'It's the sprouts,' the head of the Federal Robert Koch Institute in Berlin, Reinhard Burger, announced on 10 June. No surprise. Bean-sprout food-poisoning outbreaks occur regularly. The first big one was caused by Salmonella in the UK in 1988, when 143 people fell ill after eating contaminated mung bean sprouts. The outbreak in Sakai City in Japan in 1996 was caused by radish sprouts contaminated with E.coli O157. In the UK last year 231 people were infected with Salmonella from bean sprouts.

The majority of bean-sprout outbreak victims are women, probably because children and men are less likely to eat them. The outbreak can have been running for a while before it’s noticed. Investigations are often started only when there are an unusual number of cases caused by a rare strain of microbe. And in the majority of outbreaks the bacteria are not found on the suspect sprouts. The German outbreak has all these features. But there’s strong evidence that incriminates the sprouts as the vector for E.coli O104. A study of travel groups and club members who had eaten at a restaurant was crucial. Menus, receipts, information about ingredients and photographs showing food on plates were compared to identify what the sick had eaten but the well hadn’t. There is also a geographical relationship between case clusters and the distribution chain of sprouts from the producer in Lower Saxony.

Raw bean sprouts are a high-risk food because beans can have invisible contamination on them, and because the conditions used for sprouting – many hours in the warmth and wet – are optimal for the growth of bacteria. The only possible ways to reduce the risk are to use beans from a trustworthy grower who doesn’t get manure on his harvest, and decontamination. The best chemical decontaminant is bleach, but the highest permitted concentration won’t kill all the bugs. The only process that looks promising is irradiation. It’s hard to see organic sprouters going down this route.