The FDA tells us that it has a positive test result for salmonella on some cantaloupes produced by Agropecuaria Montelibano. The results are from samples the FDA had taken for testing at a border crossing on March 12, 2008.
Information is scarce, and it is hard to know what to make of this. The finding of salmonella is interesting because the serotype found was Salmonella Freetown, which is different from the Salmonella Litchfield strain that supposedly sickened 50 people.
We have to be careful. Cross-contamination in labs is common. It seems that the famed Chilean grapes, if they had cyanide, got it while in FDA custody, as we mentioned here. We also dealt with a food safety “outbreak” at Church Brothers/True Leaf Farms that was also prompted by lab error.
Until FDA releases the full test file and it can be examined by outsiders, we won’t know very much.
This test was done because the FDA ramped up testing before it had identified a company as responsible for the illnesses. Now the FDA views any salmonella as an adulterant. But we suspect that if we tested every cantaloupe in the country, a certain percentage would have some salmonella. The more we test, the more we find.
The more reasonable question is to ascertain if there is a reason to think that this firm’s cantaloupes have a higher incidence of salmonella than is generally to be found on the commodity.
We asked Pundit Investigator and Special Projects Editor Mira Slott to find out more:
Q: I understand FDA has obtained a positive salmonella test from a cantaloupe sample originating from Agropecuaria Montelibano. Where and when was this sample taken?
A: We do have a positive test. At the beginning of March, we had begun sampling cantaloupe imports. Based on the epidemiological analysis, we had narrowed the problem to melons but weren’t sure where the problem originated, which company was involved, so we increased the sampling to find the source. This was prior to the FDA import alert going out.
We pulled samples originating from the grower at the point of entry on March 12 that came up positive for Salmonella.
Q: Was it the same strain that was connected to the outbreak?
A: No. It was a different serotype than the one that has been making people sick. There are lots of different serotypes of Salmonella; there could be 250 variants. This one is called Salmonella Freetown. This positive test shows that cantaloupe that entered as recently as March 12 from this grower could be contaminated.
Q: Just to clarify, the positive test only relates to the Freetown strain? There have been no positive tests linked to the grower for the Litchfield strain that was related to the outbreak?
A: We haven’t found a positive sample for the Litchfield strain yet. But the epidemiological analysis and trace back analysis are implicating this firm.
Q: Who did the testing, FDA or an independent lab?
A: FDA did the testing; this was an assignment for FDA in the field, and the samples were gathered as part of that.
Q: Will FDA provide the results of the tests? Are you releasing tests so the shipper can monitor and evaluate them?
A: We are sharing the test results with the company. We follow a scientific method and we share that method and the results. We stand by our tests.
Q: Does FDA have a regular program to do sampling on a random basis to establish a baseline for comparison?
A: I don’t know what the base line is for this particular commodity. We try to sample all commodities at some rate, and it varies from commodity to commodity. In general, we try to increase samples where we think there is a greater risk. If a particular product is processed and there’s a kill step and the chance of contamination is unlikely, we may do less sampling.
We try to weight our testing so there is increased attention toward areas where we think there is potential public health concern. One type of food is more likely to support pathogen growth than another. Some pathogens will get you sick; some will kill you. If there’s a history of problems in a category, or a firm has a history of problems, we’ll focus more on that.
Suppose an average base line is every 18 months, but there’s a firm we’ve had sporadic trouble with, maybe we look at that firm every 12 months. Another firm has no problems, maybe for them 24 months. Everyone has a finite amount of resources; we’re trying to put resources where we’ll get the largest payoff. The goal is to protect people.
The goal is to have no positive result, but as a practical matter, you can’t expect that; is it 1 in every 10,000 or is it 1 in every 20,000?
Q: If every cantaloupe imported or grown in the U.S. were tested, a certain percentage would show up positive for salmonella. Therefore, increased testing would increase the likelihood to produce some positive samples (i.e., was it just terribly bad luck on the part of this grower?). Therefore, it seems important to find out the baseline.
A: I wouldn’t approach the issue from a baseline standpoint. I don’t know the baseline, but FDA’s position is presence of any salmonella renders the food to be adulterated. Clearly the better your testing capability, the more likely you are to find something. The more you sample, the more likely you’ll be to find something. Without the positive test result we had still implicated the company for the Litchfield.
Q: Do you know the concentration of salmonella discovered in the positive Freetown sample?
A: As far as concentration, I don’t know. We know as of March 12, product came in from that grower that could have been potentially contaminated. Unless it was just these samples that we grabbed, we would expect more problems. We’ve implicated their product through sampling and also through epidemiological and trace back methods.
While we were looking for one type of contamination we found another. By following GAP, you can minimize the likelihood for Salmonella contamination to happen and companies do have some ability to influence this. The 50 people that got sick don’t care about a baseline in testing. Every year, we find a few cases and we need to pursue them.
Q: Agropecuaria Montelibano goes well beyond minimum GAP standards, putting itself through rigorous third-party auditing and testing. Its food safety practices and integrity are in good standing throughout the industry. Is there no threshold that would influence FDA’s actions?
A: This positive test for a different serotype just surfaced. Yet before this, FDA’s investigation was targeted on the Litchfield serotype outbreak.
Q: From what you said in our earlier interview, the outbreak followed a bell curve shaped pattern, starting at the beginning of January and trailing off in March with the last case March 5. If the incubation period is a few days, why would it have made sense to continue widespread recalls of product believed to be connected to this serotype outbreak more than three weeks later?
A: This Litchfield outbreak is actually bimodal, with more than one peak, but trailing off in March. What if more problems exist? Basically FDA looked at the data, and we were not convinced this was over. Now with this positive test for another strain; clearly there’s something else going on, and how widespread is what we’re trying to find out. It doesn’t mean this company didn’t have a good track record.
Now FDA and CDC have a team down there investigating. I’m sure there is frustration. We need to find out why this contamination is happening. If we can help the grower determine the cause, then they can use that information to implement measures to improve pathogen control and insure product they ship to the U.S. meets U.S. safety standards. We’re down there trying to trouble shoot what went wrong. It’s often difficult to find the smoking gun because fields have been harvested. But we try and narrow it best we can.
Quite often in trace back you can’t find that smoking gun. That’s not to say there isn’t value in doing the investigation. Maybe the grower will find room for improvement in irrigation techniques, for example; whether that was the exact cause, we may never know.
Q: That being the case, how does the FDA determine when the import ban will be lifted? Is it discretionary judgment by the head of FDA? Is it based on a certain number of trailer loads testing negative? What are the criteria? Are there rules?
A: Criteria for getting an import alert lifted are in the guidance section of the alert. http://www.fda.gov/ora/fiars/ora_import_ia2202.html
Q: Imports are regulated in this way, but is there any comparable program for U.S. shipments. Do you do random sampling of domestic cantaloupes? If a problem happened with a shipper in Arizona, could FDA block him from shipping? And what would a domestic shipper do to earn his way back into the market?
A: We do domestic random testing, and if we did find a positive sample, we could block the shipper from shipping interstate commerce.
Generally speaking, once a problem is identified, a firm needs to show they’ve taken corrective actions to remedy this problem and then show so many clean shipments.
If an outbreak involves illnesses, there’s a rule of thumb… we’re probably going to have to visit the farm to be sure that the manufacturer’s processes and procedures are being properly controlled and followed. We’d like to see a plan and test results.
When the ban is lifted depends on the severity of the problem. We will be seeking some sort of assurance the problem has been corrected.
Q: How can a company give you assurance the problem has been corrected in cases where you can’t determine the source of the problem?
A: We investigate the heck out of the company and can’t find anything, then what? I’m not sure what happens in that scenario. Maybe the company is doing everything right and they have a sick worker. Maybe there is a problem with one piece of equipment that contaminates product coming from five fields. It may turn out the problem only impacts a certain percentage of their product. Discovering the source can be complicated.
Q: Is the FDA equipped with enough staff, money and resources to conduct thorough investigations on site when these problems arise? And what authority does FDA actually have to conduct investigations of these companies, especially in countries like China?
A: It’s very tough to put together investigations. We do look at all these things… irrigation, fields, processing plants. We regularly do these trace backs. If we have outbreaks with tomatoes in this country, we would investigate all those areas. We went to Guatemala during the raspberry outbreak. We’ve been to China when unapproved drugs were discovered in seafood.
But I can’t say a blanket statement that every time there’s a problem we’ll visit fields. At the same time, we can’t just show up. In the U.S. we have a little bit more leverage. For this Honduras firm, we have an import alert for this product. We have to coordinate a visit with them or suggest or be invited to come down.
All we know is that something has changed at this Honduran grower that has created a food safety problem, and now we want to figure out what it is. It doesn’t mean they weren’t trying to take preventive measures. Was there something they were missing, a control they overlooked? Something seems to have gone awry, and we’re working with the government and the company to find out the problem. This positive sample for a new serotype adds another wrinkle. It’s not uncommon when contamination takes place to find more than one type of salmonella involved.
Testing is one part of an effective food safety program. But it’s not a reliable protective measure by itself. The best we can do is try to ascertain the cause as fast as possible.
The key points are: 1. On Friday March 21, we completed testing of the grower’s product that had entered the U.S. before the import alert was put into place and that testing has now come up positive for Salmonella Freetown, another strain unrelated to the Litchfield outbreak. 2. We have a team on the ground in Honduras now investigating the cause of the problem.
There are two issues here: 1) Does the FDA reveal enough information for its tests to have credibility. Banks always have two witnesses to deposits being opened from night depositories. 2) Does a zero-tolerance policy make sense for a field grown crop such as cantaloupe?
There are real risks to this issue. Who will invest millions in farms and packing houses if you can have all the certifications, all the credibility and, even if your product has no more salmonella than the industry average, you can still be thrown into bankruptcy.
We need to really think through what we hope to accomplish with our food safety laws.