FDA seeks rapid test to detect bacteria

WASHINGTON — Wanted: Salmonella detector. Must work fast.

Send plans and specifications to Uncle Sam, care of the Food and Drug Administration.

Frustrated that conventional lab methods can now take as long as nine days to identify the most common of food bugs, the FDA is searching for a rapid test for salmonella.

Two recent outbreaks — one involving peanut butter, the other blamed on tomatoes and hot peppers — have put the agency on the spot.

Each time the FDA had pieces of the puzzle, but it took a while to fill in the complete picture. The uncertainty made consumers nervous about eating everyday foods. Food producers lost millions in forgone sales and recalled products. Lawmakers fumed. One congressman likened the government's disease detectives to the Keystone Kops.

Since other outbreaks are likely to happen, FDA officials are desperately seeking anything that would make their response more efficient.

"The goal here isn't to design some sort of 'Star Trek' gizmo," said Dr. David Acheson, assistant commissioner for food safety. "We're looking for something that can save us 12 hours here, 12 hours there. If we can shave it to five days, that would be a step forward."

Michael Doyle, head of the food-safety program at the University of Georgia, said the FDA should aim high. "To identify an outbreak can take two to three weeks; if they can get that down to three days, it would be a major step forward."

The FDA has asked the Pentagon, the Homeland Security and Agriculture departments, and the Centers for Disease Control and Prevention to lend their expertise. The Agriculture Department and the CDC also contend with salmonella outbreaks. Homeland Security has responsibility for combating bioterrorism. And the Pentagon is skilled at evaluating all kinds of technology.

"We approached these guys, and they're interested in working jointly," said Acheson.

The first step is to see whether any private companies or academic research centers are working on a rapid test. Then Acheson wants to select two or three methods to evaluate more closely. Finally he'd compare specific techniques and devices in head-to-head lab tests.

One of the reasons it can take so long to identify salmonella is that samples submitted to the lab may not have enough of the bacteria. More bacteria have to be cultured in a nutrient-rich broth to make an identification.

"I can't make the bugs divide any quicker," said Acheson. "But what if we had tools that could work off a smaller number of organisms? I think there is time to be shaved there."

If the initial screening finds salmonella, more testing is needed to match its particular genetic fingerprint to the outbreak strain.

The easiest thing would be to have a portable device that inspectors could carry with them. They could take a tomato, pulverize it, inject the juice into the device and get an answer in a matter of hours.

"That would be the Holy Grail," said Acheson.