Of microbes and milk; probing America's worst salmonella outbreak

FDA Consumer, Feb, 1986 by Chris Lecos

In the end, the task force was unable to unearth any evidence within the plant of the existence of the same Salmonella typhimurium strain that was causing the food poisonings. Nor could it prove how the bacteria got into the plant in the first place to contaminate the low-fat milk.

A report released by the task force on Sept. 14 disclosed various defects in the plant's operation and equipment and, more importantly, what the investigators felt were the most likely causes of the outbreak. But while "potential problems" were identified in the plant, the task force was unable "to reconstruct an unbroken chain of probable events that led to intermittent contamination" of the dairy's low-fat milk products. A "combination of defects" probably played a role in causing the oubreak, according to the report.

The task force concluded that the most likely source of the outbreak was a stainless steel pipe called a cross-connection. The pipe, about 10 feet long, was linked on one side to piping that carried unpasteurized milk and on the other o pipes carrying pasteurized skim milk. (The connection is also called a skim milk transfer line.) Valves at each end were supported to prevent unpasteurized milk from mixing with pasteurized products.

Investigators suspected the cross-connection early on, after it was discovered by Charles Price, the senior regional milk specialist from FDA's Chicago office, when he went to the plant on April 2. Price discovered that sometime in the past the Hillfarm Dairy had converted one of its raw milk storage tanks into a tank for storing pasteurized skim milk. Tracing the piping from the tank, he found the cross-connection.

The cross-connection was a "modification" to the plant's original engineering design and had been installed sometime between 1975 and 1979. It had several functions. It was used when the dairy wanted to route pasteurized skim milk from the storage tank for use with such products as ice cream. If there wad contaminated skim milk in the line, it did not affect those products becuase they were pasteurized afterwards. The cross-connection also was used at the end of a day's production to remove for reprocessing whatever milk was left over in the pasteurized skim milk tank.

"Just because it was there," Kozak pointed out, "did not mean it caused the contamination. There was still a question of figuring and demonstrating how contaminated milk could get through the cross-connection and then into the pasteurized skim milk line." Presumably, this was supposed to be prevented by appropriate shut-offs.

So another elaborate test was undertaken. Like other U.S. dairies, Hillfarm employed what is known as a post-pasteurization blending process to produce its low-fat (2 percent) milk--the type linked to the outbreak of the illness. This involves blending already pasteurized whole milk (3.25 percent or more milk fat) with already pasteurized whole milk (3.25 percent or more milk fat) with already pasteurization was done since the process involved the mixing of already pasteurized products. Post-pasteurization blending is not uncommon; some dairies around the country have used the process for more than 20 years.