Outbreak of Salmonella serotype Kottbus Infections Associated with eating alfalfa sprouts — Arizona, California, Colorado, and New Mexico, February-April 2001

Morbidity and Mortality Weekly Report, Jan 11, 2002

On March 12, 2001, the California Department of Health Services (CDHS) identified a cluster of Salmonella Kottbus isolates with indistinguishable pulsed-field gel electrophoresis (PFGE) patterns. During February 1-May 1, CDHS identified 23 patients with S. Kottbus infections in several California counties and an additional patient from Arizona. This report summarizes the results of the investigation of this outbreak, which identified cases in four stares and implicated alfalfa sprouts produced at a single facility.

The median age of case-patients was 36 years (range: 9-72 years); 16 patients (67%) were female. Twenty-one patients developed an acute diarrheal illness, and three patients had urinary tract infections. Three patients were hospitalized.

Using a standardized questionnaire, a matched case-control study was conducted. A case was defined as culture-confirmed S. Kottbus infection with onset after January 2001 in a California resident with an isolate having the outbreak PFGE pattern. The first 10 reported California patients were matched with two controls by age group, sex, and city prefix code. Fifteen (63%) of 23 patients ate alfalfa sprouts during the week before becoming ill. A significant association was found between eating alfalfa sprouts and illness (matched odds ratio: 5.5; 95% confidence interval = 1.2-26.1). No other food or restaurant exposure was significantly associated with illness. Following the case-control study, 32 patients infected with the outbreak strain of S. Kottbus were identified in California (24), Arizona (six), Colorado (one), and New Mexico (one).

A traceback investigation identified a single sprout producer as the source of the contaminated sprouts. Review of the sprouter's production records indicated that a single seed lot was temporally associated with the dates of illness onset. A culture of a sample of this seed lot yielded S. Kottbus. These seeds were imported from Australia in November 2000, but no further information about the distribution of this seed lot was available. Cultures from two floor drains in the production facility also yielded S. Kottbus. Patient, seed, and environmental isolates all had indistinguishable PFGE patterns.

Although the implicated seed lot was last used on March 29, the sprouter issued a voluntary recall of all sprout products on April 17, and ceased all sprout production pending further internal review of their production processes. Review of decontamination and distribution records indicated that at least some seeds underwent heat treatment followed by a 2,000-ppm sodium hypochlorite treatment for 15 minutes. The U.S. Food and Drug Administration (FDA) recommends decontamination of seeds with one or more treatments (e.g., soaking in a 20,000-ppm calcium hypochlorite for 15 minutes) that have been approved for reduction of pathogens in seeds (1,2). The effectiveness of alternative seed decontamination has not been established. The sprout producers subsequently agreed to use only the FDA-recommended 20,000-ppm soak when sprout production resumed.

Reported by: J Mohle-Boetani, MD, B Werner, MD, M Polumbo, PhD, J Farrar, DVM, D Vugia, MD, Acting State Epidemiologist, California Dept of Health Svcs. S Anderson, K Komatsu, MPH, K Tagg, N Peterson, MD, State Epidemiologist, Arizona Dept of Health Svcs. I Painter, DVM, S Van Dunn, MA, Div of Bacterial and Mycotic Diseases, National Center for Infectious Diseases; K Winthrop, MD, M Beatty, MD, EIS officers, CDC.

Editorial Note: S. Kottbus is a rarely reported cause of salmonellosis in the United States. During 1968-1998, a median of 42 S. Kottbus isolates were reported each year to CDC through the Public Health Laboratory Information System (3). This was the second outbreak of S. Kottbus since 1985 and the first outbreak associated with sprouts.

Since 1995, 15 outbreaks of Salmonella spp. and two outbreaks of Escherichia coli O157:H7 infections associated with sprouts have been reported to CDC. Despite public health advisories about the risks for eating raw sprouts, persons at high risk for systemic infection continue to eat sprouts (4). Two of the patients in this outbreak were immuno-compromised, and one was a young child. In each case, persons perceived raw sprouts as a "healthy" food item.

Sprouts may be contaminated during seed production, germination, sprout processing, or consumer handling and preparation (5,6). On the farm, sprouts seeds may become contaminated through the use of untreated agricultural water, improperly composted manure as fertilizer, excretion from domestic or wild animals, runoff from domesticated animal production facilities, or improperly cleaned harvesting or processing machines (5,6). The association of specific seed lots with illness suggests that seeds are the most likely source for this and most other sprout-related outbreaks (4). Conditions suitable for seed sprouting also are ideal for increasing pathogenic bacterial counts by several logs.

The use of a 20,000-ppm calcium hypochlorite soak before sprouting might reduce the risk for sprout-related illness (4). However, use of this high-dose soak is not completely effective, and outbreaks continue to occur (7). Cracks and crevasses in the sprout seed may trap pathogenic bacteria, making them inaccessible to lethal concentrations of disinfectants (5). Because >20,000-ppm calcium hypochlorite soaks can impair seed germination (5), alternative methods are needed to reduce the risk for human disease following sprout consumption. In this outbreak, some of the implicated sprouts were from seeds that had undergone a combination of heat treatment and a 15-minute, low-dose calcium hypochlorite soak (2,000 ppm). The subsequent outbreak suggests that this hybrid technique using a heat treatment combined with a low-dose hypochlorite solution might not reduce adequately pathogenic bacterial colony counts in alfalfa seeds. Reducing pathogenic bacterial counts on seed during production and harvest could improve the effectiveness of postharvest decontamination.