At the 2001 American Society of Microbiology Annual Meeting in Orlando, FL, session, “Microbiology of Imported Foods: Oversight and Regulation:’ panelist Dean O. Cliver, Ph.D., University of California, Davis, School of Veterinary Medicine and member of the World Health Organization Collaborating Center for Food Virology, noted that the challenges of detecting and controlling harmful bacteria, viruses and protozoa on imported produce remain complex on both the scientific and regulatory fronts. However, some of the lessons learned from a review of selected case histories of foodborne illness outbreaks associated with produce that occurred in the U.S. and Canada over the past decade provide insight into how the produce industry might proceed in the face of these challenges.

“As far as imported food is concerned, what gets imported is based on the climate wherever the food is being grown and labor costs,” said Cliver. “Part of what is driving the market in the U.S. is that some people who claim to have credentials in nutrition are telling us that ‘fresh’ food is more nutritious than its counterpart that has been frozen or canned, but I consider that an unproven assertion. From a public health standpoint, what we’re talking about is foods that we eat raw.

“As a result of these drivers, we see a lot more fresh, or perishable, produce imported into the U.S., and while each commodity group has some idea of the flow of product into and out of the country, there really isn’t any way of knowing the rates of microbial contamination or incidence of foodborne illness associated with contamination of fresh fruits and vegetables,” he continued. “What typically happens is that when an exotic pathogen like Cyclospora rides in on a food, although relatively few people are affected, it suddenly becomes an international incident. It is blown out of proportion, but what proper proportion is, is anybody’s guess.”

The case histories selected by Cliver, were chosen with regard to whether they represented chronic or recurrent problems. A few instances of sporadic cases were noted. He described the following cases associated with cantaloupe, sprouts, raspberries, strawberries, parsley and mangoes:

• Cantaloupe. In June and July 1991, more than 400 illnesses in the U.S. and Canada were attributed to the contamination of these melons with Salmonella Poona. The melons were imported by Canada and allegedly originated from two Texas counties.

In May and June 1998, an outbreak of Salmonella Oranienburg was reported in Ontario, Canada, which caused 22 illnesses. The traceback investigation could not rule out the U.S. as a source of the cantaloupe, but noted that the countries of origin could also have been Mexico or a few Central American nations.

Most recently, in May 2001, an outbreak associated with cantaloupe imported into the U.S. by two Mexican companies caused approximately 30 illnesses involving eight U.S. states. The contaminant was identified as Salmonella Poona. Investigators were able to determine that the cantaloupe did not come from the U.S., because the harvest season had not yet begun.

• Sprouts. The venues for a 1994 outbreak of Salmonella Bovismorbificans-associated illnesses were Sweden and Finland. Some 282 illnesses were reported in Sweden and approximately 40 in Finland. It was determined that the alfalfa seeds were from Australia.

Salmonella Stanley was the culprit in a 1995 outbreak which affected Finland (28 illnesses) and the U.S. (128 illnesses), in which each shared a source and a serotype. In this case, the source was a broker in The Netherlands, although the alfalfa seed could have originated either in Italy, Hungary or Pakistan.

Also in 1995, Salmonella Newport in sprouts caused 133 illnesses in the state of Oregon and one Canadian province, British Columbia. Six other U.S. states also reported illnesses associated with these sprouts. The alfalfa seeds were traced to a broker in The Netherlands.

• Raspberries. One of the most notorious of international foodborne disease outbreaks involved a recurrent problem with Cyclospora cayetanensis in raspberries exported by Guatemala. In 1996, there were more than 1,465 illnesses reported in the U.S. and Canada; in 1997, both countries again experienced outbreaks. In 1998, the U.S. decided to embargo raspberries from Guatemala, but Canada did not and had another outbreak. (Cliver noted that this latter proved to be “one of the more elegant controlled experiments” in this type of outbreak. Even so, puzzles remain: “The oocysts of Cyclospora take days to weeks to sporulate and become infectious after they’re shed in feces, and there is no known alternate source on humans,” said Cliver “so how these got onto raspberries, which have very short shelf life, can’t be washed at source, and so on, is still very much a mystery.”)

As a result, Guatemalan raspberry growers operate under a U.S.-developed system called the Model Plan of Excellence. At present, only four Guatemalan farms still are trying to export raspberries to the U.S. “This Model Plan of Excellence is much more rigorous than anything we have in the U.S.,” noted Cliver. “I think if they tried to apply this in California, you’d evoke an armed battle.” During the export season from mid-March to mid-August, he added, 10% of the workforce from each of these four farms is randomly selected to contribute fecal samples for analysis for Cyclospora. Since May is the biggest month for export, 100% of the workforce on each farm has to contribute fecal samples for analysis in the first few days of the month, or the farm will not be allowed to export raspberries to the U.S.

• Strawberries. In February and March of 1997, frozen strawberries were identified as the food that caused 256 illnesses containing hepatitis A, an outbreak that spanned several U.S. states and in which Michigan was hardest hit. The strawberries were grown in Mexico, but they were processed in San Diego, CA. It has never been shown how the contamination occurred. (“Part of the reason this outbreak garnered so much publicity was that these were served in USDA school lunches,” said Cliver, “and someone had to certify falsely that these had been grown in the U.S. in order for them to be purchased by USDA for use in the school lunch program.”)

Cliver mentioned two outbreaks categorized as sporadic in nature. The first involved an August 1998 outbreak of Shigella sonnei associated with parsley, in which 478 illnesses were reported in three U.S. states and two Canadian provinces. Indications are that the parsley came predominantly from Mexico, where a water problem existed in the packing sheds; primarily municipal water was used, but the water was not such that the locals would drink it.

The second sporadic outbreak involved mangoes contaminated with Salmonella Newport, which occurred in November and December 1999. While there were only 78 resulting illnesses, the outbreak is notable because it affected consumers coast to coast in 13 states from Maine to California. Cliver added, “This particular outbreak was detected on the basis of pulse-field gel electrophoresis (PFGE) fingerprinting, one of the new technologies that lets us connect sporadic cases to see a larger picture of what is occurring, which may prove to be a new way of detecting outbreaks. In this instance, the traceback investigation reportedly pinpointed one farm in Brazil as the source, which, if accurate, is remarkable.”

Perspectives on Produce Food Safety
As far as this set of case histories is concerned, stated Cliver, the main aim is to illustrate that very seldom has it been possible to determine how the produce became contaminated, partly due to the heterogeneity of contamination and partly due to the difficulties in adequately obtaining representative samples from large-sized lots. The lack of good methodology, in some cases, also contributes to the challenge. “This is true particularly with the nonbacterial pathogens such as Cyclospora, hepatitis A and Norwalk virus,” he said, “which generally have not been detected in the food samples. One of the objectives of my lab is to try to develop methods for detecting these agents in foods, but we’ve found that they are only useful if applied when the situation arises.

“Given that we do not have solid data on the true levels of contamination coupled with the heterogeneity of contamination, I’m not a great one for surveillance testing. At the very least, however, I do feel that when an outbreak has occurred, it would be great if the investigators were armed with valid test methods to apply to whatever commodities are available. Hepatitis A, for example, is particularly problematic, because the average incubation period is four weeks and by the time you recognize the outbreak, you’ve typically got very little sample still around to test. In other words, in my 39 years of experience, I know that seldom is an alternate etiology suspected until the food has been tested endlessly for bacteria, and by that time there’s either quantitatively nothing left to test or qualitatively, it’s in such bad shape that it isn’t worth testing.”

As a result, the U.S. has begun to encourage its trading partners to adopt Good Agricultural Practices (GAPs). “But, because we don’t know how the contamination occurred, and very often we haven’t detected the pathogens in the food, our version of GAPs is necessarily based on speculation,” said Cliver. “I think it is important to realize that the U.S. is a net exporter of food; that, in fact, this is one of the few items in our whole international trade system where our balance of payments is favorable and we’d like to keep it that way.

“On the other hand, it’s not clear that our balance of foodborne pathogens is favorable—we may or may not be exporting more foodborne pathogens than we’re importing. From that standpoint, I think that it is important that we realize that we have enough foodborne disease in the U.S. to last us a long time and it is not surprising that U.S. export commodities sometimes make people in other countries sick, just as import commodities brought into the U.S. sometimes make our citizens sick. Therefore, I suggest that we try to avoid a ‘fortress mentality,’ because I think that it is very easy for us to feel aggrieved when an outbreak occurs that is associated with an imported food and not to realize that there are other people on the receiving end of these outbreaks some of the time.”

Another perspective to be gained from these cases is there are positive advances in the science and approach of traceback investigations, noted Cliver, which can provide more data to reduce the incidence of foodborne illness outbreaks associated with produce, imported or domestic. One area of advance is the development of fingerprinting methods. “Obviously, if you can detect outbreaks that you wouldn’t have recognized otherwise by fingerprinting, then in cases in which we have positives—even if they are based on isolates from clinical samples rather than from food samples—we have some opportunity to do effective tracebacks. Indeed, in the hepatitis A outbreak, it was possible to corroborate the distribution that was otherwise hypothesized, on the basis of sequencing the RNA in the hepatitis A isolates. Also, there are now methods for fingerprinting by RAPD for Cryptosporidium and Giardia.”

A second advance is being made in the area of improved recordkeeping. “Getting records that show the distribution from the farm of origin all the way to the consumer will help food safety efforts, noted Cliver. “For example, it could reduce problems with commingling, where the product loses its identity. This commingling was part of the problem with reported illnesses caused by Salmonella Enteriditis in California-grown almonds that occurred in three Canadian provinces in 2001. More than one handler was, in fact, involved, but one handler alone had its company’s name on the packages. On the other hand, you can believe that every box of raspberries that comes out of Guatemala today has its own farm of origin indicated on it, which stays with it all the way to the consumer.”

Cliver warns that while progress is being made in better recordkeeping, these practices eventually will add to the cost of food. “We need to consider whether we’re really buying enough safety to warrant all the measures we’re putting into place. This should be considered, because some of the time all these measures do is increase costs, and at the same time, contribute little or nothing to actual safety of the product.

“The people who say that they want an absolutely risk-free food supply regardless of the cost are ignoring the fact that we have upwards of 20 million people in the U.S. who aren’t eating regularly, because they can’t afford food at present-day prices,” Cliver concluded. “When we program in more alleged safety measures that may turn out to be inadequate to achieve food safety goals, one thing we are sure of is that food is going to cost more.”