There is nothing more essential to meat and poultry safety than proper cooking to destroy vegetative bacterial pathogens of concern and cooling promptly to prevent outgrowth of spore-forming bacterial pathogens. Since 1999, U.S. federally inspected meat plants have been required to cook beef, roast beef, and cooked corned beef products to achieve at least a 6.5-log reduction of Salmonella, or to process to an alternative lethality (e.g., at least a 5-log reduction), per 9 C.F.R. 318.17(a)(1). The U.S. Department of Agriculture Food Safety and Inspection Service (USDA-FSIS) requires that cooked uncured meat patties must be processed to meet or exceed the times and temperatures listed in 9 C.F.R. 318.23, which will achieve a 5-log lethality. Cooked poultry products must be processed to achieve at least a 7-log reduction of Salmonella per 9 C.F.R. 381.150(a)(1). For other ready-to-eat (RTE) meat products, establishments must ensure that products are safe for consumption (i.e., free of pathogens) to produce unadulterated product and meet Hazard Analysis and Critical Control Points (HACCP) requirements. Also, establishments are required to cool RTE products to achieve “stabilization” per 9 C.F.R. 318.17(a)(2). “Stabilization” means there can be no multiplication of toxigenic microorganisms such as Clostridium botulinum and no more than 1-log10 multiplication of Clostridium perfringens within the product as it passes through the growth temperature range of pathogens during cooling down to refrigeration temperatures.

Updated Compliance Guidelines
To aid the industry in meeting cooking requirements, FSIS had published a document called Appendix A, Compliance Guidelines for Meeting Lethality Performance Standards for Certain Meat and Poultry Products in 1999. After revisions in 1999, 2011, and 2012, FSIS issued the most substantial revision to these guidelines in June 2017, to provide clarification of regulatory requirements for RTE products [Salmonella Compliance Guidelines for Small and Very Small Meat and Poultry Establishments that Produce Ready-to-Eat (RTE) Products and Revised Appendix A]. The revised compliance guideline also provides additional options for achieving lethality of Salmonella in RTE products, updates the lessons learned from food safety assessments, and combines and replaces information from previously issued guidance documents. In June 2017, FSIS also published FSIS Appendix B Compliance Guidelines for Cooling Heat-Treated Meat and Poultry Products (Stabilization) of the final rule, “Performance Standards for the Production of Certain Meat and Poultry Products” (64 F.R. 732) and FSIS Directive 7110.3, Rev. 1, Time/Temperature Guidelines for Cooling Heated Products, dated January 24, 1989, and the June 2017 FSIS Compliance Guideline for Stabilization (Cooling and Hot-Holding) of Fully and Partially Heat-Treated RTE and NRTE Meat and Poultry Products Produced by Small and Very Small Establishments and Revised Appendix B June 2017 Compliance Guideline.

Although both revised compliance guidelines were emphasized to not be considered requirements but rather a best practice, guidance, or recommendation, the recent updates created concern across the industry due to new clarified stipulations concerning the use of safe harbors for cooking and cooling in appendices A and B, respectively. FSIS published Notice 17-18, which stated that industry had until March 22, 2019, to review “new” revised appendices A and B to determine whether additional scientific support was needed to support the establishment’s process per 9 C.F.R. Section 417.5(a)(1) and to validate any changes to be made to the establishment’s process to follow the new guidelines or additional support over 90 calendar days [9 C.F.R. Section 417.4(a)(1)].

Revised Appendix A
Most cooked meat and poultry processors cite the Appendix A safe harbor time and temperature combinations as the scientific basis for their critical limits for cooking to eliminate Salmonella. Indeed, the use of Appendix A extends beyond meat and poultry cooking, as producers of other products such as veggie patties, meatless frankfurters, or meatless entrees sometimes also rely on these guidelines. However, since the Appendix A cooking times and temperatures were based upon thermal death curves for Salmonella in beef emulsions in tubes,[1] utilization of these parameters is not necessarily valid for every product matrix.

The impetus for the revision to Appendix A is in large part the agency’s desire to provide clarification about humidity in cooking. FSIS also referred to information from food safety assessments in several establishments that noted lack of control of Salmonella, as well as several Salmonella outbreaks related to RTE meat and poultry products, which appeared to be not related to inadequate cooking but rather to Good Manufacturing Practices or post-lethality recontamination of products. The agency’s baseline data of Salmonella in RTE products dating back to 2009 revealed an overall rate of 0.05 to 0.06 percent (Table 1). Pork products were the sources of over half (21/37) of all Salmonella-positive RTE samples.

The revised document includes a definition of pasteurization and guidance on how to label products as pasteurized. It also provides lethality requirements for certain RTE products, an overview of hazards of ingredients added post-lethality, and guidance on post-processing handling and sanitation. The guidance clarifies use of humidity in lethality processes, the time-temperature tables from Appendix A, and tables for achieving lethality in chicken and turkey RTE products and in nonintact meat chops, roasts, and steaks. The tables for chicken and turkey include time and temperature safe harbors based upon fat content. There are recommendations on how to support a 5-log reduction of Salmonella and clarification on how establishments can support an alternative lethality of at least 5-log reduction of Salmonella for roast, cooked, and corned beef as required in 9 C.F.R. Section 318.17(a)(1). FSIS provided clarification that defined a 5-log reduction of Salmonella, Listeria monocytogenes, and Escherichia coli O157:H7 as the target for shelf-stable products, which would include products such as dry cured ham and prosciutto. The document also discusses the agency’s expectations for how processors should evaluate cooking deviations, post-processing handling and sanitation, and potential corrective actions for FSIS positive test findings. Concerning deviations, the agency clarified that during cooking, the come-up time from 50 to 130 °F should occur in less than or equal to 6 hours. A longer come-up time or interruption would constitute a deviation, requiring an assessment of the potential for growth and toxin production by Staphylococcus aureus and Bacillus cereus. If there is potential for 3 log of S. aureus growth, the agency expects the processor to test for staphylococcal enterotoxins and B. cereus emetic toxin.

The study by Goodfellow and Brown,[1] from which original Appendix A tables were derived, also investigated inactivation of Salmonella on the surface of beef rounds during dry oven roasting. The study determined that if steam was injected into the smokehouse for at least 30 minutes during cooking to an internal temperature of 130 °F, then a 7-log10 reduction could be achieved. The greater efficacy of moist heating compared with dry heating for inactivation of Salmonella is well established.2–4 During convective heating, sufficient moisture by volume of air in the heating space can facilitate condensation on the surface of meat, resulting in moist heating. Conversely, in a heating vessel with low moisture by volume, evaporative cooling occurs as the moisture mass transfers out of the meat and into the air.[2] Ergo, FSIS had been concerned for several years about industry utilization of Appendix A time-temperature combinations without apparent regard to humidity within cook ovens. The importance of humidity in cooking, especially at mild temperatures, was further realized when a number of Salmonella outbreaks from jerky were attributable in part to inadequate humidity during heating.[5,6] Since then, jerky heating was specifically addressed by FSIS in a 2014 compliance guideline, but until now, detailed guidance for all other products did not exist. FSIS has stated it has not changed the humidity recommendations within Appendix A other than “re-emphasizing” that they apply to all cooked products (including poultry), unless the establishment can support that humidity does not need to be addressed. Options for supplying adequate humidity to a process that relies on Appendix A time-temperature combinations were spelled out in the guidelines (Figure 1), and both attainment and accurate measurement of these humidity requirements are the primary concerns of industry.  

For example, “impingement”-style ovens with continuous cooking have openings from which moisture and heat can readily escape, making it difficult to maintain humidity throughout a cook process. Also, some smokehouses in very small establishments lack dampers, and so humidity control would be impossible. Yet, the guidance states that measurement accuracy for cooking should include time ± 1 minute, temperature ± 1 °F, and humidity ± 5 percent. Further, it states that humidity “should be” part of Critical Control Points (CCP) monitoring of critical limits or prerequisite programs. Several types of products and/or cooking techniques are exempt from the humidity requirements and include:

•    Immersion in cooking liquid medium

•    Cook-in-bag

•    Direct heat (e.g., grill, heating coil, flame, or rotisserie) due to rapidity of surface heating

•    Semipermeable, impermeable product casing (such as frankfurters)

•    Cooking beef patties (due to direct heat)

Revised Appendix B
“Stabilization” refers to cooling cooked meat and poultry expediently to delay the germination and subsequent outgrowth of spore-forming bacterial pathogens of concern, namely C. botulinum, C. perfringens, and B. cereus. The necessity of treating the cooling, or stabilization, control step in cooked meat and poultry processing as a CCP in USDA-FSIS-regulated establishments has been a point of debate for several years. Among some of the points made by industry scientists is that in 1998, FSIS overestimated by a very wide margin the number of surviving spores in meat and poultry products after cooking because the baseline data involved raw meat and poultry samples analyzed by a method that had no prior heating step to destroy the vegetative cells, so the number of spores could not be differentiated and the colonies were not confirmed to be C. perfringens (RB Tompkin, personal communication). This led to very conservative time and temperatures being required for cooling in the 1999 version of Appendix B (i.e., no greater than a 1-log increase in C. perfringens). Several industry studies were subsequently published clarifying spore levels in raw meat and poultry samples, demonstrating growth control in actual meat and poultry products during extended cooling and die-off of vegetative cells of C. perfringens during subsequent refrigerated storage.[7–9] Indeed, the USDA risk assessment concluded that C. perfringens growth during stabilization of RTE or partially cooked (PC) meat and poultry products has a small overall effect on the likelihood of illness, and that growth during retail and consumer storage is the major predicted cause of illnesses from C. perfringens in RTE/PC meat and poultry products.[10] Although industry had contended that no outbreaks had been attributable to industrially produced meat and poultry products, FSIS cited in Revised Appendix B one outbreak associated with C. perfringens from commercially produced RTE turkey loaf product in 2000. The agency also took the opportunity to clarify a number of compliance criteria for heat-treated, not fully cooked products. FSIS acknowledged that food safety concepts associated with RTE products may also apply to heat-treated not-ready-to-eat (NRTE) products, such as ready-to-cook bacon.

The revised document includes recommendations previously provided in Appendix B, FSIS Directive 7110.3, and examples of scientific evidence for establishments to support a safe production process. Several options for cooling meat and poultry products were provided, including requesting a waiver from regulatory performance standards to allow up to a 2-log multiplication of C. perfringens within a product, provided there’s no growth of C. botulinum, if the establishment collects data from raw materials that demonstrate a relatively low level.

FSIS provided four recommendations for cooling now that limit growth of C. perfringens to be less than 1 log with no growth for C. botulinum (Table 2).

In an attempt to allay some of the industry concerns about achieving Option 2, FSIS published some recommendations for establishments that cannot meet the stabilization guidelines.11 Recommendations included pre-chilling the cooler before loading product; increasing airflow (e.g., adding a fan) to speed cooling; and reducing the amount of product in each batch or lot placed in the cooler at one time to reduce the total heat load to be removed. Additional answers to public comments were provided pertaining to additional cooling options for establishments to use based on modeling.[12] The agency also published additional information and guidance pertaining to the use of Option 3 for stabilization of products containing natural sources of nitrite and ascorbate.[13,14]

FSIS outlines corrective actions for an establishment to consider when a cooling deviation occurs, including pathogen modeling. Growth of more than 1 log C. perfringens (or 2-log growth) and greater than a 0.30-log increase of C. botulinum would mean that product should be destroyed. The agency also recommended that processors assess B. cereus growth only if C. perfringens growth is estimated greater than 3 log. To aid product disposition determinations, product testing for C. perfringens should follow n = 10, c = 2, m = 100, and M = 500. Due to the work of Mohr and colleagues,[15] the compliance guideline clarifies that ComBase’s Perfringens Predictor and the Smith-Schaffner models are most accurate for use in predicting growth of C. perfringens in meat systems. If evaluating existing or conducting new validation studies, the recommendations in the FSIS Compliance Guideline: HACCP Systems Validation should be consulted.[16] The types of scientific data that can be used for validation include in-plant data, modeling, challenge studies, and other regulatory/public health agency reports (e.g., Food Code, Canadian Food Inspection Agency, European Food Safety Authority, Codex).

Besides the industry concerns about Appendix B previously mentioned, there are additional concerns that the guideline contains general information that could be misleading. For instance, the guidance document cites from the U.S. Food and Drug Administration, “The minimum inhibitory salt concentration for C. perfringens is 7%, and 10% for C. botulinum.” However, these limits are in growth media incubated at ideal temperatures. It’s been reported that 3 percent salt completely inhibits C. perfringens in beef and ham during extended cooling for 21 hours.[17] Another term used is in reference to the use of peer-reviewed publications as scientific support for cooling. The guideline states that in order to use such scientific support as validation, processors must ensure that their cooling data “matches” that of the scientific support; however, there is no guidance or definition as to what constitutes a suitable match. Another such phrase concerning scientific references is that “if establishment does not follow all parts,” then the scientific reference may not be applicable. The term “all” could be misconstrued such that no scientific reference would be suitable and that every process would need to be validated by conducting a laboratory-scale or pilot-scale study or in-plant validation with surrogates. This would be unduly cumbersome and costly to industry, and small and very small establishments usually lack the resources to perform such work. The guidance mentions that “critical operating parameters” need to be predefined. This is of course helpful, but these types of phrases will surely lead to much debate about what is and what is not “critical” between in-plant personnel of FSIS and food company personnel attempting to explain their scientific support and validation for cooling.

Future Revisions Likely
FSIS has provided much more detail than in previous versions of compliance guidelines for appendices A and B. The agency has taken an approach toward these guidelines of trying to refine and improve upon their technical accuracy and utility for industry and in-plant personnel. FSIS has been open to industry and academic input on ways the guidelines can offer flexibility and still ensure pathogen control. For instance, an industry expert on cooking has proposed the use of wet-bulb temperature as a suitable means to ensure adequate moist heating of meat product surfaces in lieu of the humidity measurements or minimum steam injection requirements (B Hanson, personal communication). What is most clear is that FSIS is willing to accept suitable scientific support for the safety of meat cooking and cooling processes in the form of challenge studies, predictive modeling, in-plant data, and other reliable sources. Therefore, the need to rely on safe harbors in appendices A and B, and the now more prescriptive qualifying criteria for using such safe harbors, should nudge industry to seek its own scientific validation of customized processes. For cooking, newer, more accurate thermal inactivation data for Salmonella, Shiga toxin-producing E. coli, and L. monocytogenes have been collected in roast beef, turkey, and ham.[18] Such data could form the scientific basis for a cook process instead of Appendix A, especially when used in conjunction with the process lethality spreadsheet.19 Similarly, a number of scientific studies demonstrate control of growth of C. perfringens as a result of product formulations. Matching the critical operating parameters of such scientific data to those of an industrial process is necessary to consider the process or product validated, but in many instances, that may be easier than meeting the current stipulations of the safe harbors in the compliance guidelines for appendices A and B.   

Peter J. Taormina, Ph.D., is president of Etna Consulting Group. Dr. Taormina earned his B.Sc. in biology from Valdosta State University and M.Sc. and Ph.D. from the Department of Food Science and Technology and Center for Food Safety at the University of Georgia.

References
1. Goodfellow, SJ and WL Brown. 1978. “Fate of Salmonella Inoculated into Beef for Cooking.” J Food Prot 41:598–605.
2. Bengtsson, NE, et al. 1976. “Cooking of Beef by Oven Roasting: A Study of Heat and Mass Transfer.” J Food Sci 41(5):1047–1053.
3. Doyle, ME and AS Mazzota. 2000. “Review of Studies on the Thermal Resistance of Salmonellae.” J Food Prot 63(6):779–795.
4. Taormina, PJ and JN Sofos. “Low-Water Activity Meat Products” in The Microbiological Safety of Low Water Activity Foods and Spices, edited by JB Gurtler, JL Kornacki, and MP Doyle (New York: Springer, 2014), 127–164.
5. U.S. Centers for Disease Control and Prevention. 1995. “Outbreak of Salmonellosis Associated with Beef Jerky—New Mexico, 1995.” MMWR 44(42):785.
6. Eidson, M, et al. 2000. “Beef Jerky Gastroenteritis Outbreaks.” J Environ Health 62(6):9–13.
7. Kalinowski, RM, et al. 2003. “Impact of Cooking, Cooling, and Subsequent Refrigeration on the Growth or Survival of Clostridium perfringens in Cooked Meat and Poultry Products.” J Food Prot 66(7):1227–1232.
8. Taormina, PJ, et al. 2003. “Incidence of Clostridium perfringens in Commercially Produced Cured Raw Meat Product Mixtures and Behavior in Cooked Products during Chilling and Refrigerated Storage.” J Food Prot 66(1):72–81.
9. Taormina, PJ and GW Bartholomew. 2005. “Validation of Bacon Processing Conditions to Verify Control of Clostridium perfringens and Staphylococcus aureus.” J Food Prot 68(9):1831–1839.
10. Golden, NJ, et al. 2009. “Risk Assessment for Clostridium perfringens in Ready-to-Eat and Partially Cooked Meat and Poultry Products.” J Food Prot 72(7):1376–1384.
11. askfsis.custhelp.com/app/answers/detail/a_id/2027/~/part-1-of-2%3A-recommendations-for-establishment-that-cannot-meet-the-2017.
12. askfsis.custhelp.com/app/answers/detail/a_id/2028/~/part-2-of-2%3A-recommendations-for-establishments-that-cannot-meet-the-2017.
13. askfsis.custhelp.com/app/answers/detail/a_id/2030/~/part-2-of-3%3A-revised-appendix-b%3A-stabilization-option-3-for-products-containing.
14. askfsis.custhelp.com/app/answers/detail/a_id/2031/~/part-3-of-3%3A-formulating-products-containing-natural-sources-of-nitrite-and.
15. Mohr, TB, et al. 2015. “Assessing the Performance of Clostridium perfringens Cooling Models for Cooked, Uncured Meat and Poultry Products.” 78(8):1512–1526.
16. www.fsis.usda.gov/wps/wcm/connect/a70bb780-e1ff-4a35-9a9a-3fb40c8fe584/HACCP_Systems_Validation.pdf?MOD=
AJPERES.
17. Zaika, LL. 2003. “Influence of NaCl Content and Cooling Rate on Outgrowth of Clostridium perfringens Spores in Cooked Ham and Beef.” J Food Prot 66(9):1599–1603.
18. McMinn, RP, et al. 2018. “Processed Meat Thermal Processing Food Safety — Generating D-Values for Salmonella, Listeria monocytogenes, and Escherichia coli.” Meat Muscle Biol 2(1):168.
19. meatpoultryfoundation.org/content/process-lethality-spreadsheet.