Non-compliance with regulation may have fueled 2005 Salmonella outbreak in orange juice

Outbreak highlighted need for cooperation between business and regulatory organizations.

A multistate outbreak of Salmonella infections stemming from unpasteurized orange juice in 2005 demonstrated the importance of cooperation between state and local epidemiologists, CDC investigators and FDA regulation.

Seema Jain, MD, is part of the enteric diseases epidemiology branch in the division of foodborne, bacterial and mycotic diseases at the National Center for Zoonotic, Vectorborne, and Enteric Diseases at CDC. She was involved in the outbreak investigation and the publication of the corresponding article.

“CDC and the state and local health departments worked together very well during this investigation,” she said. “This outbreak also taught us that we need to work with companies.”

Initial investigation

In June and July 2005, the Michigan Department of Community Health conducted an investigation involving 14 cases of illness due to S. Typhimurium.

“The shoeleather epidemiology happened in Michigan,” Jain told Infectious Diseases in Children. “It was diligent, detail-orient epidemiology.”

Jain said that there was a well-defined hypothesis involving a product not normally associated with foodborne illness: orange juice. After the hypothesis was formed, interviews were conducted. Interviews revealed that 11 of 13 patients had consumed store-brand fresh-squeezed orange juice that had come from one of two high-end market retailers.

Further investigation by health officials in Michigan determined that one juice processor in Florida, defined by the investigators as “company X,” had supplied juice to both retailers. The investigators went through many items on a comprehensive questionnaire with a subset of patients at the outset. They identified the most common elements of behavior and determined a certain range of commonality, according to Jain.

“The initial investigation narrowed it down to orange juice, and then down to two stores,” Jain said. “Once orange juice was identified, an extra questionnaire was developed to hone in on the product and brand. Sometimes this is necessary because consumers may not be completely aware of what they are drinking. That was especially true with this product because the juice came under a lot of different labels.”

Jain said that in addition to the diversity of labels, a confounding factor particular to this investigation was that many people drink orange juice by the glass at restaurants and have no idea where it is coming from. “We went to pretty good lengths to figure out the source of the juice,” she said. “This impacted the initial investigation, as well as the case-control study.”

Role of technology

A key component of this investigation was the use of technology. CDC requested that state public health laboratories conduct pulse-field gel electrophoresis (PFGE) analysis on all S. typhimurium and S. saintpaul isolates that had been identified from May 1 to Dec. 6, 2005. Local health departments were informed through an e-mail listserv and through Epi-X, the Epidemic Information Exchange system, that patients should be interviewed using a standardized questionnaire.

“This outbreak was detected because of an initial investigation by the Michigan Department of Community Health,” said Herbert L. DuPont, MD, chief of medicine at St. Luke’s Episcopal Hospital in Houston. “It was also detected because of the unique PFGE pattern of the Salmonella isolates studied in the state laboratories.”

The cases contained PFGE patterns that were indistinguishable by two restriction enzymes.

“It was important to do typing using PFGE because S. typhimurium is very common,” Jain said. “Initially, it was hard to see the signal. Typing helped narrow down the search. This specific pattern was unique.”

The S. typhimurium strains involved in the outbreak had not previously been seen in the national molecular subtyping network for foodborne pathogens known as PulseNet.

“A unique PFGE pattern was identified that could then be employed as a fingerprint to look at the extent of the multistate outbreak,” DuPont said.

FDA investigation

FDA officials revealed that company X had not complied with the juice Hazard Analysis and Critical Control Point (HACCP) regulation. The investigation uncovered S. Saintpaul isolates from the orange juice of company X. However, the precise route of contamination remains unknown.

At the time of the investigation, Sherri McGarry was the foodborne emergency coordinator at the Center for Food Safety and Applied Nutrition at the FDA. “We saw deviations in compliance with HACCP regulations,” she said. “However, because we arrived on the scene after the contamination had occurred, we can only speculate on the nature of the conditions. They may have changed in natural time.”

According to the terms of the juice HACCP regulations, all processors are required to apply control measures capable of achieving a 100,000-fold pathogen reduction in the number of microorganisms of interest to public health. Company X had clearly not taken these steps. However, after a nationwide alert on July 8, 2005, the company voluntarily recalled all juice that may possibly have been contaminated a day later.

“There was a lot going on around this time,” McGarry said. “Epidemiologic information was coming in from Michigan and other places. Other states had other pieces of the puzzle. There was a trace-back to be conducted, laboratory work to be conducted. We needed to find points of service and obtain records about what kind of product was available. Throughout all of this, there needed to be uniformity and consistency. The FDA put all of this together.”

The environmental investigation revealed that all of the orange juice from company X was produced in one plant. Oranges were brought to the plant through three avenues: directly from orchards, from the cold storage of external packing houses or from the cold storage of company X. Fruit brought from external packing houses was sorted, damaged fruit was culled and the remaining fruit was washed and waxed. Damaged fruit was not culled and remaining fruit was not washed or waxed when taken from the cold storage of company X.

The fruit from company X underwent surface treatment. Although the details of this process were not available, washing with water and a sanitizer, washing with a brush and steaming were included.

Extracted juice was chilled to a temperature between –2.2°C and –1.1°C. Final product was shipped for delivery within 72 hours and had a shelf life of 12 to 22 days, depending on storage temperature.

Orange juice from the company contained the words ‘all-natural fresh-squeezed’ on the label. However, there was no indication as to whether the product was pasteurized or treated with pathogen-reducers.

“What we found was that the same product was being distributed under different labels,” McGarry said. “Although different customers wanted different brand names, all of the juice came from a common source.”

Distribution of the juice from company X reached 31 states, Puerto Rico, Canada, France and Japan.

Outbreak details

There were 152 cases of S. Typhimurium or S. Saintpaul identified in 23 states. The researchers acquired detailed information for 95 of those cases, about half of whom (48%) were female.

The median age of patients was 23 years.

Results from investigation in five states indicated that for 38 patients and 53 age-group matched controls, orange juice consumption was associated with illness (90% vs. 43%; OR, 22.2; 95% CI, 3.5-927.5).

Further analysis using a logistic regression model demonstrated that consuming unpasteurized orange juice from company X was also associated with infection (53% vs. 0%; OR, 38.0; 95% CI, 6.5-infinity).

Legacy

Jain said that the 152 cases were likely just the beginning, and there may have been as many as 6,000 salmonella infections associated with this outbreak nationwide. She said it has had an effect on the health care community.

“On a basic level, this investigation identifies gaps and ways to improve public health,” Jain said. “It shows the utility of outbreak investigations.”

DuPont said that in the United States, the FoodNet and PulseNet programs, which were integral to the investigation of company X, have been extremely effective in identifying foodborne pathogens. “These programs will be the foundation of new efforts being developed to prevent and control future outbreaks in this country,” he said.

Jain said that the listserv and Epi-X made a dramatic difference in the outcome of the investigation.

McGarry is currently the director of the Division of Public Health and Biostatistics and center emergency coordinator at the Center for Food Safety and Applied Nutrition at the FDA. She and Jain said individuals should be more aware of the products they are consuming, and labels should include information as to whether the product is pasteurized or not.

“There has been historical interest in products that are pasteurized or unpasteurized vs. so-called pure or natural products,” McGarry said. She said that consumers will wonder whether a product is truly fresh if it is pasteurized. “Unless you are going to follow some mechanisms to reduce potential pathogens, an unpasteurized product can be a risk. We need to be following the science and methods for pathogen reduction. This is outlined in the juice HACCP.” – by Rob Volansky

Jain S. et al. Clin Infect Dis. 2009;48:1065-1071.