TOPICS > Science

Scientists Track Source of Salmonella Outbreak Mystery

September 10, 2008 at 6:30 PM EDT
Loading the player...
A salmonella outbreak in May sparked a lengthy investigation that first called tomato crops the culprits, causing farmers to lose $100 million, and later found the true source to be Mexican peppers. Tom Bearden reports on the salmonella probe.

JIM LEHRER: Now, tracking the source of this summer’s salmonella outbreak. NewsHour correspondent Tom Bearden has our Science Unit report.

TOM BEARDEN: TV audiences love watching investigators use a combination of high-tech equipment and old-fashioned shoe leather to solve complex mysteries.

ACTOR: Ah, the old flower delivery trick. Works every time.

TOM BEARDEN: Detectives like these on “CSI” follow promising leads, and sometimes find themselves at a dead end. Except for the Hollywood glitz, that’s also a pretty fair job description for the people who investigate outbreaks of food-borne disease, the men and women who work in public health in state and federal agencies.

Fictional CSI investigators usually start with a body. In public health, it usually starts in a doctor’s office. A person comes in complaining of stomach problems. If it appears to be food poisoning, the doctor orders lab tests. In the most recent outbreak, which began at the end of April, salmonella turned out to be the culprit.

It’s a rod-shaped bacteria that can be found in meat, poultry and eggs. But, in recent years, it’s also been found in fresh produce. Once ingested, it can take several days before people become sick.

Complicating all this is that fact that, on average, only one person in 38 who gets sick actually visits a doctor.

Beginning the investigation

ALICIA CRONQUIST, Colorado Department of Public Health: Once the lab has a positive salmonella culture, they're supposed to report it to either the state or the local health department, whereby we start our investigations.

TOM BEARDEN: Alicia Cronquist is an epidemiologist with the Colorado Department of Public Health.

ALICIA CRONQUIST: In Colorado, we have local public health departments that do a lot of our disease investigation. So, they hear about this positive salmonella case. And a local public health nurse will call the person and interview them to try to figure out when they became ill, what foods he or she ate, what other sorts of activities they did. So, we ask a wide range of questions about what that person did in the week before he or she became ill.

TOM BEARDEN: In the meantime, scientists in state health departments test stool samples to determine which one of the some 2,600 strains of salmonella has made people sick.

They use an electrical pulse to break down the DNA to get its unique fingerprint.

David Butcher heads up the Colorado State Health Labs.

DAVID BUTCHER, Colorado Public Health Laboratory: The electrical current has caused the DNA fragments to travel down this gel. And the different fragments separate off. And that's what gives you these -- these hash marks down the -- the rendering here that are what's referred to as the fingerprint.

TOM BEARDEN: Once that fingerprint is determined, it is posted on a Centers for Disease Control database, which is available to all public health agencies across the country. That allows scientists to figure out whether the same fingerprint has emerged elsewhere.

Fingerprinting the source's DNA

TOM BEARDEN: Most salmonella cases -- there are 40,000 reported annually in the U.S. -- are small local affairs involving improper food handling in someone's kitchen or restaurant. When the source is identified, the investigation stops.

But, if a cluster of cases involving the same strain pop up in different places, the investigation escalates dramatically. Beginning in early May, scientists noticed that people in several states had been sickened by the Saint Paul strain of salmonella. That triggered a second round of interviews.

ALICIA CRONQUIST: Hopefully, at that point, we have some idea of some things that those folks have in common, that, hopefully, we have some hypotheses.

And then, based on those, we try to test those hypotheses, and we will create yet another questionnaire, where we will ask people some extremely detailed questions about those food items.

TOM BEARDEN: A multistate outbreak also brings the Food and Drug Administration into the investigation. Their job is to trace the actual source of the contamination.

This is where the shoe leather comes in. Investigators visit hundreds of restaurants and grocery stores, gather invoices, and try to figure who bought what from whom.

David Acheson is head of food safety at the FDA.

DR. DAVID ACHESON, Assistant Commissioner For Food Protection, Food and Drug Administration: So, at this point, you're going from one restaurant to three suppliers. And, from there, you then go to nine suppliers, because each one's getting from three, and it just mushrooms out.

TOM BEARDEN: As they follow the paper trail, they also gather samples of suspect food items. Those are sent to FDA laboratories, where teams of scientists, like microbiologists Eric Brown, Marianna Naum, and Christine Keys test for bacteria.

Salmonella doesn't really thrive on the surface of vegetables, and, in the past, sometimes didn't show up in traditional tests. So, the FDA devised a new method, soaking specimens in a nutrient solution for a day, growing the bacteria to reach detectable levels.

ERIC BROWN, Food and Drug Administration: We nurse the bacteria back to health and allow it to start entering a growth phase, so we can begin to study it.

TOM BEARDEN: After the salmonella is confirmed by a test where it reacts to an antibody, it is then cultured in different kinds of growth media. If it turns a certain color, it's salmonella.

MARIANNA NAUM, Food and Drug Administration: It does not show the specific type of salmonella, but it does give us a definitive positive of salmonella.

TOM BEARDEN: So then the bacteria's DNA fingerprint is determined, using the same process that identifies it in the original human samples.

Blaming tomatoes

ERIC BROWN: What we have been able to do now by putting into play an integration of all these three -- these different techniques, from genetic science, microbiology, and biochemistry, we can now answer the question, is, when people do get sick from eating produce, are these just rare lightning-strike events that happen in a local growing area, or are we talking about a single producer, a single distributor that's causing a real problem to public health?

TOM BEARDEN: This summer, patient interviews showed a high statistical probability that the source was tomatoes. But thousands of tomato samples didn't test positive for the type of salmonella that had made them sick.

CHRISTINE KEYS, Food and Drug Administration: If it's one of these mixed food types, of it it's something that is ubiquitous -- you know, every American household is going to have tomatoes in it, and those tomatoes come from all over the place -- you know, it's more difficult to trace back, when we have to screen through lot of samples to find the one.

In salmonella Saint Paul, there are over 500 specific PFGE patterns, and so we're looking for one specific pattern out of all of those.

DR. DAVID ACHESON: Meanwhile, the cases were continuing, going into more states. We were heading up to 1,000 cases. Like, this isn't stopping. This is not stopping. We put advice out, don't eat tomatoes. It wasn't stopping, maybe slowing down a little, but that was making the state and the local health departments and CDC saying, well, maybe we should go back and do another case control study. Let's -- let's repeat this whole process. So, they did.

Mysteries remain

TOM BEARDEN: The new interviews pointed to jalapeno and serrano peppers. So, investigators started over, going back through the food supply chain. Two months after the initial outbreak, the intersecting lines began to converge at a food distribution center in McAllen, Texas. This time, the lab tests revealed an exact match.

DR. DAVID ACHESON: So, we -- we trace back into Mexico, and we go to a distribution center in Mexico. And that -- that took us back to a farm in -- in Mexico. So, we go up and we inspect our farm, and we do testing. So, we had taken it from these clusters through a distribution center in Texas back to a farm in Mexico.

TOM BEARDEN: The FDA then issued a warning not to eat Mexican peppers.

So, case closed? Well, real life isn't quite as tidy as a TV show. The Mexican government says their tests at the same farm didn't find salmonella on the produce. And, as for the tomato growers, they claim to have lost $100 million due to the scare. Health agencies, however, insist the fruit may still have been responsible for the earliest outbreak.