Between 2009 and 2017, FDA and their partners at the Centers for Disease Control and Prevention (CDC) identified 28 foodborne illness outbreaks of Shiga-toxin producing E. coli (STEC) in the United States with a confirmed or suspected link to leafy greens—an average of more than three outbreaks per year.
This eight-year period came on the heels of the 2006 E. coli O157:H7 spinach outbreak followed by the industry’s development and implementation of the initial leafy greens marketing agreement. Then in 2018, as the industry prepared for Food Safety Modernization Act (FSMA) compliance and inspections, additional E. coli O157:H7 outbreaks occurred that were linked by the FDA’s traceback investigations to Arizona and California’s desert growing region and California’s Central Coast. In each of these outbreaks, FDA deployed investigators to the regions and found the outbreak strains in the leafy green production environment. During the FDA’s environmental assessment conducted in response to the early 2018 outbreak associated with desert-grown romaine, three samples of irrigation canal water analyzed by whole genome sequencing were found to contain an E. coli O157:H7 strain with the same rare molecular fingerprint as the strain that produced human illnesses (the outbreak strain). These samples were collected from an approximate 3.5-mile stretch of an irrigation canal in the Wellton area of Yuma County that delivers water to several of the farms identified in the traceback investigation as potential shippers of contaminated product. Later in the year, another foodborne illness outbreak sent FDA investigators to California’s Central Coast where an irrigation reservoir was found to contain E. coli O157:H7 with a molecular match to the outbreak strain.
What links these two outbreaks, other than they are both associated with romaine, is that in each instance FDA found the same strain as the outbreak pathogen in water and sediment in the area of investigation. It cannot be said with certainty that the water contaminated leafy green crops. And we do not know how the water and/or sediment itself may have become contaminated. But these findings, coupled with a high potential for water of unknown quality to transfer pathogens if applied to a crop, have prompted both the FDA and industry to re-examine the preventive controls associated with water used in fresh produce operations.
It is well understood that water, if it contains fecal material and contacts edible portions of the crop, may contaminate leafy greens during production and harvest operations. Contamination may also occur by means of water-to-soil followed by soil-to-leafy greens contact. In its November 2018 letter to industry, FDA requested that we assess the need for updating and developing additional commodity-specific procedures, policies, and best practices to enhance leafy green safety. FDA specifically highlighted the importance of assuring that the agricultural water (ag water is defined as water that directly contacts the harvestable portion of the crop) used in production and harvest is “safe and adequate for its intended use.”
After assessing the need, industry groups have determined that change is necessary. United Fresh and the Produce Marketing Association have organized a task force to further review and assess issues associated with recent outbreaks while actively working to develop resources, which might manifest as white papers, guidance, recommendations, or other deliverables in several areas including preventive practices, labeling, traceability, investigation, and outbreak response. The task force includes many fresh produce supply chain members and is progressing as rapidly as their large structure allows them to deliver results.
Western Growers, recognizing that the Leafy Greens Marketing Agreements in Arizona and California are the quickest path to any implementation of new preventive controls, has been facilitating a process to develop new food safety practices and standards for adoption and implementation by industry. The facilitative process includes WG working directly with a small drafting committee to develop initial language and then asking broad industry for feedback on the drafts. As industry provides commentary and suggestions, WG works to incorporate that feedback and take it back to the drafting committee for its input. After cycling through this iterative process often enough to feel comfortable with a stable final draft, Western Growers will then hold an industry-wide webinar to explain the draft and take further input. Upon conclusion of those steps, Western Growers will take the proposed revisions to the Arizona and California LGMA Technical Committees that, in turn, will review the changes and, when comfortable, recommend adoption to their respective boards. If adopted by LGMA boards, the work will become the controlling language for all leafy green handlers engaged in the Arizona and California LGMA (the vast majority of the fresh romaine supply).
Because both FDA and industry have prioritized water as an issue that needs to be addressed, the current effort is to develop a new paradigm for water quality relating to food safety. In food safety programs commonly deployed by industry, water that is “safe and adequate for intended use” means that it meets or exceeds the standards for recreational water as measured by testing for generic E. coli and monitoring for shifts in baseline levels. Recent research and outbreaks have taught us there are inadequacies in the current ag water quality management strategies and that change is necessary. First, generic E. coli is inconsistent, and therefore ineffective, when used as the sole indicator for the presence of pathogens, such as E. coli O157:H7, in ag water sources. In addition, not all ag water from all sources are of equivalent food safety risk—a municipal source or a deep, protected well do not present the same risk as surface waters. Historically, we have attempted to account for the quality of water systems by testing water as close to the point of use as possible. But the manner in which water is stored and conveyed can be a significant contributor to the safety of the system and warrant separate assessment and evaluation. In reality, it is the water source + the storage of water + the conveyance system used that dictate the water’s ultimate “fitness for use.” Any new paradigm must consider all these factors, provide metrics to assist with their evaluation and spell out best practices to ensure system integrity.
The metrics currently being drafted and discussed are intended to prioritize risk by classifying ag water into two systems designated for specific uses within leafy greens operations. In essence, we are asking individuals to assess their water sources, storage, and conveyance and then categorize water systems as Type A or Type B.
A Type A system is one in which the source water is known to be free from fecal contamination (such as a municipal, regulated reclaimed, tested deep well or verified treatment supply) that is then stored and conveyed in a manner that does not expose the source to the outside environment. Water from Type A systems can be used in any irrigation scenario including overhead irrigation close to harvest.
A Type B system would essentially be all other water and its uses would be restricted. For example, a surface water source stored in an open reservoir and then pumped into a sprinkler set could not be used within 21 days of harvest. These are examples of how the water source + the storage + the conveyance must be fully evaluated to determine the “fitness for use” or that water is “safe and adequate for its intended use.” Of course, there is much more that must go into the development of a new ag water paradigm, but the high-level concept is sound. And incorporating industry-led changes into the LGMA requirements is the most effective way of changing practice in the field, which is what many are trying to do.
This raises the final point of concern: It is not ideal to move into siloed activity to address these outbreaks and improve our preventive programs. As a result of the most recent outbreak, the United/PMA Task Force has been created, regional groups have formed to address issues, and select parts of the supply chain have grouped together to suggest change. In the spirit of “hang together or hang separately,” it seems a better model should be developed and institutionalized going forward. In recent discussions the concepts of a “strike team” or a National Transportation Safety Board-style team has been offered, and I suggest this could be as simple as organizing a standing group of association leaders representing all aspects of the supply chain that meets and coordinates action every time there is an outbreak. It is an extra and unneeded chore to try to coordinate all the siloed efforts underway today. As trade organizations, I proposed that when it comes to consumer safety, we get over our need to individually demonstrate value and come together to make a difference for the industry and the consuming public that we serve.