NEW YORK -- After a few people turned up dead, officials sprang into action -- interviewing the next of kin, identifying suspects and pursuing leads. It sounds like your classic murder investigation, except in this case, the suspect may be a leafy vegetable, a juicy hamburger or some errant cow manure, and the investigators are wearing lab coats instead of trench coats.
Call it E.S.I. -- Epidemic Scene Investigation. The recent E. coli outbreak in Germany, which has resulted in 22 deaths and sickened over 2,200 people, has sparked the largest epidemiological investigation in living memory, say researchers. And just like in any run-of-the-mill crime drama, it has resulted in mistaken suspects, false leads and plenty of frustration.
German officials have said they are almost certain that the source of the outbreak was an organic bean sprout farm in northen Germany, after incorrectly blaming Spanish cucumbers at first. But so far, samples of the suspected sprouts have turned up negative, prompting criticism that the German handling of the investigation has been haphazard and has even helped exacerbate the crisis.
Sprouts have been a common culprit in such outbreaks, having been linked to 30 outbreaks over the past 15 years. In the previous worst case, which occurred in Japan in 1996, 12 people died and more than 9,000 fell ill from tainted radish sprouts. Sprouts are grown in hot, humid conditions -- ideal breeding grounds for E. coli bacteria.
In 1997, E. coli detected in alfalfa sprouts infected 108 people in Michigan and Virginia. Health department specialists and epidemiologists in both states quickly isolated the specific strain of E. coli found in victims and conducted interviews to determine the source of the sprouts, finding that the alfalfa fields at the four suspect farms were possibly contaminated by cattle manure, run-off water or deer feces.
Frances P. Downes, a lab director at the Michigan Department of Health, helped investigate that outbreak and explained the procedures followed in such a probe.
Once a person gets ill, their doctor orders a culture to find out what bacteria is the cause of their illness. If they find E. coli, salmonella or certain other bacteria, the doctor is required by law in most states to forward the information to the local or state health department.
"Once it gets here, we do additional testing to confirm that that's what it really is and then we do additional testing -- is it a specific type of salmonella or E. coli?" said Downes. The bacteria's chromosome must then be isolated and chopped into pieces through DNA fingerprinting.
"You separate them based on size -- it looks like a barcode -- and that gives you the genetic makeup of that particular strain of bacteria. From that information, you can focus an investigation on seven people if you get matching genetic makeups for that bacteria," Downes said.
Epidemiologists then interview the victims to look for commonalities in the food they've eaten, since it's highly probable that the bacteria that sickened them came from the same source. "You call those seven people and go through their food history, what restaurants they ate at, what day did you eat there, where did you buy that milk or ground beef," said Downes. "Maybe they all went to the same water park. It's exactly like detective work. You're IDing a suspect and pursuing the leads."
Such sleuthing can be complicated since victims can't always remember specifics about their diets, just like crime victims can have a hard time recalling specifics about their assailants.
"Sometimes it's a long time between when they got sick and when you're interviewing them -- you might be just now hearing about people who got sick at the end of April," said Diane Woolard, an epidemiologist at the Virginia Department of Health. "Sometimes we ask them to get a calendar or their checkbook or store receipts to remind them what they might have eaten. If they don't have that, you ask them to think about that time of year -- in a typical week during that time, what would you have eaten? Strawberries? What store would you have bought it in?"
By gathering as much information as possible from sick people, scientists can look for common threads and use that data to hypothesize about the likely source. But if it's not easy to find overlaps, then scientists check with the Centers for Disease Control in Atlanta, which maintains a database of illness reports from around the country. "If our seven people didn't have enough recall but people from four other states could give a history that helps identify the suspect, that helps," Woolard said.
The investigation then moves back into the laboratory. To test their hypotheses, scientists set up a study, giving questionnaires to the sick people and a control group of healthy people who share characteristics with the sick people -- matched by age group, same area code or attendance at an office party. Again, it can be complicated -- sometimes researchers have to ask restaurant managers for the phone numbers of other parties who had a reservation that night.
"You have to have that comparison group because maybe everybody ate Wonder Bread -- it doesn't mean the bread made people sick. Maybe it was popular," said Woolard. "You have to have a way to compare the exposure sick people had with the well people."
The failure to conduct that type of test led to the conflicting conclusions of the research into the German outbreak, said Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota. He explains that that only by tracing back the infected items to a single supplier can you really determine the likely cause of the outbreak.
Once a likely source is identified, the investigation starts to provide vital information that can prevent an outbreak from happening again. "Say the sprouts come from a particular store -- we work with them to find out what lots were sold on that day and get those off the shelves," said Downes.
Once those are removed, the researchers go back to where the culprit was produced. "It's not just the whodunit but also to prevent it from happening again," Downes added. If the sprouts, for example, came from a farm, researchers test the water and seeds and examine the agricultural practices, employee habits and how waste is treated.
As with hardened police detectives, it's important for scientists to keep an open mind. "Sometimes, it looks like it might be a smoking gun but you have can't assume anything. We didn't think that E. coli could come from apple cider because it was too acidic but there were several cases that proved us wrong," said Downes.
Sometimes, investigators hit dead ends. "We want to figure them all out but you can't," said Woolard. She thinks it's unfair to blame the German researchers for their difficulties tracking the source of the outbreak. "The public expects ready answers and they're not always so ready."