Only about two-thirds of states require laboratories to submit salmonella specimens that contain the DNA fingerprints that confirm an outbreak. In the rest, it's merely voluntary, an msnbc.com survey showed.
Some states test every salmonella sample they collect using pulsed-field gel electrophoresis, or PFGE, while others check only some. In Wisconsin and Texas, for instance, only about half are screened.
The result? Outbreaks like the current one can go undetected, delaying warnings about illnesses and recalls of poisonous foods. People became ill from eating tainted peanut products as far back as Sept. 1, but it was November before the outbreak was detected and early January before it hit the public health radar.
“It’s that whole idea of finding needles in haystacks,” said Dr. Robert Tauxe, deputy director for foodborne illnesses at the federal Centers for Disease Control and Prevention. “We would like virtually all of the salmonella to be tested if we can.”
More information might have raised alarms earlier, Tauxe said, about the unusual Salmonella Typhimurium types linked to tainted peanut products that have sickened at least 655 people in 44 states and Canada, contributed to nine deaths and forced national recalls of more than 2,200 peanut products.
The outbreak has been blamed on problems at a Blakely, Ga., processing plant run by Peanut Corp. of America.
Mandatory testing urged
Food safety advocates have lobbied for years to require mandatory PFGE testing to detect the growing number of national outbreaks caused by foods ranging from spinach and peppers to peanut products.
“Having less than 100 percent compliance lowers the sensitivity of outbreak detection,” said John Besser, clinical lab manager for the Minnesota Department of Health. “The current system was designed to test local events such as the church potluck. The way you make the system better is by getting salmonella isolates tested.”
But states with voluntary programs say cooperative laboratory arrangements keep them on top of salmonella surveillance and that strained budgets and limited staffing force them to make hard choices. The same staffers who work on salmonella often are also monitoring HIV infections and tuberculosis, for instance.
"If there's a problem, we don't just sit around and wait. We ask them relatively strongly to test for the isolates," said Doug McBride, press officer for the Texas Department of State Health Services. "But do you do an isolate of salmonella and risk not getting an isolate of something more serious?"
Staff at the CDC’s PulseNet, a cooperative network of public health and food agency labs, first noticed the current outbreak on Nov. 10, when PFGE testing detected the same unusual salmonella isolates — two kinds officials hadn't seen before — in 13 patients in 12 states.
CDC officials won't say yet which states first signaled the outbreak and whether submission and testing is mandatory there.
Still, the reports were enough to raise fears about a multi-state outbreak in a commercial food product. Interviews in early January that compared foods eaten by sick patients with those of well people quickly implicated peanut butter as the source of infection.
Strains linked to outbreaks aren't always obvious
But in other situations, the cause is not so clear. Last summer, an outbreak of Salmonella Saintpaul linked first to raw tomatoes and then to fresh peppers sickened more than 1,400 people, hospitalized nearly 300 and contributed to two deaths.
One of the problems in that case was that Texas, which confirmed more than 550 cases, did not require mandatory testing of isolates, said Dr. Tim F. Jones, the state epidemiologist in Tennessee and an international food illness expert.
“It led to a delay,” Jones said.
But Linda Gaul, an epidemiologist who heads the Texas foodborne illness team, said their system detected the problem within a day of neighboring New Mexico, which requires testing.
"I honestly don't know in Salmonella Saintpaul last summer how we could have done it any differently," she said.
The state is considering changing to mandatory submission, but Gault says she has confidence in a surveillance system that receives isolates from about 1,500 of the state's 3,000 salmonella cases each year.
"If we're getting 50 percent we're not going to miss anything," she said.
Salmonella is a common infection confirmed in about 40,000 cases annually, but likely responsible for perhaps 38 times that many illnesses, according to the CDC. There are about 2,500 subtypes of salmonella, but only about 400 circulating at any one time.
More states have moved to require mandatory submission of salmonella isolates in recent years. In 2005, only about half of laboratories said they were governed by laws requiring submission, according to the Association of Public Health Laboratories. By 2007, that had jumped to more than 60 percent.
An mnsbc.com survey of all states and the District of Columbia showed that two-thirds of states now require labs to submit salmonella isolates to the state for testing.
What’s not clear, however, is how many states actually test for PFGE patterns. Smaller states such as Washington, where submission is mandatory, are able to test every specimen in the 650 to 850 cases confirmed each year, said Dr. Marcia Goldoft, a state epidemiologist.
“If you don’t have mandatory testing, you don’t know what you’ve got,” she said.
Even Idaho, which relies on a voluntary system, received and tested all of its 155 cases in 2007, said Dr. Christine Hahn, the state epidemiologist. That success makes it hard to argue for mandatory changes in a system that works well, Hahn said, but she added that the trouble in some states is obvious.
“You’ll notice on the national map some states where no or very few cases have been reported,” she wrote in an e-mail. “It’s hard to believe that those states truly had nobody diagnosed with salmonella during that time.”
In Florida, where one case has been confirmed as part of the current outbreak, officials began looking for the rare salmonella type in January, only after the crisis was identified, said Doc Kokol, communications director for the state health department. They received 45 isolates, instead of the usual five or 10 they might have gotten. Of those, 20 were Salmonella Typhimurium and one was the outbreak strain.
In California, where submission of salmonella isolates is required, PFGE testing is not. With 5,000 cases of salmonellosis reported in the state each year, laboratories perform PFGE tests when there’s a spike over baseline levels, spokesman Ken August said.
'Totally at the mercy of the states'
Mandatory testing is only one of several changes urged by health officials and advocacy groups looking to safeguard the nation’s food supply. More detailed interviews of salmonella patients and quicker testing could speed up detection and response to outbreaks, they say.
Even under best conditions, it can take two to three weeks from initial illness to confirmation, health officials said. In some cases, it might be much longer, said Dr. Kirk Smith, head of the Minnesota Department of Health unit that investigates food illnesses.
“When there is recognition that there is a multi-state outbreak, we need to go into this emergency mode,” he said, adding later: “People rip on the CDC for not solving these things quickly enough, but they’re totally at the mercy of the states.”
Requiring state reporting would help detect outbreaks earlier, reducing human infection, illness and death, said Ernest Julian, a Rhode Island health official who also heads a workgroup for the Council to Improve Foodborne Outbreak Response. There's no doubt in his mind that would have been the case with the current outbreak.
"I think you would have picked it up even faster," he said. "You have to have the mandatory reporting. That's key."
While states complain about dwindling budgets and PFGE tests that run $100 apiece, the costs when an outbreak occurs can dwarf those figures. “Obviously, one bad player can kill a whole industry,” Julian said.
Early industry estimates suggested tomato growers lost at least $100 million when federal officials first implicated the produce as the cause in the Saintpaul outbreak, then switched to peppers. Although firmer figures aren't available, that number is likely a "gross underestimate," said Julia Stewart, a spokeswoman for the Produce Marketing Association.
Improving surveillance nationwide would cost far less, perhaps $50 million, said Besser, the Minnesota food safety expert.
“It’s possible to do this on a national level and it wouldn’t be that expensive,” he said. “Put your money where you can make a difference.”