Seed Safety

ISS Seed Screening for Human Pathogen Testing

Over the last decade commercially produced sprouts have been implicated in outbreaks involving thousands of people.

"In almost all cases, contaminated seed was the source of the contaminated sprouts both for Salmonella sp. and E. coli 0157:H7."  Lester M. Crawford, D.V.M., Ph.D. Acting Commissioner of the FDA.

International Specialty Supply developed a method of screening sprouting seed for human pathogens that has been confirmed, referenced or recommended as necessary by top food safety regulatory agencies and experts, including:

  • The World Health Organization (WHO)
  • United States Food and Drug Administration (FDA)
  • United States Department of Agriculture (USDA)
  • California Department of Health
  • New South Wales Food Authority
  • South Australian Research and Development Institute
  • Ontario Ministry of Agriculture
  • The Center for Science in the Public interest
  • Resources for the Future
  • Food Standards Agency (UK)
  • The Camden Research Group
  • Food Safety Authority of Ireland, 
  • National Center for Food Safety and Technology
  • Illinois Institute of Technology
  • Center for Food Safety and Applied Nutrition
  • Dozens of researchers and food safety experts. 

The ISS Seed Screening Program has already prevented several lots of contaminated seed from being used in commercial sprout production.

The Source of Contaminated Sprouts is Contaminated Seed

The suspected scenario in historical outbreaks of illness tied to sprouts is that the seed is contaminated in the field by manure used as fertilizer, or by grazing animals, or in silos or bins by bird or mouse droppings and urine.  The process of sprouting allows nearly ideal conditions of food, moisture, and warmth in which a single pathogen cell can proliferate into over one hundred thousand pathogen cells.

So Why Not Just Decontaminate the Seed?

A great majority of the research on alleviating sprout related outbreaks has focused on decontaminating the seed.  And although the FDA went so far as to recommend that all commercially grown sprouts be produced from seed that has been sanitized with 20,000 ppm chlorine, no sanitizer has yet been found to be reliable at eliminating pathogens from seed without destroying  germination.

If the pathogens were sitting on the seed coat, it would not be a problem -- 200 ppm chlorine would do the trick.  But the bacteria that are trapped in cracks and crevices in the seed coat appear to be protected from disinfectants, and there is evidence that some bacteria can be internalized within the seed itself.  When a single pathogen cell remains alive, it will multiply back to levels as though no sanitation had ever been done.

To date, no sanitation technique has been shown to completely eradicate pathogenic micro-organisms from either the seed or the sprouts.

Avoiding Outbreaks is Not Difficult if the Seed Is Not Contaminated in the First Place

In 2000, Bob Sanderson of Jonathan Sprouts came up with a simple, yet brilliant, idea. "If the problem is pathogens in the seed, why not just make sure there aren't any pathogens in the seed in the first place?"

In a collaborative effort between International Specialty Supply and Jonathan's Sprouts, a method was developed in which sprout growers and seed suppliers could substantially reduce the odds of there being pathogens in seed.

Seed is Generally Tested for Quality, Not Safety:

  • Seed sampling for seed borne plant pathogens has been practiced for many years and the statistical probabilities are well documented.

    For over a hundred years seed has been tested for germination, purity, hard seed and other properties that relate to the quality of the seed.  These are addressed in terms of percentage.  For example, a seed lot might have a quick germination of 90% with 3% hard seed and 1% abnormal seeds, for a total germination of 94%. The statistical probabilities used to determine these percentages are extremely accurate. But these percentages are not particularly relevant to testing seed for human pathogens.

    Testing Seed for Human Pathogens

    When testing for pathogens, one is not looking for a percentage, but for any at all. If a single pathogen is detected, the lot is contaminated and cannot be used for sprouting purposes. 

    To test adequately, the entire sample, which in a truckload of seed can be over 20 Kg (44 lbs) (nearly a thousand times larger than a normal lab sample), is sprouted, which increases the bacteria level approximately 5 logs (100,000 times) in 48 hours. Then the runoff water is sampled, enriched, and tested for Salmonella, E.coli 0157:H7, and generic E.coli.  A portion of the sprouts themselves are crushed and tested for Salmonella, E.coli 0157:H7, Listeria monocytogenes, Shigella and generic E.coli.  They are also tested in a separate lab for Enterotoxigenic E.coli (ETEC) if generic E.coli is found.

    ISS Seed Screening Procedures

    The seed screening procedures, developed by International Specialty Supply for ISS Screened Sprouting Seed, can substantially reduce the risk of food-borne illness related to commercial sprout production.  The process involves: seed sampling, seed inspection, sprout growing (enrichment), spent water sampling, enrichment of sampled water, and pathogen testing.

    Every step is critical, must be precisely done, and accurately documented.

    Seed Sampling

  • In order to detect a pathogen in a lot of seed, you first need to "capture" it in the seed sample you intend to test.  The probability of picking up at least one contaminated seed in a 40 lb seed sample is very high at an average contamination level of 4 CS/kg (4 Contaminated Seeds per 2.2 pounds of seed), but very low with smaller samples and smaller contamination rates.

    Probability of Capturing a Pathogen in Alfalfa Seed Contaminated with Various Pathogen Levels Per Kilogram
     Seeds Sampled Pathogens (cs/kg)
    Kilos Seed Count 0.1 1 2 3 4 5 10 100
    0.000002                 1 0% 0% 0% 0% 0% 0% 0% 0%
    0.00002               10 0% 0% 0% 0% 0% 0% 0% 0%
    0.0002             100 0% 0% 0% 0% 0% 0% 0% 2%
    0.002           1,000 0% 0% 0% 1% 1% 1% 2% 18%
    0.02         10,000 0% 2% 4% 6% 8% 10% 18% 86%
    0.2       100,000 2% 18% 33% 45% 55% 63% 86% >99%
    1       500,000 10% 63% 86% 95% 98% >99% >99% >99%
    2    1,000,000 18% 86% 98% >99% >99% >99% >99% >99%
    3    1,500,000 26% 95% >99% >99% >99% >99% >99% >99%
    5    2,500,000 39% >99% >99% >99% >99% >99% >99% >99%
    10    5,000,000 63% >99% >99% >99% >99% >99% >99% >99%
    12    6,000,000 70% >99% >99% >99% >99% >99% >99% >99%
    25   12,500,000 92% >99% >99% >99% >99% >99% >99% >99%
    50   25,000,000 >99% >99% >99% >99% >99% >99% >99% >99%
    100   50,000,000 >99% >99% >99% >99% >99% >99% >99% >99%
  • Probability (P) = 1-(C/T)^N, where (C/T) = assumed ratio of Clean seeds to Total, and N = Number of seeds sampled.

  • Table. The Probability of Capturing a Contaminated Seed in a Twenty-Ton Lot of Alfalfa Seed Contaminated with Various Contamination Levels Per Kilogram.

    The above table is theoretical and assumes even distribution in the bags that contain contamination.  The number of contaminated seeds in an outbreak has not been investigated, nor has pathogen distribution.  Although pathogen distribution in a lot is usually not homogeneous, distribution within the bags themselves likely is.

    Seed Inspection

    sprouting_seed_inspectionAfter a representative sample is taken, the composite sample of seed is inspected for indicators of contamination.  This includes inspecting the bags for mouse urine or dropping, holes in the bags, insect larva, bird droppings, etc. The seed is then carefully inspected with both a magnifying glass and microscope to determine its fitness for human consumption.  Again, we look for traces of visitation by animals and insects.  In the process we have also found glass, seed that was blended with seed that had been treated with a fungicide, and other things.

    Sprouting (Enrichment)

    The entire sample is sprouted.  The seed is not sanitized prior to sprouting.  In 48 hours the pathogens, if present, should have increased about 1,000,000 times, substantially increasing the probability of detection.

    Water Sampling and Testing

    At about 48 hours, a sample of the runoff water is collected and enriched to make the pathogens multiply about 5 log (100,000 times) and the water is tested for salmonella, E.coli, and E.coli 0157:H7.  The tests are run in duplicate.

    Sprout Testing New!

    Some of the sprouts produced from the composite sample are sent to a separate lab.  That lab crushes the sprouts and tests for Salmonella, E.coli 0157:H7, Listeria monocytogenes, Shigella and generic E.coli.  They are also tested in a separate lab for Enterotoxigenic E.coli (ETEC) if generic E.coli is found.


    Each step is thoroughly and accurately documented and signed by the person taking responsibility for completing each step of the process.  The adage is, "It if isn't documented, it wasn't done."


    We add an extra step as well.  Before we receive a shipment, we bring in a sample and inspect it, sprout it, and test it.  We stop at the point of rejection.  That is, if the seed does not pass visual inspection, we reject it without taking the time to sprout it out and test it for pathogens. 

    Generally if we receive seed with mouse dropping in it, that same processor will send bad samples time after time, and year after year.  After a while, one learns who to stay away from. 

    We are not only screening seed, we are screening seed processors.

    Don't All Seed Companies Follow These Procedures?

    ISS developed this program, and appears to be the only company in the world who does this extensive testing.

    Some seed companies will take a handful of seed from a bag and send it to a lab for pathogen testing.  The lab will take 20-grams (less than an ounce) from the sample and test it for pathogens.  Yet in previous outbreaks in which contaminated seed was recovered the contamination levels have been as low as 2 pathogens per pound of seed (4cfu/kg). 

    What do you think the likelihood of finding those two pathogens is when testing less than an ounce of seed?  It is less than 8%. 

    So why do labs only test 20-25 grams?  They do not have commercial sprout equipment!  And what are they going to do with the 360 lbs (163 kg.) of sprouts produced from a 36 pound sample of seed?

    So is ISS Proposing That Sprout Growers Don't Need to Sanitize their Seed or Post-Test for Pathogens?

    Far from it, but if seed companies had been selling seed that was not contaminated, the sprout industry would not be in the situation it is.  The FDA would not be alarmed at all, and sprouts might actually be considered one of the safest products in the produce department.

    That is water over the dam, and it will take many years of no outbreaks before sprouts are considered a safe product again.  But because pathogens can be internalized in seed or protected in biofilms on sprouts, the sprout industry cannot count on sanitization alone.  The best solution is to start with seed that is well screened and to verify through post testing that the seed screening and sanitizing was effective.

    What are the Weaknesses of Seed Screening?

    The screening process outlined above is based on the probability of capturing a contaminated seed.  But once captured, the pathogen needs to be detected.  As of this date, no test for salmonella or E.coli O157:H7 is perfect.  According to the papers highlighted in the previous sentence, the best one can expect from a test kit is detecting contamination 97% of the time it is present in a sample.  To overcome this shortcoming, the water samples need to be tested in duplicate, which brings the probability of detection closer to the probability of capture.

    Similar to ISS' Seed Sampling, the probability that at least one of the duplicates is positive is greater than the probability of a single test being positive (given that the pathogen/organism is present).

    So, if a pathogen test has a sensitivity of 97% (i.e. probability of a positive test given the sample is positive) then a testing scheme utilizing duplicates should have an overall sensitivity of 1 - (1-0.97)^2 = 0.9991 - which, is virtually 100%.

    A weakness of the ISS Seed Screening Program is the possibility of basing decisions on inaccurate information.  You are forced to rely on seed lot information provided by other people.  A seed lot is supposed to be one uniform lot.  Suppose someone along the line-- a farmer, seed processor, seed supplier, etc. -- retags seed. The reliability of the information provided by the test would be lower than the probability charts.  For instance, if you purchase 800 bags of lot A,  and a seed supplier decided they could get rid of the last 30 bags of lot B, the last 18 bags of lot C, and the last 2 bags of lot D by retagging them all as lot A.  Your test may still be quite accurate for the 750 bags that originally made up lot A, but not very reliable for the fifty bags that made up lots B, C, and D.  Although this marginalizes seed screening, it would have still been effective on 750 bags, which is 94% of the lot and therefore still a useful risk reduction step.

  • Probabilities of detection require knowledge of the quantity and distribution of pathogens in the particular seed lot. This information is not known, so actual probabilities cannot be predicted. However, sampling on some seed lots that have been implicated in outbreaks has been able to isolate pathogens, which strongly suggests that if this sampling had been done beforehand, the seed would have been diverted to non-food uses, and the outbreak would not have occurred.

    ISS Seed Screening Does Not do the Following Things:

    • It does not cost much.
    • It does not produce hazardous wastes.
    • It does not put production workers at risk.
    • It does not effect germination, vigor, yield, or the quality of the sprouts.
    • It does not reduce background flora.
    • It does not introduce the possibility of resistance.
    • It does not disenfranchise organic sprout growers.
    • It does not negatively affect or take away from other food safety procedures, such as decontamination.

    What ISS Seed Screening Does:

    • It helps the seed industry identify practices that introduce pathogens into foods.
    • It can prevent contaminated blended lots from entering the market if sampling is done prior to blending.
    • It can prevent some seed with visible contamination from entering the market.  This includes glass and other things that would not show up on a pathogen test.
    • It is most effective when sanitizers are least effective.
    • It helps identify farms and seed processors lacking Good Agricultural Practices.
    • It is used identically on all types of seed.
    • It affects commercial sprout growers of all sizes, with all levels of sophistication and using all types of equipment. It can be applied evenly throughout the sprout industry.
    • It can be used on seed destined for home sprouters, and may be the only realistic defense that home sprouters have.  Chlorine is not a very good option for home sprouters.
    • It shifts some of the responsibility of providing safe seed to the seed industry itself.
    • It has substantially reduced the number of food borne illnesses from sprouts. It is only a risk reduction step though.  Seat belts reduce the risk of injury in traffic accidents, but people still get hurt. 

    Who Should Do These Procedures?

    It is the growers who are ultimately responsible for producing safe sprouted products.  They can do these procedures themselves, they can get screened sprouting seed from International Specialty Supply, or they can do both.  The more the seed is sampled and tested, the less likely it is to be contaminated.  No matter who does it, the grower needs to have good documentation of exactly what was done, when it was done, and who did it.  Without this documentation, growers (and health inspectors) need to consider the seed unscreened and unsafe.

    These procedures do not eliminate or reduce pathogens in the seed.  If seed is contaminated before it is screened, it will still be contaminated after it is screened.  The procedures reduce the risk that the seed may contain the human pathogens that they were screened for.  If International Specialty Supply does these procedures, it does not guarantee that there are no pathogens in the seed.  It does guarantee that ISS did everything it signed off on in writing for a particular lot of seed.  Once the seed is properly sampled, inspected and tested, a document is signed by ISS making it "ISS Screened Sprouting Seed"

    These tests are extensive and vary from seed to seed and lot-to-lot.  ISS cannot practically test every lot in every instance.  Drop shipments, for example, are almost never tested.  With rare exception, all of the seed that comes into ISS' warehouses in Tennessee and California are tested.  But the only way to know for sure is to receive the documentation from us for the particular lot of seed you are interested in purchasing.

    ISS strongly encourages growers to get the safety documentation regarding each lot of seed they purchase.  A copy of ISS' documentation is attached.  A copy of the ISS Seed Sampling and Testing Flow Chart and its Seed HACCP plan are available upon request.  A copy of the ISS Seed Screening Documentation is available by pressing the navigation bar at the top of this page.

    Who Else Believes in the ISS Seed Screening?

    • The FDA invited ISS to present their seed screening procedures at the five year meeting in May, 2005.  The FDA will very likely adopt ISS' seed screening procedures as recommendations for all seed sold as sprouting seed.
    • The FDA has advocated the "Six-step procedure developed by ISS" in numerous reports and presentations.
    • In his keynote address to the June 9, 2009 Scientific Session of the ISGA Convention in Chicago Dr. Martin Cole (Chairman of the International Commission on Microbiological Specifications for Foods (ICMSF)and Director, National Center for Food Safety and Technology (NCFST)), discussed the theoretical possibility that ISS Seed Screening protocol may be equivalent to a 2.34 log reduction or better.
    • The New South Wales Food Authority implemented seed screening in their Plant Products Safety Manual and noted that "The seed pre-screening sampling process is based on the method developed by International Specialty Supply of Cookeville. TN, USA and Jonathon Sprouts of Marion, MA USA." (Pages 28, 31 and 33)
    • Its applicability to Australian conditions was reviewed and confirmed by the South Australian Research and Development Institute.
    • The Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA), adopted ISS' seed screening program as recommendations for seed companies and sprouters in its "Sprouted Seeds Good Manufactures Guidebook." 2007.
      • "Using similar testing methods, other seed suppliers should also be able to supply safe seed."
    • The World Health Organization commented that:
      • "Outbreak investigations have indicated that microorganisms found on sprouts most likely originate from the seeds"
      • "There is currently no treatment available that can guarantee pathogen free seeds."
      • "Seed producers, distributors, and sprout producers should test lots of seeds for microbial pathogens"
      • "Sprouting seeds before testing increases the possibility of finding pathogens that may be present. If lots of seeds are found to be contaminated, they should not be sold or used for the production of sprouts for human consumption."
      • "Failure to find contamination does not guarantee that the seeds are pathogen free. However, if contamination is found at this stage, it allows seeds to be diverted or destroyed before entering sprout production for human consumption."
    • Codex Alimentarius, the Code Of Hygienic Practice for Fresh Fruits and Vegetables, notes in Testing of seed lots before entering production that:
      • The seed sample selected for testing should be sprouted prior to analysis to increase the potential to detect pathogens if present. Analysis may be performed on the sprouted seeds or the water used to sprout the sample.
      • Seed samples for microbial analysis should not be subject to any microbiological decontamination treatment at the sprouting facility.
    • The Center for Science in the Public Interest advised the FDA:
      • [ISS] "has demonstrated that seed screening has many benefits and is very inexpensive.  FDA should mandate that seed distributors conduct this kind of screening and certify their seeds as safe [Actually, ISS certifies them as being screened], in order to [help] prevent contaminated seed lots from entering the market. This could significantly reduce the number of outbreaks associated with sprouts."
    • Resources For the Future, a Washington DC based think-tank, in its seminar titled "Achieving A Safe Food Supply in Increasingly Global Markets" wrote:
      • [ISS] "developed a program of seed screening that has allowed it to avoid all sprout-related outbreaks or recalls in North America since 2000, causing health organizations worldwide to start recommending ISS screening procedures be used on any seed destined for sprouting."
    • FSA (Food Standards Agency, UK) notes that:
      • "Outbreak data indicate that the seed used for sprouting is the most significant source of the pathogens implicated. ?bacteria that are trapped in cracks and crevices in the seed coat appear to be protected from disinfectants and there is evidence that some bacteria can be internalized within the seed date no technique has been shown to completely eradicate pathogenic micro-organisms from either the raw or sprouted seed...Given the difficulty of decontaminating seed another important aspect the detection and isolation of pathogens from the seeds used to produce sprouts."
    • The Campden Research Group in England in their 2004 report titled "Review of microbiological risks associated with sprouted seeds" concluded that
      • "Absence of pathogens in seeds is critical and, consequently, microbiological testing of seeds prior to use for production of sprouts is essential."
    • Seed sampling is also suggested in the Codes of Practice for Food Safety in Ireland.
      • "As seeds are thought to be the most likely source of contamination?The dried seed should be sampled and tested microbiologically upon arrival. Seed contamination is thought to be sporadic, at low levels or unequally distributed throughout seed lots. Therefore, a negative result does not guarantee the absence of pathogens; however a positive result allows a producer to avoid using contaminated seed lots."
    • Leading researchers for the USDA and FDA, William F. Fett, Tong-Jen Fu, and Mary Lou Tortorello in "Seed Sprouts, the State of Microbiological Safety" Chapter 6 of Microbiology of Fresh Produce, Edited by Karl R. Mathews, 2006 pp202-207.
      • "Proper sampling inspection and testing of seeds for pathogens can substantially reduce the chance of using contaminated seeds for sprout production and therefore help prevent foodborne illness.  However, improper seed screening protocols may still result in outbreaks."
      • "A comprehensive seed sampling and testing procedure has been developed [ISS]."  Then they explain ISS' process described above and close with: "The adoption of these seed-screening procedures has already prevented at least four potential sprout-related outbreaks do to Salmonella and E.coli O157:H7."
    • Food Safety Researcher and Professor Keith Warriner supported ISS' approach in "Interventions to Improve Food Safety of Sprouted Seeds", his PowerPoint Presentation for British Columbia's Premier Food Safety Conference - October 18, 2007.
    • The California Department of Public Health highly recommends sampling in their Advisory to Commercial Sprout Producers, May 1, 2008 (Comment #2).  Their advice is based on ISS' seed screening program.
    • Trevor Suslow, Plant Sciences and Postharvest Technologist, UC Davis, in his presentation titled "Evaluating Seed Borne Pathogens as a Risk Factor in Produce Outbreaks" for the Plant Disease Seminar, Nov, 13, 2007 concluded his presentation with this:
      • "Recommendations for Seed Industry GAPs & BMP's -  "Most critically, before engaging in any seed testing program, a uniform and standardized test procedure must be determined.  Statistically sound sampling protocols need to be established...The experience of the sprout seed industry with pathogen-free certification programs can provide insight into approaches and problems (SproutNet has valuable and practical information on testing programs and access to abstracts of published seed food safety research)." Trever Suslow
    • In the study "Factors Influencing the Growth of Salmonella during Sprouting of Naturally Contaminated Alfalfa Seeds" FDA and National Center for Food Safety and Technology researchers referenced the ISS Seed Screening Program when they point out that:
      • "To prevent the consumption of contaminated sprouts, the absence of pathogens in seeds must be assured. Screening of seeds by seed suppliers for the presence of pathogens can help to reduce the number of contaminated seeds entering the marketplace."

    sprout page seporator

    NOTHING ELIMINATES RISK:  ISS Seed Screening or the use of ISS Sprouting Seed is one of several methods that should be employed by commercial sprout producers to reduce the risk of human pathogens in sprouts.  Seed should also be properly sanitized prior to production and the sprout run-off water should be tested for human pathogens between 24 and 48 hours of production.

    For more information on seed sampling and testing see "Seed Sampling and Testing: A risk-reduction strategy for sprouts."