Interim guidelines for the control of verotoxinogenic escherichia coli including E. Colio157:h7 in ready to eat fermented sausages containing beef or a beef product as an ingredient

Guideline no. 12
Issued by Food Directorate
Health Protection Branch
Health Canada

February 24, 2000

1 Scope

Raw fermented sausage has recently been implicated in several foodborne outbreaks. Verotoxigenic strains of Escherichia coli and in particular E. coli O157:H7 has been identified in finished product. Research has found that some methods used to manufacture fermented sausage do not control or eliminate this pathogen from the finished product. These products are considered ready-to-eat and have from time to time posed a risk to the consumer. Infection with these organism has serious and sometimes fatal consequences and therefore, additional interventions are proposed to protect the public health. To date, outbreaks of E. coli O157:H7 reported in association with dry/semi-dry fermented sausages have been linked to beef meat ingredients. This guideline describes the additional interventions that are recommended for the production of ready-to-eat fermented sausages containing beef as an ingredient or where there is a risk of cross-contamination from beef. Following appropriate consultation with the industry and consumers groups etc. these guidelines will be developed into a regulation.

2 Introduction

Raw fermented sausage was found to be the cause of an outbreak of E. coli O157:H7 in the US in 1994. In October 1995, it became clear that the E. coli strain that was isolated could survive the acidic conditions required for manufacturing of raw fermented sausage. Health Canada took steps to notify the Meat Industry, Agriculture Canada and Regional Directors of Health Protection Branch of this new potential hazard.

In December 1996 and again in September 1998 the Canadian Food Inspection Agency notified dry and semi-dry fermented sausage manufacturers of the results of a US Blue Ribbon Task Force on Raw Fermented Sausage and recommended that establishments adopt one of five interventions proposed by the USDA to address the E. coli O157:H7 concern in these types of products.

However, in the spring of 1998, another Escherichia coli O157:H7 outbreak was traced to a naturally fermented Genoa Salami product manufactured by a registered establishment in Ontario. As part of the follow-up to this outbreak the natural fermentation process was identified as a "high risk" process and a survey was proposed to identify other manufacturers that were using similar high risk processes. The survey was to include a thorough review of the manufacturing process and were appropriate samples of the final product were to be taken for microbial analysis.

Due to resource limitations higher priority activities the survey was never completed. Again in November 1999, another E. coli O157:H7 outbreak in Western Canada was traced to a similar type of raw, fermented, Hungarian-style sausage. Over 150 people became sick and at least five developed hemolytic uremic syndrom. In the investigation that followed this outbreak it became clear that the interventions previously recommended to improve the safety of dry and semi-dry fermented sausages were not being followed in some establishments. As a result regulatory action is being taken to address these concerns. It is our hope that this action will result in an improvement of the safety of these types of products and that it will establish equivalent requirements for registered and non-registered establishments.

3 Guidance for the Safe Manufacturing of Fermented meat products

3.1 Introduction

The Canadian Food Inspection Agency's Meat Hygiene Manual of Procedures (Chapter 4 section 10) provides a summary of pertinent information for the safe manufacturing of fermented meat products. This section includes a discussion of the criteria used to assess different types of sausage (i.e. dry, semi-dry, shelf-stable) based pH and the water activity levels in the final product and pathogens of concern. The need for quality raw materials and ingredients (i.e. by continuous monitoring) is stressed as well as the need to demonstrate control over the fermentation production process such as degree-hour measurements. This document spells out the requirements for facility and equipment, and the manufacturing controls and specifications for ingredients that are required for federally registered establishments (Chapter 4.10.c and d). These sections are useful in for the inspection of registered and non-registered manufactures of ready-to-eat fermented meat.

3.2 Requirements for Non Fermented Sausage

Please note that, with the exception of meat products made by a retort process, shelf-stable non-fermented meat products must have a finished product water activity (a_w) of 0.85 or less or a pH of 4.6 or less. If the final water activity of a non-fermented sausage product exceed 0.85 or the pH is higher than 4.6, it should be stored under refrigeration and be labeled appropriately (i.e. "Keep Refrigerated")

4.0 Controls to address hazards related to verotoxinogenic E. coli (e.g., E. coli O157:H7) and Salmonella in fermented sausages

In order to suitably control these hazards and prevent incidents of food borne disease, establishments which manufacture fermented sausages are required to use one of the following interventions for the control of verotoxinogenic E. coli includingE. coli O157:H7 and Salmonella when they make this type of product.

If an establishment does not follow one of the interventions described, they are automatically considered to be using intervention 3, end product testing. Product may be detained and a Health Risk Assessment requested when an establishment following intervention 3 refuses to do the required testing on the finished product.

4.1 Intervention 1: Include as part of the manufacture of the sausage, one of the following heat process which is recognized as controlling E. coli O157:H7.

Under this intervention, it is not required to test for E. coli O157:H7. Time and temperature controls should be documented in the same manner as is required for other similar cooking processes.

Heat Processes Recognised as Controlling E. coli O157:H7

Minimum internal temperature maintained during the entire process		Minimum processing time in minutes after the minimum temperature has been reached
(°F)	(°C)
Table 1 footnotes Table 1 footnote 1 This table is identical to the roast beef cooking table with one exception: the minimum processing time for a minimum internal product temperature of 145°F/62.8°C is 4 minutes instead of "instantaneous". This difference is because the sausage product's smaller size results in a much quicker cooling and decreased cumulative lethality. Return to table 1 footnote 1 referrer
130	54.4	121
131	55	97
132	55.6	77
133	56.1	62
134	56.7	47
135	57.2	37
136	57.8	32
137	58.4	24
138	58.9	19
139	59.5	15
140	60	12
141	60.6	10
142	61.1	8
143	61.7	6
144	62.2	5
145	62.8	4Table 1 footnote 1

4.2 Intervention 2 : Use a manufacturing process (combination of fermentation, heating, holding and/or drying) which has already been scientifically validated to achieve a 5 D reduction of E. coli O157:H7.

Manufacturing processes used to make fermented sausages are only considered effective against E. coli O157:H7 if it is shown that they reduce the level of E.coli O157:H7 by 5 logs (for example from 100,000 cfu/g to less than 1 cfu/g). This is referred to as a 5D process or a 5 log reduction. The manufacturing process used must be evaluated in a scientific manner consistent with the challenge study recommendations (Annex 1).

Under intervention #2, it is not required to test each lot for E. coli O157:H7 or Salmonella. The operator shall nevertheless conduct some degree of testing for these organisms as a verification procedure for their process.

The operator must maintain suitable records to demonstrate that all of the critical control points (CCP) for the process have been met (for ex., casing diameter, fermentation room (green room) thermographs, pH at the end of the fermentation step of the process, a_w, etc.)

The following processes have been scientifically validated as achieving a 5D or greater reduction of E. coli O157:H7.

Combination of Process Parameters Validated to Achieve a 5-D Reduction of E. coli O157:H7

Fermentation chamber temperature		pH at the end of fermentation process	Casing diameter	Subsequent process (dry, hold or cook)	Ref .
°F	°C
Table 2 footnotes Table 2 footnote 1 Nicholson, R., et al, Dry fermented sausage and Eshcerichia coli O157:H7.National Cattlemen's Beef Association, Research Report No. 11-316, Chicago, IL, 1996. Return to table 2 footnote 1 referrer Table 2 footnote 2 Hinkens, J.C., et al, Validation of Pepperoni Processes for Control of Escherichia coli O157:H7, Journal of Food Protection, Vol. 59, No. 12, 1996, pp. 1260-1266. Return to table 2 footnote 2 referrer
70	21	≥ 5.0	≤ 55 mm	HEAT (1hr @ 110°F and 6 hrs @ 125°F)	Table 2 footnote 1
90	32	≤ 4.6	≤ 55 mm	HOLD @ 90°F for ≥ 6 days	Table 2 footnote 1
90	32	≤ 4.6	≤ 55 mm	HEAT (1hr @ 110°F then 6 hrs @ 125°F)	Table 2 footnote 1
90	32	≤ 4.6	56 to 105 mm	HEAT (1hr @100°, 1hr @110°F, 1hr @120°F, then 7hrs @ 125°F)	Table 2 footnote 1
90	32	≥ 5.0	56 to 105 mm	HEAT (1hr @100°, 1hr @110°F, 1hr @120°F, then 7hrs @ 125°F)	Table 2 footnote 1
96	36	≤ 5.0	≤ 55 mm	HEAT (128°F internal product temperature x 60 minutes) and DRY (at 55°F and 65% Relative Humidity to a Moisture Protein Ration of ≤ 1.6:1)	Table 2 footnote 2
110	43	≤ 4.6	≤ 55 mm	HOLD @ 110°F for ≥ 4 days	Table 2 footnote 1
110	43	≤ 4.6	56 to 105 mm	HOLD @ 110°F for ≥ 4 days	Table 2 footnote 1
110	43	≤ 5.0	56 to 105 mm	HOLD @ 110°F for ≥ 7 days	Table 2 footnote 1

4.3 Intervention 3: Microbiological end-product testing must be done on each production lot and the lot held pending reception of results where the manufacturing process does not correspond to one of the processes set out under intervention 1, 2 , 4 or 5

1. Definition of "lot": The definition of "lot" for purposes of sampling must be statistically sound and must correspond to product manufactured under the same conditions.
2. Sampling plan: For each lot, the operator shall take 30 samples of finished product and submit them for analysis. The sample plan must be representative of the lot.
3. Sample size: Each sample shall consist of at least 25 g of product. Samples must be taken in accordance to standard microbiological techniques to avoid contamination of product and sampling of intact product packages is strongly recommended. It is unacceptable to take multiple sample from one intact package as this is not considered statistically representative of the lot.
4. Compositing of samples by the laboratory for analysis: It is acceptable to combine a maximum of three (3) samples into a composite for purposes of analysis when testing is done for E. coli O157:H7 and Salmonella.
5. Organisms to be tested: At a minimum, each composite sample shall be tested for the presence of E. coli O157:H7 and Salmonella.
6. Laboratory requirements: CAUTION ! - Since E. coli O157:H7 are pathogenic to humans, the tests should be done in a laboratory that has the proper equipment for containment by personnel trained in handling level 2 pathogens.
7. Method used: The method used to analyze the end product samples shall be one of the methods listed in Health Canada's Compendium of Analytical Methods, Vol. 3 Health Canada (ISBN 0-921317-17-4).
8. Reporting of results: Results shall be reported in writing. Results shall be identified to the lot of product being tested and shall include individual results for each test performed, method used, minimum sensitivity of the test used, lot for which these results apply.
9. Release of product: Product will be held under the control of the operator until the written results of analysis have been reviewed and found acceptable (i.e. negative for the presence of E. coli O157:H7 and Salmonella).
10. In case of a positive result for either E. coli O157:H7 or Salmonella: the entire lot must be held and either submitted to process verified to achieve a minimum 5D reduction or the product must be destroyed. Possible cross-contamination of other lots shall also be assessed.
11. Keeping of records: Records of test results shall be kept for a minimum of 24 months beyond the release date of the product.

4.4 Intervention 4: Implement a HACCP system at the establishment which includes testing of raw meat and batter, and use a manufacturing process (fermentation and holding, heating and/or drying) which has been scientifically validated as achieving at least 2 D reduction of E. coli O157:H7.

To be eligible to use this intervention, the operator must have implemented a HACCP system which meets the requirements of the CFIA's FSEP approach (Related information could be found on CFIA's Web site at http://www.cfia-acia.agr.ca/english/ppc/haccp/haccp.html). Sampling of raw batter must be done in accordance to the requirements set out in parts (a) to (k) below.

Manufacturing processes used to make fermented sausages are considered partially effective against E. coli O157:H7 if they can be shown to achieve between 2D and 5D reduction of E. coli O157:H7. The manufacturing process used must be evaluated in a scientific manner consistent with the Challenge Protocol (see section Annex 1).

1. Definition of "lot": The definition of "lot" for purposes of sampling must be statistically sound and must correspond to like production practices. Provided that effective controls for tracing product are in place and all corresponding dry fermented sausage manufacturing processes have been validated as achieving at least a 2D reduction of E. coli O157:H7, it would be acceptable to conduct one single series of sampling on batter which may be used thereafter in different sausages.
2. Sampling plan: For each lot, the operator shall take 15 samples of raw batter and submit them for analysis. The sample plan must be representative of the lot.
3. Sample size: Each sample shall consist of at least 25 g of product. Samples must be taken in accordance to standard microbiological techniques to avoid contamination of product. It is unacceptable to take multiple samples from one site as this is not considered statistically representative of the lot.
4. Compositing of samples by the laboratory for analysis: It is acceptable to combine a maximum of three (3) samples into a composite for purposes of analysis when testing is done for E. coli O157:H7 and Salmonella.
5. Organisms to be tested: At a minimum, each composite sample shall be tested for the presence of E. coli O157:H7 and Salmonella.
6. Laboratory requirements: CAUTION! - Since E. coli O157:H7 are pathogenic to humans, the tests should be done in a laboratory that has the proper equipment for containment by personnel trained in handling level 2 pathogens.
7. Method used: The method used to analyze the end product samples shall be one of the methods listed in Health Canada's Compendium of Analytical Methods, Vol. 3, Health Canada (ISBN 0-921317-17-4).
8. Reporting of results: Results shall be reported in writing. Results shall be identified to the lot of product being tested and shall include individual results for each test performed, method used, minimum sensitivity of the test used, lot for which these results apply.
9. Release of product: Product will be held under the control of the operator until the written results of analysis have been reviewed and found acceptable (i.e. negative for the presence of E. coli O157:H7 and Salmonella).
10. In case of a positive result for either E. coli O157:H7 or Salmonella: the entire lot must be held and either submitted to a 5D reduction process or be destroyed.
11. Keeping of records: Records of test results shall be kept for a minimum of 24 months beyond the release date of the product.
12. The following processes have been scientifically documented as achieving a minimum 2D reduction in E. coli O157:H7.

Process Parameters Documented as Achieving a Minimum 2-D Reduction in E. coli O157:H7

Fermentation chamber temperature		pH at the end of fermentation	Casing diameter	Subsequent process (dry, hold or cook)	Ref.
°F	°C
Table 3 footnotes Table 3 footnote 1 Nicholson, R., et al, Dry fermented sausage and Escherichia coli O157:H7.National Cattlemen's Beef Association, Research Report No. 11-316, Chicago, IL, 1996. Return to table 3 footnote 1 referrer
70	21	≥ 5.0	56 to 105 mm	HEAT (1hr @ 110°F and 6 hrs @ 125°F)	Table 3 footnote 1
90	32	≤ 4.6	56 to 105 mm	HOLD @ 90°F for 7 days then dry	Table 3 footnote 1
90	32	≥ 5.0	56 to 105 mm	HOLD @ 90°F for 7 days then dry	Table 3 footnote 1
110	43	≥ 5.0	≤ 55 mm	HOLD @ 110°F for 7 days then dry	Table 3 footnote 1
110	43	≥ 5.0	56 to 105 mm	HEAT (1hr @ 110°F and 6 hrs @ 125°F)	Table 3 footnote 1

4.5 Intervention 5: Use an alternative manufacturing process which is scientifically validated against E. coli O157:H7.

The manufacturer may make a request for the evaluation of an alternative manufacturing process by Health Canada, Director, Bureau of Microbial Hazards, Food Directorate, HPB Ottawa (i.e. a 5 D process that differs from those outlined in Intervention 2 or a 2D process with raw batter testing that differs from intervention 4. To allow the process to be evaluated, manufacturers shall use the same challenge protocol that was developed by the USDA and described below under Annex 1 Challenge Protocol. Because of the complex nature of the protocol, it is strongly recommended that the services of an experienced food technology center be retained.

Upon completion of a successful evaluation, the establishment shall be receive a letter of no objection indicating that the process has been evaluated for its ability to control E. coli O157:H7 and found acceptable. Until such confirmation is received, the operator will have to manufacture product in accordance to one of the other four interventions outlined above.

Health Risk Assessment when Fermented Sausage is Found Positive for E. coli O157:H7

If fermented sausages containing beef or beef ingredients are manufactured by processes other than those specified in intervention 1, 2 or 4 above and the establishment is not in possession of a letter of no objection, final product should be considered in violation of Section 4 of the Food and Drugs Act and is considered a Health 1 concern. Finding E. coli O157:H7 in the final product is considered a Health 1 concern. If E. coli O157:H7 is found in raw batter, any ready-to-eat sausages manufactured from this batter must receive a minimum 5D process or the entire batch of batter represented by the testing must be destroyed.

The following is a characterize of a Health 1 health risk:

Health 1 The health hazard identified represents a situation that could cause serious adverse health consequences or death. Appropriate action should be taken against the product to limit or prevent exposure in the population to the product. Such action should ensure that the product is no longer sold and the population does not consume what they have at home (eg. action at the consumer level if the product has been distributed). Follow-up action should ensure that the cause has been determined and appropriate corrective action has been taken to correct the problem.

Annex 1

Challenge Protocol for the Evaluation of a Fermented Sausage Manufacturing Process for the Ability to Control E. coli O157:H7

1. Biosafety requirements: CAUTION! - This protocol is a laboratory-based validation procedure that employs cultures that are pathogenic to humans. THE VALIDATION SHOULD NOT BE CONDUCTED WITHIN AN ACTUAL FOOD MANUFACTURING FACILITY. Work should be conducted in a biosafety level II facility by appropriately trained personnel. Following use, autoclave all inoculated product and sanitize processing equipment. Follow appropriate procedures for the disposal of waste.

2. Types and numbers of strains of E. coli O157:H7 to use as an inoculum: at least five (5) strains of E. coli O157:H7 should be used including representatives of strains associated with human illness and strains isolated from meat and poultry products. One isolate from an outbreak associated with a dry fermented sausage product must be included.^{Footnote 1}

3. Methods of production, enumeration and standardization of inoculum: Individual cultures of each strain should be prepared by inoculating an appropriate growth media, such as Tryptic Soy or Trypticase Soy broth, supplemented with 1% glucose and incubating for 18 to 24 hours at 37°C to obtain stationary phase cells. The additional glucose is added to ensure that the inoculum is pre-adapted for acid tolerance. Cultures should be grown the day prior to product inoculation with a minimum holding period prior to actual use. Each strain should be centrifuged, washed and resuspended in 0.1% peptone broth. Dilutions of each strain should be made to yield approximately equal numbers of each of the five strains. The five strains should be thoroughly mixed prior to being used as an inoculum. After the mixed working inoculum is prepared, the viable count of the mixture should be determined by direct surface plating on MacConkey sorbitol agar (MSA). Each of the individual strains in the inoculum should contribute about 20 percent of the total inoculum.

4. Size of inoculum to be used: the final concentration of E. coli O157:H7 in the meat mixture should be no less than 2.0 x 10⁷ cfu/g of meat mixture. The actual inoculum level in the meat mixture should be confirmed by sampling the inoculated meat mixture immediately after the inoculation using the above media. At this concentration, product can be serially diluted and direct plated without the need for enrichment to recover low levels of inoculum. The initial inoculum level was chosen to allow direct enumeration of at lest a 5 log reduction in the level of the inoculum between the initial count in the meat mixture and the finished product.

5. Method of inoculation to be used: the inoculum must be added to the meat and mixed prior to the addition of the other ingredients or a starter culture to the meat mixture. The use of a non-inhibitory, food grade, green dye added to the inoculum may aid in determining the uniform distribution of inoculum. The following procedure is recommended:

1. Add inoculum to meats while grinding or chopping the meats to the desired consistency
2. Mix in cure (if used), salt and spices.
3. Blend in starter culture (if used) near end of mixing cycle.
4. Stuff batter into casings.

6.Stuffing product into casings: Inoculated product should be stuffed into casing as usual to approximate normal production procedures. A shorter length may be used as long as the length is approximately twice the diameter of the stuffed casing.

7.Sample size, sampling time, sampling location and number of samples to test: Select two sausage sticks at the end of the drying period (finished product). From each stick selected, cut multiple cross-sectional slices from multiple locations on each stick to a final analytical sample weight of 25 g per stick.

8. Methods of microbial analysis: Blend each of the two 25 gram samples (one per stick) in separate 225 ml portions of buffered peptone water. Serially dilute the homogenates in buffered peptone water and surface plat 0.1ml portions from the dilutions onto MSA plates in duplicate. Count plates after incubation at 42°C overnight. Confirm 5-10 randomly selected colonies by serological and biochemical methods as necessary. Report count per gram of finished product. Report initial inoculum level.

9.Number of replicates: a minimum of three replicates of the study should be performed. Three separate formulation batches can, however, be processed concurrently following stuffing.

Therefore, total number of samples for microbiological analysis =

Time zero (0)	=	2
After fermentation	=	0
During drying	=	0
End drying	=	2
		4
Number of replicates	x	3
Total samples		12

10. Measurement of process parameters used to determine when a product is finished at each stage of production (process control criteria): Duplicate uninoculated samples of the product which are collected after stuffing and at each production stage should be assayed for moisture, fat, protein, salt content, pH, a_w, and titratable acidity.

Therefore, total number of samples for additional analysis =

Time zero (0)	=	2
After fermentation	=	2
During drying	=	2
End drying	=	2
		8
Number of replicates	x	3
Total samples		24