CFIA policies, regulations and standards (including the 1997 Feed Ban itself), standards developed by professional organizations, and generally accepted inspection practices.  The following are the criteria that were applied to the lines of enquiry for this review:

For the design of the 1997 feed ban regulations, the review criteria represent a set of expectations that organizations should meet when developing regulations. Accordingly, the reviewers expected that in developing the feed ban regulations CFIA would have:

• followed a proactive, anticipatory approach in developing the feed ban regulations;
• carried out appropriate analyses and gathered comprehensive, relevant information;
• communicated and consulted effectively with stakeholders.

For the second line of enquiry involving CFIA's implementation of the 1997 feed ban, the following generally accepted criteria for assessing regulatory-type programs:

Timely, efficient and effective implementation of program structures and activities includes the following requirements:

• A sound program framework and prescribed standards for assessing compliance should be properly documented and communicated to staff and regulated parties.
• Appropriate inspection tools to assess compliance should be available for use by trained and professional inspection staff.

For the third line of enquiry the following criteria were used:

• Feed mills and rendering plants have complied with the feed ban regulations, as evidenced by compliance with inspection requirements.
• Facilities have taken steps to address cases of non-compliance.
• Incidents of non-compliance are resolved in a timely manner.

The Review Team

The team was lead by Corporate Planning, Reporting and Accountability Branch of the CFIA along with assistance from the Centre for Public Management and Jordan, Nesbitt and Associates Limited.  The team members are as follows:

Canadian Food Inspection Agency:
Tom Beaver, Executive Director, Corporate Planning, Reporting and Accountability
Matthew Gaetz, Administrative Coordinator, Operations Branch
Allen Good, National Feed Coordinator, Feed Section, Programs Branch
Dr. Harpreet Kochar, Ph.D, Operations Coordinator, Animal, Operations Branch
Judy Thompson, Feed Manufacturing Coordinator, Feed Section, Programs Branch
Tianna MacInnes, Audit Officer, Corporate Audit and Performance Measurement

Centre for Public Management
Tom Hopwood, M. Eng, P. Eng, Partner
Bill Kessels, CA, CMC, CISA, Principal

Jordan, Nesbitt and Associates Limited
Malcolm Jordan

For information, please contact Tom Beaver, at (613) 225-2342

Appendix 2 Inspection Tasks for Feed Mills and Rendering Plants

The following is a summary of the tasks related to the feed ban regulations:

Feed mills

• Assess adequacy of feed and feed ingredient labels (prohibited material)
• Assess adequacy of feed invoices
• Assess the adequacy of written procedures and documentation for the disposition of returned* and recalled feeds containing prohibited materials (*feeds returned by customers and retailers)
• Assess the adequacy of written procedures and documentation for the disposition of flush or recovered materials* containing prohibited material (*materials recovered from spillage, dust collectors, etc.)
• Assess the written procedures and records regarding the reuse of used packaging  for prohibited material 
  • Assess the adequacy of written clean-out procedures and production records
  • Verify that employees are following procedures in the following areas (within this task there are the following tasks):
    • Receiving equipment
    • Ingredient storage and handling equipment
    • Ingredient processing equipment
    • Mixing equipment
    • Pelleting/extruding equipment (including pellet mill/extruder, cooler and sifter/shaker)
    • Packaging equipment
• Bulk finished feed storage and handling equipment
• Loading and unloading of bulk delivery vehicles

For the 13 tasks listed there are 86 standards

Rendering Plants

• Assess the adequacy of on-site production records.
• Assess the adequacy of distribution records - prohibited material (product invoices, bills of lading etc. for bulk and bagged product deliveries)
• Assess the adequacy of labelling - prohibited material (if labels used to accompany bulk deliveries or on bags)
• Assess the adequacy of written procedures and records for clean out of all processing lines and equipment used in the manufacture of finished products (within these tasks there are the following subtasks):
  • raw material delivery vehicles/equipment 
  • raw material receiving 
  • raw material handling and storage
  • cooking 
  • pressing and milling 
  • finished product packaging equipment
  • finished product packaging (reuse of product packaging)
  • finished product handling and storage
  • finished product bulk loading and shipping
  • finished product delivery vehicles/equipment operated by the rendering plant   
  • finished product delivery vehicles/equipment not operated by the rendering plant 
• Assess the adequacy of written procedures and records for identifying and handling flush material
• Assess the adequacy of records of manufacturing errors and corrective actions taken
• Assess the adequacy of written procedures and records for the handling and end-use of recovered* materials within the rendering plant.- prohibited material (* materials recovered from spills, leaks, manufacturing errors [rework, returns, recalls], waste water treatment etc.)

For the 18 tasks listed there are 62 standards

Appendix 3 - Historical Compliance Trends:  Feed Mills

Overall feed mill compliance has been showing an upward trend over the past three years.  The objective of this appendix is to present this trend data, both on an overall and by-task basis in order to demonstrate the compliance rates of the feed mill industry over the past three years.  The following areas will be covered:

• overall compliance by task on a three-year basis;
• percentage of unsatisfactory tasks, by task, on a three-year basis;
• percentage of feed mills with no unsatisfactory tasks on a three-year basis;
• major and minor unsatisfactory tasks by feed mill, on a three-year basis

As discussed earlier in the report, the scope of this analysis includes all inspection data available in the MCAP data base for the period from April 1st 2002 to January 14th 2005, with the exception of inspections undertaken in fiscal year 2002-2003 under a previous version of the inspection report.  This data has been excluded as there is no direct mapping of the tasks from the previous inspection report to the new inspection report used for the majority of the analysis period.

Figure 1: Overall Compliance, by Task, on a Three-Year Basis for Feed Mills

	[D]
click on image for larger view

Overall task compliance demonstrates the total number of "satisfactory" and "not applicable" tasks, as a percentage of total inspection tasks.  It shows an overall high level of compliance, which has increased over the past three years.  Task 8, with the lowest percentage of compliance, relates to the adequacy of the details included on the feed invoices and is considered a "Minor Unsatisfactory" item in the Major/Minor classification system used in this report.

Figure 2:Percentage of Unsatisfactory Tasks, by Task, on a Three-Year Basis for Feed Mills

	[D]
click on image for larger view

This chart presents the unsatisfactory tasks as a percentage of total tasks, with the "not applicable" tasks removed. "Not Applicable" tasks indicate that the facility does not use prohibited material, or, if they do, the same equipment is not used to process or handle both prohibited and non-prohibited material.

It also demonstrates the reduction in unsatisfactory tasks over the last three years.  In many cases, a mill may have only one or two applicable tasks, with the others not applicable because the plant does not use a particular machine or process.  For this reason the overall compliance rate presented in Figure 1 (previous chart), which includes these not applicable tasks, presents a more balanced view of compliance.

Figure 3: Percentage of Feed Mills and Rendering Facilities with no Unsatisfactory Tasks on a Three Year Basis

	[D]		[D]
click on images for larger view

One weaknesses of examining compliance by task is that it does not present the complete picture of industry compliance.  Because all compliance deviations are counted, it is possible that a small number of plants with a high relative number of deviations each could skew the results.  The analysis of compliance by plant in this section demonstrates that the percentage of plants with no deviations has risen from approximately 77% in 2002-2003 to approximately 87% in 2004-2005.   While there has been progress to date, significant work remains in this area.

Table 1: Major and minor unsatisfactory tasks by feed mill, on a three year basis

	[D]
click on image for larger view

This analysis further breaks down the unsatisfactory tasks by plant into major and minor categories.  The current inspection system does not categorize unsatisfactory tasks.  These tasks were classified based upon the information contained in the MCAP system and hard copy inspection reports when available.  For this reason, some "Other" tasks were not classified.  However the analysis provides useful information on the breakdown of tasks.  The number of feed mills with at least one major unsatisfactory task has been steady in the 3-5% range over the last two years, while the majority of the improvement in the overall rate of plant compliance (outlined in the chart in the previous section) has been gained through a reduction of minor tasks.

Table 2: Major and minor unsatisfactory tasks by rendering facility, on a three year basis

	[D]
click on image for larger view

This analysis further breaks down the unsatisfactory tasks by plant into major and minor categories.  The current inspection system does not categorize unsatisfactory tasks.  These tasks were classified based upon the information contained in the MCAP system and hard copy inspection reports when available. Rendering facilities with at least one major unsatisfactory task  has been steady at one plant in each of the last two years.

• Main Page - BSE in North America