Associated Press Writer

W

ASHINGTON -- A century ago, at the dawn of food safety laws, inspection amounted to little more than opening a bin of flour and looking for something wiggly.

It's a different story now.

Solving the case of the poison processed peanuts took marathon work by federal scientists, clues in Canada, Oregon, Ohio and Connecticut, and a breakthrough in Minnesota at the hands of public health hotshots known as Team Diarrhea.

So labyrinthian has the nation's food production and distribution network become that a salmonella outbreak that has sickened 575 people in 43 states and resulted in the recall of more than 1,500 foods is traced to one plant making a mere 1 percent of the county's peanut products.

It's a far cry from 1906, year of the Pure Food and Drug Act, when Americans got most of their food from local farmers, grew some of their own and turned to processed products mainly for simple staples.

Those weren't necessarily the good old days. Shady operators packed sawdust and talcum into grains and sugar to add bulk. Toxic chemicals were used as preservatives. Much was foul and filthy before modern refrigeration and sanitation.

But when maggots munched in a sack of barley, it wasn't much of a mystery who let that happen.

In contrast, the peanut potboiler has posed a series of mysteries: a whatdunit, a whodunit, and the still-urgent question, where did it all go? Now, the government has added another layer of intrigue to one of the largest recalls in history. It's accused the peanut plant of shipping products the company knew had tested positive for salmonella.

n

The whatdunit

A chain of events has to play out for the feds to discover an outbreak of foodborne illness. Sick people have to go to the doctor. The doctor must order tests. The lab must perform the tests correctly, then report the results to state or local authorities who must tell the federal government.

Then the real detective work begins.

The peanut case quickened the pulse of federal scientists on Nov. 12, more than a month after people started getting sick, when the federal Centers for Disease Control detected a cluster of salmonella cultures with an unusual genetic fingerprint reported from 12 states.

That was "a blinking light," Dr. Ali Khan, assistant surgeon general, told Congress.

PulseNet, a national network for finding patterns in widely dispersed foodborne bacterial illness, offered additional clues when four more states reported the cluster.

Then, on Dec. 2, scientists began examining a second salmonella cluster with a similar genetic makeup, reported from 17 states. The two clusters turned out to be the same.

"December is when the alarm bells go off," Khan said.

To solve a whatdunit, public health officials need to know what type of salmonella has caused an outbreak and what food is carrying it. There are more than 2,500 kinds of salmonella, each divided into subtypes.

For a while, chicken was suspected, as well as peanut butter.

Enter Team Diarrhea.

Last year, aggressive gumshoe work by Minnesota health officials helped pinpoint hot peppers as the source of a national salmonella outbreak wrongly blamed on tomatoes.

Investigating the new outbreak, they discovered in late December that some of the patients they interviewed lived or had eaten at one of three places -- a nursing home, another long-term care institution and an elementary school.

Those places shared a food distributor in North Dakota. And they had only one product in common on their shelves -- an Ohio brand of peanut butter sold to institutions.

That was one eureka moment. More would come.

Minnesota officials subsequently found salmonella in an opened five-pound container of peanut butter at the nursing home. It turned out to be the same strain as the one in the outbreak.

Not everything was falling into place. The peanut butter could have been tainted at the nursing home after it was opened. Not all victims in the country had eaten at an institution. Indeed, some hadn't eaten peanut butter at all out of a jar.

Gabrielle Meunier, whose 7-year-old son Christopher spent nights in the hospital sick from the outbreak in Vermont, told lawmakers the mystery poison might have been identified much sooner if the government had a secure Web site where victims could communicate with each other.

"Had I had an opportunity to talk to other mothers whose children were sick, and compare what they had eaten, I have no doubt we could have cracked this case back in early December," she said.

The inevitable delay between someone's illness and its reporting to the feds meant that no one had an up-to-date picture at any one time.

As Khan put it: "I look at this as akin to driving while looking at the rear view mirror."

Still, the outbreak strain of salmonella typhimurium had been identified, and peanuts were known to be the culprit in some way.

Leads were taking federal officials and their state counterparts across the country ever closer to the source. Only by knowing who caused this could it be stopped.

n

The whodunit

The tainted peanut butter found in Minnesota carried the King Nut label, and that company was quickly investigated by the Food and Drug Administration, as were other suppliers to institutions. King Nut issued a precautionary recall.

This trail also led farther back to the plant that supplied the peanut butter to King Nut -- the Blakely, Ga., operation of Peanut Corp. of America. An initial round of recalls was announced by Peanut Corp. The company's products leave the plant in containers of up to 1,700 pounds for peanut butter, and 35 pounds to entire tanker trucks for its peanut paste.

Those supplies go to hundreds of companies for reprocessing as ingredients in well over 1,000 foods.

In mid-January, officials in Connecticut found the outbreak strain in a previously unopened jar of King Nut peanut butter, made by the Blakely plant. Federal officials say this was the first strong indication that the peanut butter was contaminated before it left the plant.

How to explain the fact that some victims hadn't eaten peanut butter in a school, at a nursing home or out of a jar at all?

Scientists were looking hard at that question. First, they determined that many victims had eaten other products containing peanut butter as an ingredient.

Then they focused on two brands of prepackaged peanut butter crackers. Both were made at one plant, and that plant had been supplied with paste by the Peanut Corp. operation.

Now the investigation had crossed borders. The Canadian Food Inspection Agency retrieved intact packages of crackers from a patient in Canada who had bought them in the U.S. A test found the outbreak strain of salmonella in the crackers.

Three more intact packages bought by a victim in Oregon were tested, too, and found to be contaminated.

One mystery solved. And rapidly unfolding events at the Blakely plant would soon settle the whodunnit.

The FDA said its inspection, ending Jan. 27, found two salmonella strains at the plant. Although different from the outbreak strain, the discovery was telling. So was the observation of roaches, mold, a leaking roof and other sanitation problems.

The government, which says the outbreak might have contributed to eight deaths, has started a criminal investigation. The company denies any wrongdoing.

n

Where it went

As well as tracing the contamination back to the source, officials have to follow serpentine trails forward to try to figure out all the final destinations. The list of recalls, and possibilities, keeps growing.

Recalls now include cookies, crackers, cereal, candy, ice cream, pet treats and much more.

In addition to having its products spread through the marketplace, Peanut Corp. has been a supplier to the government.

Federal officials said Friday they are shipping 660,000 new emergency meal kits to Arkansas and Kentucky after discovering many packages they sent earlier, to help people recover from an ice storm, contained the recalled peanut butter.

The Agriculture Department said it had shipped some of the company's potentially contaminated peanut butter and peanuts to eight states, including school lunch programs in California, Minnesota and Idaho, in 2007. The department has suspended business with the company.

Also Friday, the government said a closer examination of company records shows that in 2007, it shipped chopped peanuts after salmonella was confirmed by private lab tests. In other cases, officials said, the company sold products that had tested positive without waiting to receive a second round of testing that eventually came back negative.

That raised another question about Peanut Corp. executives, one familiar to Washington from its own history of political scandal: What did they know, and when did they know it?

Q&A on computer network that warned of salmonella

By MIKE STOBBE

AP Medical Writer

ATLANTA -- U.S. health officials were first alerted to the current peanut-based salmonella outbreak through a government computer network called PulseNet. But the system relies on doctors testing patients, and it takes interviews with patients to identify the food source. Here is an explanation of how this system works:

Q

: What is PulseNet?

A

: It's a national network of public health labs coordinated by the U.S. Centers for Disease Control and Prevention. When state and local health officials get lab results of people sick with food poisoning, they post information about cases on the WebBoard, the PulseNet listserv. State and national health officials use this information to look for patterns.

PulseNet gets its name from a lab technology called pulsed-field gel electrophoresis, which enables investigators to do DNA "fingerprinting" of the infection bacteria and its strain subtype.

It's somewhat like an FBI database that can compare fingerprints from an arrest record in California to fingerprints at the scene of a crime in Florida and determine if it's the same crook. It takes about two days to run a genetic fingerprint on a sample. Each year, PulseNet identifies more than 300 clusters of patients infected with salmonella, E. coli, shigella and listeria.

Q

: How comprehensive is it?

A

: This network can't

catalog every case in an outbreak. Only some people who get sick go to doctors, and not all doctors run tests to confirm what infection the patient has. Health officials estimate that the actual number of illnesses in an outbreak may be 10 times higher than the lab network reports.

But by detecting an outbreak while it's still going on and then identifying the food that's spreading it, health officials believe they can prevent countless illnesses and some deaths.

Other countries have similar systems. The CDC routinely shares data with Canada, but not with other countries.

Q

: How long does it take to identify a foodborne germ?

A

: It generally takes two to four weeks from the time the first person in a cluster gets ill until the cluster is detected by PulseNet. In the case of the peanut butter outbreak, CDC first detected a national pattern in November, the month after substantial numbers of lab-tested illnesses first emerged.

CDC officials did not disclose the outbreak until January. They said it took more than a month for health officials to interview sick patients to determine that peanut butter was the food they all had in common. Then, a test from a peanut butter container in Minnesota found the outbreak strain of salmonella.

Q

: Can't this process be sped up?

A

: The PulseNet system itself is relatively rapid. But investigation timelines are driven by what patients and doctors do about illnesses, and how quickly health officials in individual states react.

"The PulseNet system is a great system. The problem is it's a 22nd-century system resting on pillars of epidemiologic research that go back to the 1800s," said Bill Marler, a Seattle plaintiff's attorney who specializes in national food poisoning cases.

Such limitations may explain why Florida has had no lab-confirmed cases in the current salmonella outbreak. Unlike some other states, Florida does not require doctors and hospitals to send salmonella samples to a state lab for analysis and does not budget to handle such a workload. At least 270 samples that may be the outbreak strain have been collected, but only 25 were sent to a state lab for genetic fingerprinting.

Q

: How long has this network been in place?

A

: PulseNet was launched in 1996 but didn't have the participation of all states until 2002. The cost is shared by the CDC and states. CDC says it spends about $5.6 million annually but has no figure for what states spend.

The origin of the system lies with a 1993 outbreak of E. coli food poisoning in the western United States. CDC successfully used DNA fingerprinting to nail the strain of E. coli O157:H7 that sickened hundreds and killed four children; it was traced to hamburger patties served at Jack in the Box restaurants. That's when the idea of a collaborative system among the states

and federal government began to take shape.

You must be logged in to comment.