Multistate Outbreak of Human Salmonella Typhimurium Infections Associated with Aquatic Frogs --- United States, 2009

Multistate Outbreak of Human Salmonella Typhimurium Infections Associated with Aquatic Frogs --- United States, 2009

During April--July 2009, the Utah Department of Health identified five cases of Salmonella Typhimurium infection with indistinguishable pulsed-field gel electrophoresis (PFGE) patterns, predominantly among children. In August, CDC began a multistate outbreak investigation to determine the source of the infections. This report summarizes the results of this ongoing investigation, which, as of December 30, had identified 85 S. Typhimurium human isolates with the outbreak strain from 31 states. In a multistate case-control study, exposure to frogs was found to be significantly associated with illness (63% of cases versus 3% of controls; matched odds ratio [mOR] = 24.4). Among 14 case-patients who knew the type of frog, all had exposure to an exclusively aquatic frog species, the African dwarf frog. Environmental samples from aquariums containing aquatic frogs in four homes of case-patients yielded S. Typhimurium isolates matching the outbreak strain. Preliminary traceback information has indicated these frogs likely came from the same breeder in California. Reptiles (e.g., turtles) and amphibians (e.g., frogs) have long been recognized as Salmonella carriers (1,2), and three multistate outbreaks of human Salmonella infections associated with turtle contact have occurred since 2006 (3,4). However, this is the first reported multistate outbreak of Salmonella infections associated with amphibians. Educational materials aimed at preventing salmonellosis from contact with reptiles should be expanded to include amphibians, such as aquatic frogs.

The five cases identified in July 2009 by the Utah Department of Health all had isolates indistinguishable by pulsed field gel electrophoresis and were identified with XbaI pattern JPXX01.0177. The cases had occurred during April--July. On September 29, PulseNet, the national molecular subtyping network for foodborne disease surveillance, identified a national increase of isolates with this PFGE pattern (37 isolates from 19 states in 60 days). Multiple-locus variable-number tandem repeat analysis (MLVA) provided additional discrimination of the outbreak strain. For this investigation, a case was defined as S. Typhimurium infection with illness onset on or after April 1, 2009, with 1) PFGE pattern indistinguishable from the cluster-defining pattern and 2) MLVA pattern either matching that of the main outbreak strain, or MLVA unknown.

The multistate investigation began with in-depth, open-ended interviews of salmonellosis patients regarding exposures in the week before illness onset. A total of 11 interviews with patients were conducted through November. All 11 persons reported consumption of cheese-flavored crackers; eight reported exposure to aquatic animals, including fish and aquatic frogs.

As of December 30, 2009, S. Typhimurium isolates with the outbreak strain had been identified in 85 patients from 31 states, extending from Massachusetts to California, with week of illness onset ranging from March 22 to November 29 (Figure 1). Among the patients, 52% were male; median age was 5 years (range: 3 weeks--54 years), and 79% were aged <10 years. Among 47 patients with outcome information available, 16 (34%) had been hospitalized; no deaths were reported.

Case-Control Study

To examine possible associations between illness and consumption of cheese crackers and exposure to aquatic pets, CDC conducted a nationwide case-control study during November 30--December 7. Patients infected with S. Typhimurium with the outbreak strain who had specimen collection dates after July 15 were enrolled. Controls were persons with recent infection of Salmonella strains other than the outbreak strain and matched to case-patients by age and county of residence. Exposure histories were collected for 7 days before illness onset for case-patients and for 7 days before interview for controls.

Investigators sought to match each case-patient with two controls. A total of 19 case-patients (18 with stool specimens and one with a urine specimen) and 31 matching controls were enrolled from 15 states. Case-patients were found to be significantly more likely than controls to have had exposure to an aquatic pet, including fish and frogs (74% of case-patients versus 35% of controls; mOR = 4.7 [95% confidence interval (CI) = 1.2--27.0]). More specifically, illness was found to be associated with exposure to frogs (63% of case-patients versus 3% of controls; mOR = 24.4 [CI = 4.0--infinity]). Exposure to fish was not statistically significant (58% of case-patients versus 29% of controls, mOR = 3.1 [CI = 0.8--14.2]). No association was found between illness and consumption of any food item, including cheese crackers.

Among 39 patients interviewed as of December 9, including some of the 19 case-patients enrolled in the case-control study, 14 knew the type of frog involved in their exposure, and all 14 identified the frog as an African dwarf frog (Figure 2). When asked about potential for Salmonella infection, 19 of 36 (53%) patients reported awareness of association between contact with reptiles and Salmonella infection, but only 11 of 36 (31%) reported awareness of association with amphibians.* Among 20 patients from whom the information was available, the frog's aquarium was cleaned in the kitchen sink in the homes of six persons (30%) and in the bathroom sink in the homes of seven others (35%).

abrir aquí para acceder al documento MMWR CDC completo, del que se reproduce un 10%:
Multistate Outbreak of Human Salmonella Typhimurium Infections Associated with Aquatic Frogs --- United States, 2009