JOpinion

Editor’s note: Gary M. Weber, president of G.M. Weber Consulting LLC, submitted this column for original publication in Food Safety News.

Numerous E. coli O157:H7 outbreaks linked to Romaine lettuce have been a tragedy for both consumers and the leafy greens industry. The leafy greens industry is aggressively addressing approaches to prevent future outbreaks. This analysis is intended to share observations and recommendations for consideration that may contribute to the prevention of E. coli O157:H7 and other Shiga Toxin producing E. coli (STEC) contamination of leafy greens.

A discussion of previous efforts to reduce STEC contamination of beef, and resulting illnesses, can contribute valuable insights that relate to the detection and prevention of contamination of leafy greens. Several Food Safety Modernization Act (FSMA) rules will contribute to improving the safety of leafy greens including Standards for the Growing, Harvesting, Packing and Holding of Produce for Human Consumption (Produce Safety Rule, PSR), Current Good Manufacturing Practice and Risk-Based Preventive Controls for Human Food (Preventive Controls for Human Food), and Sanitary Transportation of Human and Animal Food.

The intent of this review is to contribute to the leafy greens related illness prevention dialog. I’m confident the industry made a concerted effort to prevent STEC related illnesses in the past. Outbreaks provide additional insights into future prevention options. The perspectives shared in this analysis, and recommendations, are solely the views and opinions of the author. 

My perspective on this issue is predicated upon the following experiences:

  1. Served as the Executive Director for Regulatory Affairs for the cattle industry (1994 -2006). I was initially hired to focus on E. coli O157:H7 related issues, including meat inspection reform and establishing a science-based hazard analysis and critical control points (HACCP) approach to beef safety. During this time period I was appointed to the Secretary’s Advisory Committee for Meat and Poultry Inspection1.
  2. President of the U.S. food safety division (2008-2012) of a global company that developed the first E. coli O157:H7 vaccine for cattle to reduce shedding into the environment.
  3. Served as the Prevention Manager for the U.S. Food and Drug Administration’s (FDA) Coordinated Outbreak Response and Evaluation (CORE) network (2013-2017). During that time, we conducted many leafy greens related E. coli O157:H7/STEC illness cluster analysis efforts, outbreak responses, and prevention discussions.
  4. Assisted in coordinating the Produce Safety Rule oriented On-Farm Readiness Review (OFRR) training programs under the leadership of the National Association of State Departments of Agriculture (NASDA), the FDA, and produce oriented professionals from several State Cooperative Extension Services2.
  5. Currently serving as a food safety consultant, focused on prevention efforts linked to achieving compliance with FDA Preventive Controls for Human and Animal Food, Foreign Supplier Verification, and the Produce Safety Rule. In addition, I have served as an expert on several foodborne illness litigation cases.

Based on these experiences, and analysis, the information and opinions included here are intended to stimulate a discussion regarding recommendations for consideration by leafy greens growers, processors, and the food service sectors. 

A farm to table risk reduction strategy should be considered as critically important to the prevention of E. coli O157:H7/STEC illnesses linked to leafy greens in general and Romaine in particular.

E. coli O157:H7 outbreaks, prevention efforts associated with beef

The history of beef E. coli O157:H7 prevention efforts provides insights into facets that should be considered when developing plans to address E. coli O157:H7 illness prevention related to produce. 

From November 15, 1992 to February 28, 1993 more than 500 laboratory-confirmed infections caused by E. coli O157:H7 linked to ground beef were identified. By the time the outbreak was over, more than 700 people had been infected. 

This incident would forever be a life changing event for many, and the force behind a policy, legislative, and regulatory sea-change in meat inspection and prevention of foodborne illnesses.

The E. coli O157:H7 outbreak resulted in the most significant change in meat inspection legislation since 19063. It was the catalyst for the establishment of a government and industry-wide focus on a Hazard Analysis and Critical Control Points (HACCP) approach to beef safety. 

The risk of E. coli O157:H7, and resulting human illnesses associated with beef, has decreased dramatically since 1993. In 1997 the Centers for Disease Control and Prevention (CDC) set a goal of less than 2.1 cases/100,000 by 2010. By 2005 the incidence rate had dropped by 42 percent to 0.9 cases/100,0004. Since 2005 incidence rates have climbed slightly, primarily due to increased leafy greens related outbreaks and more laboratory testing of patients. Rangel et al. 20054 reported that 41 percent of outbreaks were linked to beef and 21 percent to leafy vegetables. In 2018, CDC reported5 the incidence rate for E. coli O157:H7 was 0.72/100,000 and non-O157 at 0.97/100,000. Recent outbreaks associated with beef products indicate the risks continue to pose a prevention challenge.

The reasons for the decline in risk associated with beef products are multifaceted and include:

  1. USDA declaring E. coli O157:H7 an adulterant in ground beef.
  2. USDA requirements that industry establish a HACCP-based approach to reducing contamination.
  3. Industry establishment of multiple hurdles that each reduce the risk of
    E. coli O157:H7 contamination such as: antimicrobial rinses, hot water washing and steam applications.
  4. USDA development of more highly sensitive culture methods for detecting E. coli O157:H7.
  5. Changes in the Food and Drug Administration’s Model Food Code that recommend the food service sector cook ground beef to 155° F. Consumers were encouraged to cook ground beef to 160° F.
  6. Industry driven transportation and storage requirements ensuring that beef trim in transit to processors remain at or below 40° F to reduce the potential for growth of pathogens.
  7. Industry requirements for a test and hold program for ground beef and associated certificates of analysis (COA). Some industry sectors routinely verify the COA data. 
  8. Educational initiatives for food service and consumers to help ensure proper cooking.
  9. Continued research and development of improved sampling techniques such those published by Wheeler and Arthur, 20186.

Considerations for a farm-to-table approach to

Reducing the Risk of E. coli O157:H7/STEC Associated with Leafy Greens
The leafy greens sector has invested a great deal of resources in the development and implementation of initiatives such as the California Leafy Greens Marketing Agreement (LGMA) and the Arizona LGMA. I believe these efforts have utilized the most up-to-date, and science-based information. However, information gleaned from recent outbreaks has resulted in additional efforts to enhance these initiatives. 

In addition, produce marketing associations, Cooperative Extension Service professionals, and private sector consultants have been contributing to other efforts to reduce the risk of pathogens associated with leafy greens.

There are also many audit schemes ranging from the USDA GAPs to third party audits that are all intended to contribute to the production of safe leafy greens, and other produce.

With the passage of the Food Safety Modernization Act, and development of the Produce Safety Rule (PSR), States and the FDA are beginning to inspect leafy greens producers to ensure compliance with the PSR.

The National Association of State Departments of Agriculture (NASDA) has been coordinating, in concert with the FDA, the On-Farm Readiness Review (OFRR) program2. This effort brings a coordinated, standardized approach to training of State regulatory, Cooperative Extension and produce growers. This initiative’s focus contributes to “Educating Before You Regulate” to help ensure compliance with the PSR.

As States and the FDA begin PSR compliance inspections they will operate under a mindset of “Educate Before and While You Regulate”. This approach will increase the likelihood that produce growers learn about PSR compliance requirements while they are being inspected. This is intended to contribute to both higher levels of producer compliance as well as reductions in specific food safety risks. 

With all these activities and initiatives underway a natural question is why do we continue to have so many E. coli O157:H7/STEC outbreaks associated with leafy greens in general, and Romaine in particular?

Potential weak links in farm-to-table produce safety

1. Analysis of Individual Farm Production Risks
I have observed that many food production firms, in general, often look to others to set their food safety benchmarks. These benchmarks might be information found in the various FDA hazard guides, or in the case of produce, LGMA oriented information, USDA GAP, and third-party audit scheme checklists.

It is important to keep in mind, based upon analysis of numerous outbreaks, that adhering to these “hazards guides” and audit schemes alone may not fully protect growers and processors from being accused of causing a problem and potentially ending up in court. 

These “hazard guides” should only be viewed as a starting point to develop an individual firm or farm hazard analysis and food safety plan. A comprehensive farm and processor specific hazard analysis are essential to help ensure the production of safe food. 

2. Developing and implementing a farm specific produce safety plan is a significant investment.
Farm profitability can be a challenge and effective investments in food safety might be compromised as a result of compliance costs associated with the PSR and simultaneous third-party audit schemes. If these costs are not providing a return on investment in terms of real reductions in food safety risk, they may indirectly be contributing to foodborne illness outbreaks. It is important investments in food safety practices and technology are highly correlated with achieving food safety objectives.

For example, in 2018, the Economic Research Services7 estimated the PSR compliance costs would range from just over $1,700 for small farms to over $37,000 for very large farms. These costs appear reasonable although the PSR compliance requirements must be viewed as a foundation level of food safety risk reduction approaches. 

In 2017, The Economic Research Service8 surveyed produce growers to determine the costs related to food safety under their food safety plans, including costs of third-party or other audit schemes. Growers reported on costs for food safety staff, foremen food safety time, audits, lost product due to animal intrusion, and water testing.

The survey found that between 27 and 38 percent of all food safety related costs were linked to third-party audit schemes. These costs ranged from a low of $27,150 to a high of $305,403 per farm annually.

The survey also found total food safety plan costs on farms (food safety staff, foremen food safety time, audits, lost product due to animal intrusion, and water testing) ranged from $120,501 to $1,297,063. If you were paying that kind of money and still ended up with a food safety problem, it’s time to consider other options. 

Recommendation
The produce industry and its customers should consider the cost and benefits of multiple third-party audit schemes.  These schemes understandably go well beyond the requirements of the FDA PSR, some of the required processes may represent potentially costly expenditures as reported in the ERS survey. The concern is if all these costs, food safety audits and related expenses are not focused on a farm’s real produce safety risks, the processor, retailer, and public remain at risk. 

Specific, individualized farm/firm hazard analysis, and resulting food safety plans should provide the framework that is evaluated against risk-based food safety benchmarks, not checklist-based auditing schemes. The expense of multiple, generic, far-ranging audit schemes may be contributing to a lack of resources focused on the real food safety risks. Some producers may believe if they are complying with the PSR, audit schemes, or marketing agreement requirements their produce will be safe. History is telling us something very different. 

Sources of E. coli O157:H7/STEC and potential transfer routes to produce

Numerous studies have documented that the primary environmental reservoir of E. coli O157/STEC is cattle. Cattle shed E. coli O157:H7/STEC into the environment at varying rates from individual animals and times of the year. 

Reducing the risks presented by cattle operations have focused on establishing setbacks from cattle (beef and dairy) operations and water testing. As outbreaks continue, these setbacks and water analysis methods are under review. 

Proximity to cattle operations must be viewed as an E. coli O157:H7/STEC hazard. Routes of pathogen movement to produce should be analyzed. If no route of transfer exists, there is no hazard reasonably likely to occur.

Cattle operations are under numerous regulatory requirements, not the least of which is that the EPA does not allow any discharge of contaminated water into the surface waters of the United States. Thus, water may not represent the most likely route of transfer of the pathogen to produce. The EPA does not regulate fugitive dust from cattle operations, although it has been evaluated9.

The FDA Yuma environmental assessment10 (FDA EA) relating to E. coli O157:H7 contamination of Romaine found the three illness-causing strains in the Wellton canal. One was to the west of a feedlot, one adjacent, and another to the northeast. The stretch of the Wellton canal where the positive samples were taken was approximately 3.5 miles in length. 

To what extent these positive samples indicate water was the source of the hazard, and water use the route of transfer or exposure, should be carefully evaluated. Is Wellton canal water the most likely route of transfer of the hazard?

Water as a risk factor in a hazard analysis
The FDA EA did not mention the flow rate of water in the Wellton canal at the time of harvest associated with the outbreak, nor during their sampling in June 2018. An analysis of the flow rate in the Wellton canal and potential location of farms that may be implicated in shipping of contaminated produce, needs to be considered.

Currently we do not know where farms suspected of producing contaminated Romaine are relative to being “up-stream or down-stream” from the feedlot mentioned in the FDA EA and where the positive E. coli O157:H7 Wellton canal water samples were taken.  

From a hazard analysis perspective, consideration of the rate of water flow passing Yuma and ultimately the cattle feedlot, from the west to the east, in February and March was as follows11: 

Water Flow Unit February 2018 March 2018
Acre Feet/Day Average 830.7 1,139.6
Cubic Feet/Second Average 419 575
Gallons Per Day 270,667,908 371,339,800
Gallons Per Second 3,134 4,298

The actual water flow rate passing the feedlot would be lower than these rates due to flow into laterals in the system. However, flow rates passing the feedlot would still be significant. Given these water flow rates it is difficult to imagine how a contaminant entering the canal, near a feedlot, would remain in proximity to the point of initial contamination.  In addition, the water flow rates in the canal, raise the issue of how would you take a representative sample of water that is flowing at a rate of 3,000 gallons per second?

Laboratory Methods
Another issue worth discussing is that there isn’t a perfect correlation between generic

E. coli measurements and E. coli O157:H7. This issue is discussed in several publications including by Cooley et al. 200712 relating to the 2006 contaminated Spinach event. 

Application of water to the harvestable portion of a crop is an issue. Consequently, water tests associated in “real time” with the water being applied are important. 

If the hazard being monitored is E. coli O157:H7/STEC, laboratory methods should be used that have the proper sensitivity and specificity to detect the pathogen. In

addition, sampling methods need to represent a statistically valid number of samples. 

The FDA EA doesn’t include information regarding methods the Centers for Disease Control and Prevention (CDC) used to identify the three illness-causing E. coli O157:H7 “strains” from Wellton canal water samples. That information would he helpful in an analysis of testing methods.

One of the most significant steps USDA took to deal with the threat of E. coli O157:H7 in beef was the development of very sensitive tests used for regulatory purposes. The current USDA E. coli STEC test description can be found in the February 2019 updated USDA Laboratory Handbook, MLG 5C.0013 method. It employs “enrichment in a selective broth medium, application of a rapid screening test, immunomagnetic separation (IMS) in paramagnetic columns, and plating on a highly selective medium, modified Rainbow Agar (mRBA)”.

Recommendation
If produce growers believe there is a risk of contamination of produce due to E. coli O157:H7/STEC contamination of water applied to the harvestable portion of the crop, careful consideration of the laboratory methods used to test water should be considered. This includes the number of samples that would be necessary to provide a statistically valid sample. The time period for testing should be aligned with application of the water to the harvestable portion of the crop. Testing for generic E. coli strains may not provide information necessary to make a truly informed decision regarding water source and use patterns relating to a potential “hazard reasonably likely to occur”.

Weather events as a factor in hazard analysis
During my tenure with FDA CORE it became clear that weather events played a significant role by contributing to the movement of pathogens (route) from their source (hazard) on to or into food. A hazard coupled with a route of movement equates to a hazard reasonably likely to occur. 

For example, it is suspected that weather events in Tuna fisheries may contribute to sewage entering the ocean and resulting contamination of Tuna with Salmonella and other pathogens. The exact route of contamination is difficult to determine but weather events provided a way for pathogens in human waste to enter the ocean during a storm event. It could be the fish themselves became contaminated internally or externally due to being in contaminated water. The 2015 hurricane season in the Pacific was the second most active on record14. Several storm events impacted the Baja Peninsula. This may have led to Salmonella Poona contamination of cucumbers grown on the peninsula.

Romaine growers in the Yuma Arizona area reportedly told the FDA EA team10 a significant weather event occurred in February 2018. It included high winds and below freezing temperatures. It is important to note that with storm events, the wind can shift 360 degrees as a low-pressure weather event occurs. In the Yuma area, could the weather event winds have shifted to the southeast, south, or southwest moving dust from a feedlot over fields, depositing it on produce, resulting in microbiological contamination?

This naturally raises the following questions for consideration in a hazard analysis:

  1. Is there a source of dust associated with cattle in the vicinity of produce fields (a hazard)?
  2. How much dust do cattle produce?
  3. What is the size of the dust particles?
  4. Can dust carry pathogens?
  5. How far can dust from cattle operations travel during a wind event (route of hazard movement)?

Some of the answers to these and other questions can be found in published literature and information from the Environmental Protection Agency (EPA).

For example, in 2015 the EPA published a summary document titled “Fugitive Dust from Beef Cattle Operations9. In this document, the EPA estimated total dust emissions from cattle feedlot operations of 17 tons/1,000 head/year. The smaller particles of concern are classified as PM2.5 and PM10. It is estimated that 15% of the total emissions would be PM2.5 – PM10. The PM designation relates to the size in microns. This equates to an estimate of 2.5 tons of PM2.5 -PM10 per 1,000 head/year. 

A study conducted by McEachran et al. in 201515 used an estimate of 28.5 grams of PM10/head/day from feedlots, which equates to 62.8 lbs./1000 head/day or 11.5 tons/year/1000 head. This would equate to 6,278 lbs. of PM10 and smaller particles per day/100,000 head of feedlot cattle. 

In a study by Hiranuma et al. (2011)16 found measurable amounts of PM material traveled as far as 2.2 miles from a feedlot location. This study did not establish the maximum distance measurable PM10 or smaller particles can travel. 

The question of how far PM25 and smaller particles of dust can travel is perhaps best illustrated by the movement of PM type dust from the Sahara. PM25 and smaller dust particles from the Sahara routinely travel across an expanse of 3,000 miles of the Atlantic Ocean and are deposited on the Amazon Basin. Swap et al. 199217 estimated that during individual storm events as much as 480,000 tons of the PM material originating from the Sahara are deposited on the Amazon Basin.

Questions have been raised regarding if microbes survive transport on dust due to desiccation or UV radiation. Dust has been documented to be capable of carrying viable E. coli O157:H7. Berry et al. 201518 studied the movement of dust from a feedlot and reported that over a two-year time period, E. coli O157:H7 and total E. coli were recovered at all produce plot distances, up to 590 feet. This study did not identify the upper limit of dust movement harboring viable pathogens. They noted the risk of movement is increased when pen surfaces are dry. Nighttime movement of contaminated dust would reduce the potential for decontamination due to UV light prior to deposition. 

The FDA conducted an EA after a 2010 outbreak associated with E. coli O145 in shredded lettuce grown in the Yuma area19. Carter et al. 201620 reported on the studies of E. coli O145 found in a major produce region in California. Some of the strains have greater virulence than others. A review of the prevalence of STEC in dairy cattle by Hussein and Sakuma 200520 clearly indicate the hazards presented by STEC are not limited to beef cattle.  

Recommendation
From a hazard analysis perspective, a beef feedlot or dairy operations in proximity to produce fields, represents an STEC hazard. Weather and associated wind events can carry dust harboring pathogens significant distances. From a hazards analysis perspective there is a source and a route of transmission (wind) resulting in STEC’s originating from a feedlot or dairy as a hazard reasonably likely to occur.  Consequently, steps must be taken to understand the risk and control this hazard, including weather event related produce safety actions. 

Microbial testing considerations as a risk identification strategy

Let’s assume you have determined what hazards are reasonably likely to occur during the production, processing, distribution or use of a food item at retail. From this analysis you have developed a food safety plan including measures to reduce the risks presented by the hazards. How you monitor the risks reduction steps is important.

Microbial testing has always been a sensitive topic. I have been in many meetings where individuals have discouraged firms from testing. The issue has been concerns regarding what such testing might find or what testing implies about risks. These concerns increased with the establishment of the Reportable Food Registry22 requirements. I’m not an attorney but it seems logical that some microbial testing and associated efforts to verify you are at least attempting to reduce risk would be viewed as industry expected due diligence.

Most firms purchasing food items, either raw or finished ingredients, require a Certificate of Analysis (COA) including several testing requirements. These requirements often include microbiological, mycotoxins, heavy metals and pesticide residue tests. These tests may be required as a function of Hazard Analysis and Risk Based Preventive Controls (HARPC). These plans usually include verification of the COA information. 

On September 27, 2007 Congress approved the Food and Drug Administration Amendments Act of 200722. Section 1005 of that legislation required FDA to establish a “Reportable Food Registry”. The RFR provides an electronic portal for industry to report when there is reasonable probability that an article of food will cause serious adverse health consequences. The RFR is intended to help the FDA better protect the public health by tracking patterns and targeting inspections.

The RFR, in general, makes sense, however in the real-world, if a firm were to report to the RFR a low level of Listeria monocytogenes (LM) in say ice cream, what might FDA do? Recent history illustrates they would determine any level of LM in a ready to eat product represents a “reasonable probability that an article of food will cause serious adverse health consequences”. 

FDA has essentially set a zero tolerance for LM in ready to eat foods, even though the Codex standard remains at 100 CFU/gram as a safe level of the pathogen, arguably, only for not at-risk populations. An LM outbreak associated with ice cream caused the death of at least 5 patients in 2016. Analysis of ice cream linked to the outbreak found low levels of LM in samples. The evaluation showed 92% of samples 

were contaminated at less than 20 MPN/g23. Prior to this outbreak, ice cream was not viewed as a high risk for LM contamination since it would not support growth when stored at cold temperatures. Each outbreak provides an opportunity to better understand food safety hazards. The FDA published updated LM guidance for industry shortly after this event.

The RFR created a situation where testing a ready to eat food item and finding any level of a pathogen could present a “reasonable probability that an article of food will cause serious adverse health consequences.” This reality added additional resistance to conducting culture-based microbiological analysis of food items. The uncertainty regarding how FDA would interpret the information considering the “reasonable probability that an article of food will cause serious adverse health consequences” added to these challenges.

Consequently, the RFR provided an incentive for the development of microbial testing systems that were based on DNA/PCR analytical platforms. There are several companies offering DNA/PCR based testing platforms. In general, these platforms provide a report indicating the relative likelihood a food item has a level of a pathogen presenting a “reasonable probability that an article of food will cause serious adverse health consequences.”

These tests technically reduce the pressure to report findings to the RFR because they are not actually detecting viable cells of a pathogen, rather only their DNA. The challenge for firms is to fully understand the DNA/PCR technology and what the resulting numbers mean in terms of risk to public health. 

From my experience, most leafy greens produced in the U.S. are subject to COA requirements. The most common analytical methods used to meet the COA requirements are DNA/PCR based tests. 

In the case of E. coli O157:H7, DNA/PCR tests are likely not as sensitive as the testing methods used, for example by USDA, that employ “enrichment in a selective broth medium, application of a rapid screening test, immunomagnetic separation (IMS) in paramagnetic columns, and plating on a highly selective medium, modified Rainbow Agar (mRBA)13”.  Thus, it is likely these methods will not provide a true representation of risk posed by a sample of leafy greens. 

Recommendation
Microbiological testing of leafy greens and other food items, in the face of an E. coli O157:H7/STEC hazard determined to be reasonably likely to occur, must use methods sensitive enough to detect the pathogen. There is virtually no safe level of STEC’s in ready to eat foods. 

Impact of temperature, time on pathogen growth in processed leafy greens

As foodborne illness clusters are identified, often the first cases provide clues to the geographic origin of the pathogen. For instance, in the Jensen Farms of Colorado Cantaloupe LM outbreak, the first cases identified were in Colorado. Eventually cases were identified across the United States. 

When the Romaine outbreak of 2018 occurred, the first cases were in the Northeast. This led to some preliminary concern the Romaine may originate in a greenhouse system in that region. Of course, the outbreak ultimately resulted in cases in 36 States. This illness outbreak pattern in the Northeast was very similar to the 2010 E. coli O145 outbreak linked to chopped leafy greens originating from the Yuma area, as reported by the CDC21. In the 2018 outbreak it was quickly determined the source of the outbreak was Romaine from the Yuma Arizona area.

This led to asking the question, could processing of Romaine and subsequent transportation and storage from distribution centers to retail locations be contributing to risk?

It is a common practice to monitor the temperature of leafy greens from harvest through transportation to processing. However, it is less clear what temperatures processed/chopped leafy greens experience over their remaining shelf life, up to14 days after initial harvesting. The FDA is in the process of enforcing the final rule on Sanitary Transportation of Human and Animal Food24. This will generate incentives to control the temperature of leafy greens during transport. Risks may remain if temperatures at retail or at home are not controlled. 

A study by Luo et al.26 2010, demonstrated that E. coli O157:H7 inoculated, chopped Romaine and Iceberg lettuce, stored in N2 flushed bags, showed no additional growth at 5° C, when sampled periodically until their “Best If Used By” dates. Conversely, the chopped Romaine and Iceberg lettuce stored at 12° C were found to have more than a 2 log CFU/g increase of the pathogen within 3 days. These research results are consistent with work published by Zeng et al. 201427. 

Coleman et al. 201328 studied handling practices of fresh leafy greens in restaurants, including product receiving requirements. They reported 49 percent of the shipments were received at temperatures over 41° F, 24.3 percent were received at between 42°- 45° F, 18.9 percent were received at 46°-54° F, and 8.1 percent were received at greater than 55° F.

This raises the question, could a low level of E. coli O157:H7/STEC, one low enough to not cause illness itself, multiply post processing to a level that could cause significant illnesses? The answer is yes, contamination of ready to eat foods with STEC or other pathogens can increase over time if the products are exposed to temperatures and time intervals that support growth. Risks can increase from the farm to the table as even a very low number of pathogens can   increase by several logs over time, increasing the risk to public health.

Recommendation
The risk of illness caused by E. coli O157 and other STEC can increase from the farm to the table under some circumstances. Even if sophisticated testing indicates an absence of these pathogens on leafy greens at harvest, temperature of the product needs to be reduced as quickly as possible and maintained at a temperature of at least 40° F or less throughout its shelf life to control risk. 

Summary: recommendations for leafy greens, produce sectors to reduce risk of E. coli O157/STEC contamination and human illness
  1. The concept of a multiple hurdle approach to reducing the food safety risk of leafy greens is not a new idea. Recently, Mogren et al. 201829 published a review of potential hurdles to microbial contamination and potential reductions of risk associated with leafy greens. Analysis of recent outbreaks has contributed to identifying the additional risk factors in the current hazard analysis for leafy greens production, processing and distribution systems.Farms, processors, distributors, and the retailers need to conduct a hazard analysis and fully understand that achieving food safety goals requires steps along the entire path from the farm to the table. Although not required by the PSR, it is highly recommended individualized food safety plans should be the result of this effort.
  2. Development of individualized farm/firm specific hazard analysis-based produce safety plans is vitally important. Reliance on FDA guidance, audit schemes or marketing agreement requirements alone, has not, and will not comprehensibly ensure the safety of produce.
  3. It is critically important produce growers understand the E. coli O157/STEC hazards presented by beef and dairy cattle operations, particularly in dry environments. Evidence suggests wind blown dust can carry pathogen laden particles, in the PM10 range for miles, depositing them on crops. The upper limit of their potential movement, and viability of pathogens being transported, has not been established. Risk of these particles moving should be based upon published atmospheric science research relating to the movement of PM10. This risk has not been comprehensively addressed in existing production guidance or other peer reviewed publications.
  4. Weather events and post event response plans need to be included and defined within individual farms/firms hazard identification and control plans.
  5. Microbial testing of water that contacts the harvestable portion of produce must be more robust if the risk of E. coli O157:H7/STEC is identified. Water testing should occur prior to application to the harvestable portion of the crop. Generic E. coli testing is not highly correlated with the presence E. coli O157:H7/STEC. Statistically valid sampling of water sources should account for source flow rates.
  6. If E. coli O157:H7/STEC is considered a hazard reasonably likely to occur, microbiological testing of produce contributing to a COA, including as a field release criterion, must use highly specific and sensitive methods. Methods such as the FSIS MLG 5C.009 should be considered when evaluating appropriate methods.
  7. Temperature of produce, especially during and post processing, is a factor in controlling the risk of E. coli O157:H7/STEC. Virtually undetectable levels of E. coli O157:H7/STEC on chopped produce can rapidly multiply to infectious levels when temperatures rise above 40° F where rapid growth is supported. From the farm to the table, the temperature of produce, especially chopped produce, must remain at or below 40° F. Firms processing produce are subject to the temperature control provisions of the Preventive Controls for Human Food rule. The FDA Sanitary Transport rule requires temperature controls as well. 
  1. When receiving produce, it is important that storage and shipping temperatures are verified to have been at or below 40° F. Shipments should be rejected if they have been subject to temperature abuse.  

references
  1. Department of Agriculture Food Safety and Inspection Service [Docket No. 99–044N]

National Advisory Committee on Meat and Poultry Inspection. 1999. Federal Register 55225 Vol. 64, No. 196. October 12, 1999. https://www.fsis.usda.gov/wps/wcm/connect/4bac9a96-39bb-43a1-870a-500a8850c1b7/99-044N_226.pdf?MOD=AJPERES

  1. On-Farm Readiness Reviews https://www.nasda.org/foundation/food-safety-cooperative-agreements/on-farm-readiness-review
  1. FSIS Timeline of Events Related to E. coli O157:H7. 2005. https://www.fsis.usda.gov/wps/portal/fsis/topics/regulatory-compliance/haccp/updates-and-memos/timeline-of-events-related-to-e-coli-o157h7/e-coli-timeline
  2. Rangel, J.M., Sparling, P.H., Crowe, C., Griffen, P.M. Epidemiology of Escherichia coli O157:H7 Outbreaks, United States, 1982–2002. https://www.researchgate.net/publication/7908247_Epidemiology_of_Escherichia_coli_O157H7_Outbreaks_United_States_1982-2002
  3. National Enteric Disease Surveillance: Shiga Toxin-producing Escherichia coli (STEC) Annual Report, 2016. Data current as of April 16, 2018. https://www.cdc.gov/ecoli/surv2016/index.html
  4. Wheeler, T and Arthur, T. 2018. Novel Continuous and Manual Sampling Methods for Beef Trim. Journal of Food Protection, Vol. 81, No. 10:1605–1613. 2018
  5. Bovay, J., Peyton Ferrier, P., Chen Zhen, C., 2018. Estimated Costs for Fruit and Vegetable Producers to Comply with the Food Safety Modernization Act’s Produce Rule.  USDA Economics Research Service. Economic. Information Bulletin Number 195. https://www.ers.usda.gov/publications/pub-details/?pubid=89748
  1. Calvin, L., Helen Jensen, H., Karen Klonsky, K., Cook, R. 2017. Food Safety Practices and Costs Under the California Leafy Greens Marketing Agreement. Economic Research Service. Economic Information Bulletin Number 173. https://www.ers.usda.gov/publications/pub-details/?pubid=83770
  2. EPA: Fugitive Dust from Beef Cattle: https://www.epa.gov/sites/production/files/2015-08/documents/feedlots.pdf
  1. FDA Yuma Environmental Assessment 2018.
    https://www.fda.gov/food/outbreaks-foodborne-illness/environmental-assessment-factors-potentially-contributing-contamination-romaine-lettuce-implicated
  1. Wellton Canal Water Flow Rate Data, February, March 2018. Pages 4-5. https://az.water.usgs.gov/projects/LowerColoradoRiverReports/reports/2018/2018_02.pdf
    https://az.water.usgs.gov/projects/LowerColoradoRiverReports/reports/2018/2018_03.pdf
  1. Cooley, M., Carychao, D., Crawford-Miksza, L., Jay MT, Meyers, C., Rose, C., Keys, C., Farrer, J., Mandrel, R.E., 2007. Incidence and Tracking of Escherichia coli O157:H7 in a Major Produce Production Region in California. PLOS ONE November 2007 Issue 11, e.1159. https://www.ncbi.nlm.nih.gov/pubmed/18174909
  2. USDA Laboratory Handbook MLG 5C.00 https://www.fsis.usda.gov/wps/wcm/connect/7ffc02b5-3d33-4a79-b50c-81f208893204/MLG-5B.pdf?MOD=AJPERES
  1. National Hurricane Center Annual Summary 2015 Eastern North Pacific Hurricane Season. https://www.nhc.noaa.gov/data/tcr/summary_epac_2015.pdf
  2. McEachran, A.D., Blackwell, B.R., Hanson, J.D., Wooten, K.J., Mayer, G.D., Cox, S.B., Smith, P.N. 2015. Antibiotics, Bacteria and Antibiotic Resistance Genes: Aerial Transport from Cattle Feed Yards vis Particulate Matter. Environmental Health Perspectives Volume 123: Number 4: 337-343. https://dx.doi.org/10.1289/ehp.1408555
  3. Hiranuma, N, Brooks, S.D. Gramann, J., Auvermann, B.W. 2011. High Concentrations of Coarse Particles Emitted from a Cattle Feeding Operation. Atmos. Chem. Phys., 11: 8809–8823. https://www.atmos-chem-phys.net/11/8809/2011/ doi:10.5194/acp-11-8809-2011. https://www.atmos-chem-phys.net/11/8809/2011/
  4. Swap, R., Garstang, M., Greco, S., Talbot R., Kallberg, P., 1992. Saharan Dust in the Amazon Basin. TellusB: Chemical and Physical Meteorology, Volume 44:2:133-149. https://www.tandfonline.com/doi/abs/10.3402/tellusb.v44i2.15434
  5. Berry, E. D. Berry, Wells, J.E., Bono, J.L., Woodbury, B.L., Kalchayanand, N., Norman, K.N., Suslow, T.V., López-Velasco, G., Millner, P.D. 2015. Effect of Proximity to a Cattle Feedlot on Escherichia coli O157:H7 Contamination of Leafy Greens and Evaluation of the Potential for Airborne Transmission. Applied and Environmental Microbiology February 2015. Volume: 81, Number: 3:1101-1110. https://www.ncbi.nlm.nih.gov/pubmed/25452286
  6. Environmental Assessment Report FDA Foods Program Non-O157 Shiga Toxin-Producing E. coli (STEC) Environmental Assessment. https://wayback.archive-it.org/7993/20171114155100/https://www.fda.gov/Food/RecallsOutbreaksEmergencies/Outbreaks/ucm235477.htm
  1. Carter, M.Q., Quinones, B., He, X., Zhong, W., Louie, J.W., Lee, B.G., Yambao, Y.C. Mandrell, R.E., Cooley, M.B. 2016. An Environmental Shiga Toxin-Producing Escherichia coli O145 Clonal Population Exhibits High Level Phenotypic Variation That Includes Virulence Traits. Applied and Environmental Microbiology, February 2016. Volume 82. Number 4:1090-1101. https://aem.asm.org/content/82/4/1090
  2. Hussein, H.S. and Sakuma, T. 2005. Invited Review: Prevalence of Shiga Toxin-Producing Escherichia coli in Dairy Cattle and Their Products. J. Dairy Sci. Volume 88:450-465. https://www.journalofdairyscience.org/article/S0022-0302(05)72706-5/pdf
  3. Food and Drug Administration Amendments Act of 2007.
    https://www.govinfo.gov/content/pkg/PLAW-110publ85/pdf/PLAW-110publ85.pdf
  1. Chen,Y., Burall, L.S., Macarisin, D., Pouillot, R. Strain, E., J. De Jesus, A.J., Laasria, A., Wang, A., Ali, L., Aptatavarthy, A., Zhang, G., Hu, L., Day, J., Kang, J.,  Sahu, S., Srinivasan, D., Klontz, K., Parish, M., Evans,P.S., Brown, E.W., Hammack,T.S., Zink, D.L., DATTA, A.R., 2016. Prevalence and Level of Listeria monocytogenes in Ice Cream Linked to a Listeriosis Outbreak in the United States. Journal of Food Protection, Vol. 79, No. 11, 2016, Pages 1828–1832. https://jfoodprotection.org/doi/pdf/10.4315/0362-028X.JFP-16-208
  2. Multistate Outbreak of Human E. coli O145 Infections Linked to Shredded Romaine Lettuce from a Single Processing Facility (FINAL UPDATE) https://www.cdc.gov/ecoli/2010/shredded-romaine-5-21-10.html
  3. FDA Final Rule on Sanitary Transport of Human and Animal Food. 2016.
    https://www.fda.gov/food/food-safety-modernization-act-fsma/fsma-final-rule-sanitary-transportation-human-and-animal-food
    https://www.fda.gov/regulatory-information/search-fda-guidance-documents/guidance-industry-sanitary-transportation-human-and-animal-food-what-you-need-know-about-fda
  4. Luo, Y., He, Q., McEvoy, J.L., 2010. Effect of Storage Temperature and Duration on the Behavior on Escherichia coli O157:H7 on Packaged Fresh-Cut Salad containing Romaine and Iceberg Lettuce. J. Food Sci. 2010 Sep; 75(7):390-397. https://www.ncbi.nlm.nih.gov/pubmed/21535546
  5. Zeng, W., Vorst K., Brown W., Marks B.P., Jeong S., Pérez-Rodríguez, F., Ryser E.T. 2014. Growth of Escherichia coli O157:H7 and Listeria monocytogenes in Packaged Fresh-Cut Romaine Mix at Fluctuating Temperatures During Commercial Transport, Retail Storage, and Display. J Food Prot. 2014 Feb;77(2):197-206. https://www.ncbi.nlm.nih.gov/pubmed/24490913
  6. Coleman, E., Delea, K., Everstine, K., Reiman, D., Ripley D. 2013. Handling Practices of Fresh Leafy Greens in Restaurants: Receiving and Training. Journal of Food Protection, Vol. 76, No. 12, 2013, Pages 2126–2131. https://www.cdc.gov/nceh/ehs/ehsnet/docs/jfp-handling-leafy-vegs.pdf
  7. Mogren, L., Windstam, S., Boqvist, S, Vågsholm, I, Söderqvist, K., Rosberg, A.K., Lindén, J., Mulaosmanovic, E., Karlsson, M., Uhlig, E., Håkansson. Å., Alsanius, B. 2018. The Hurdle Approach–A Holistic Concept for Controlling Food Safety Risks Associated with Pathogenic Bacterial Contamination of Leafy Green Vegetables. A Review. Front Microbiol. 2018 Aug 24; 9:1965. doi: 10.3389/fmicb.2018.01965. https://www.ncbi.nlm.nih.gov/pubmed/30197634

About the author: Gary M. Weber, PhD, is with G.M. Weber Consulting. Weber has more than 30 years experience working on food safety and agricultural issues for the USDA, FDA, the private sector, and trade associations. He has extensive experience working with the media and has testified under oath numerous times before Congress. He serves as a consultant and expert witness. He served over four years as Prevention Manager for the FDA Coordinated Outbreak Response and Evaluation Network (CORE) where he handled more than 100 significant food-borne outbreaks.

(To sign up for a free subscription to Food Safety News, click here.)