Terrorist Use of Biological Weapons - Deadly Cultures: Biological Weapons Since 1945 - page 289

cluding both first-wave restaurants). 8 Only 11 cases resulting from sec-

ondary transmission were confirmed.
The two waves of infection appear to have been the result of at least
two (probably more) separate series of deliberate contamination of foods
at salad bars: the first at 2 restaurants around 8 September, and the other
at 10 or more restaurants around 21 September. New cases continued to
appear for a week or two afterward, probably the result of multiple fac-
tors: variable incubation periods (normally 6-48 hours, but sometimes
a bit longer); a common practice of reusing foods from salad bars over
several days rather than discarding them at closing time; insufficiently
low temperatures to prevent bacterial proliferation, allowing fresh food
mixed with a small amount of old to develop high levels of contamina-
tion; and some secondary transmission. However, continuing incidents of
deliberate contamination may have been responsible for some of the con-
tinuing illness, as suggested by detailed analysis of cases associated with
individual restaurants.
Although restaurant-acquired salmonellosis is common, it is rarely as-
sociated with salad bars. Furthermore, this outbreak had a number of
anomalous features that should have suggested deliberate contamination
as the likely source. These included:

• In only one of the 10 restaurants definitely identified as sources of the
Salmonella did employees become ill before customers (in contrast to the
usual pattern, in which employee infection is the source of customer in-
fection).
• Two of the involved restaurants served a total of 20 private banquets
during the outbreak, to nearly 900 participants, none of whom became
ill. The banquet food was prepared in the same kitchens by the same
people as material for the public rooms, but banquet rooms contained
dedicated salad bars.
• Illness, though associated in most instances with salad bars, was associ-
ated with a range of food items at each restaurant.
• No source of any food item was common to all affected restaurants, and
no food distributor serviced more than 4 affected restaurants.
• Affected restaurant employees had little social contact with one another,
but most had eaten at their own restaurant's salad bar or at that of an-
other affected restaurant.