The hit-and-miss struggle of German health authorities to identify the contaminated food behind one of the deadliest E. coli outbreaks in recent years underscores the difficulties of following a pathogen through the complex food supply chain, as well as deficiencies in even the most modern health systems in diagnosing this deadly illness.

After mistakenly suggesting that Spanish cucumbers were the likely culprit several days ago, German authorities focused Sunday on bean sprouts from a German farm, only to report on Monday that the first 23 of 40 samples from that farm had tested negative for E. coli. The results from the remaining samples had yet to come back. That does not entirely eliminate the farm as the outbreak's origin, since even one positive test is sufficient to make the connection.

But determining the origins of an outbreak that has killed 22 and left 600 people in intensive care presents a difficult mystery to unravel, with vital clues disappearing day by day as contaminated food is thrown away and farm and factory equipment is cleaned. Patients -- whose illnesses first alerted health authorities to the outbreak -- may have only cloudy memories of the meal that landed them in the hospital. Did the sandwich last month in Hamburg contain sprouts, tomatoes, lettuce, cucumbers -- or all four?

After E. coli infection, diarrhea can take a week or more to emerge and it takes another week before the most serious complications, like kidney failure and anemia, occur.

That means that as German investigators interview patients and visit farms to hunt for traces of the germ, the smoking gun may be long gone. Finding the offending food "is sometimes going to be easy and sometimes going to be difficult and I think this is one of those," said Dr. Robert Tauxe, deputy director of the division that handles food-borne diseases for the United States Centers for Disease Control.

"What did you eat four weeks ago?" Dr. Tauxe said. "You're dealing with memory here -- so it's hard to pull apart."

He said that even if hundreds of patient stories pointed investigators to a particular food -- say, bean sprouts -- it might be impossible to prove conclusively that they are to blame. To do so, scientists must visit the restaurant, farm or food processing plant and find the germ in water, or on food or other material.

"Even if all the samples are negative, maybe you just missed it," Dr. Tauxe said. "You can go to a place reeking of chlorine, and find nothing."

Indeed the largest serious outbreak of E. coli, which sickened more than 8,000 people in Japan in 1996 has been widely attributed to eating contaminated radish sprouts, but scientists were never able to prove contamination in the laboratory.

To identify suspected sources, scientists painstakingly question victims about what they have eaten, searching for statistically significant patterns.

Some scientists complain that health systems in the United States and Europe are not using all available tools to better diagnose toxic E. coli; earlier detection would allow for more effective treatment and limit the scope of outbreaks, they say. Routine tests for patients with bloody diarrhea look for other germs -- shigella, salmonella and campylobacter -- but not E. coli. Such tests are readily available but more expensive.

"This suffered from a lack of primary diagnosis, and that's crucial because no one is looking for it," said Dr. Flemming Scheutz, the Copenhagen-based head of a World Health Organization collaborating center that specializes in E. coli. "The first alert didn't go out until people were hospitalized."

Likewise, Dr. Scheutz expressed concern that the decentralized German health system had made doctors initially unaware of the scope of the outbreak, familiar only with the cases in their region. A nationwide reporting system collects such information in the United States.

Dr. Lothar H. Wieler, a professor of veterinary medicine at the Free University of Berlin, said he was reluctant to criticize health officials. But he said German vets had a faster warning system for infectious disease than doctors who treated humans did. Germany has strict laws on sharing patient data.

"It would certainly be good to have a reporting system where officials get data very fast," he said. "That is technically possible but must be implemented."

But over all, said Frederic Vincent, a European Commission spokesman for health and consumer policy, German officials had acted logically in confronting "one of the worst E. coli outbreaks ever."

Mr. Vincent defended their decision to publicly blame Spanish cucumbers, even though it proved unfounded and has decimated Spain's farm economy. He said that 80 percent of the ill Germans had recalled eating the vegetable, that many of Europe's cucumbers were grown in Spain and that initial tests had revealed the presence of E. coli on Spanish cucumbers -- though it later proved to be a different strain.

E. coli strains range from benign to highly toxic. Accordingly, infections range from mild disease to a potentially deadly illness. The toxin-producing strains can result in severe bloody diarrhea and, in some cases, a syndrome characterized by kidney failure and anemia.

There were about 3,500 confirmed cases involving those strains in the European Union last year, Mr. Vincent said, and nearly 1,000 in Germany in 2008, according to the European Center for Disease Prevention and Control. But many if not most cases probably go undetected. A few countries, like Ireland and Denmark, test aggressively for the disease, so their numbers appear high, Dr. Scheutz said.

The outbreak in Germany was caused by a relatively rare strain of E. coli, O104, which possesses 2 potentially deadly qualities encoded into its genes: it produces the so-called shiga toxin and it also sticks to intestinal walls.

Dr. Alfredo Caprioli, who runs the European Union reference laboratory for E. coli in Rome, said the strain had previously been identified as the culprit in a few cases, including two in Germany in 2001; in those no food source was ever implicated.

Because the strain is rare, scientists remain uncertain of just how much toxin it produces or even where it hides in nature, which make tracing and combating outbreaks more difficult. The high number of severe illnesses in the current outbreak could be because the strain is extraordinarily virulent, because the food that caused the outbreak was highly contaminated or because a huge number of people were exposed to it, Dr. Tauxe of the American disease centers said.

But the outbreak does have some classic features. Women have been infected far more frequently than men have, which is common when the contaminated food involves raw vegetables, rather than meat -- the other common source of E. coli. "In the past, when we see the majority of patients are women, we think salad," said Dr. Scheutz.

Two weeks into the outbreak, German health officials have a lot to sort out. "These can take a very long time -- we may never find it," said Mr. Vincent of the European Commission, noting that a 2006 E. coli outbreak that resulted from contaminated spinach originating in California took over a month to work out.

"But hopefully we can get very lucky," he said.

First Published: June 7, 2011, 4:00 a.m.