We compared three isolates of S. Enteritidis recovered from Salmonella-based rodenticides. Two isolates (of the Biorat strain) were from Biorat samples collected in 1995 and 2001. The label from the Biorat product obtained in 2001 states that Biorat contains 1.25% “monopathogenic” Salmonella and 0.02% hydroxycoumarin. Pooled samples of the 1995 Biorat product yielded 1x10⁸ CFU S. Enteritidis per gram of Biorat granules, and the 2001 product yielded 200,000 CFU per gram. A third isolate, S. Enteritidis var. Danysz (of the Ratin strain) was recovered from Ratin by the Danish Veterinary Laboratory in the 1920s or early 1930s.

Isolates were serotyped and biochemically characterized (8), subtyped by pulsed-field gel electrophoresis (PFGE) by using the restriction enzymes XbaI and BlnI (9), and phage-typed. Phage-typing was performed at the National Laboratory for Enteric Pathogens, National Microbiology Laboratory, Canadian Science Centre for Human and Animal Health.

Figure 1

Figure 1. Pulsed-field gel electrophoresis of Salmonella enterica serotype Enteritidis isolates from Biorat and Ratin products using XbaI (lanes 2–5) and BlnI (lanes 7–10). Lanes 1, 6, and 11, molecular weight standard strain...

Both the Biorat strains and the Ratin strain were identified as S. Enteritidis, phage type (PT) 6a. The two Biorat strains were indistinguishable from each other by PFGE with the restriction enzymes XbaI and BlnI. PFGE patterns of the Ratin strain differed from those of the Biorat strain by three bands with XbaI and by five bands with BlnI (Figure 1).

Neither the Biorat strains nor the Ratin strain was shown to decarboxylate lysine. By contrast, four S. Enteritidis PT 6a isolates in the CDC culture collection from sources unrelated to Biorat or Ratin were positive for lysine decarboxylase. Threlfall et al. reported that the strain of S. Enteritidis PT 6a from Biorat is indistinguishable from those of the Ratin and Liverpool rodenticide strains by plasmid profile typing (5), as both strains contained plasmids of approximately 59, 4.0, and 3.0 MDa.

Since the mid 1980s, S. Enteritidis has caused a global pandemic of foodborne illness associated with eggs and poultry as a result of infection of the internal organs of chicken (10). Because of this pandemic, S. Enteritidis has become the most common serotype of Salmonella isolated from humans worldwide (11).

The S. Enteritidis strain found in Biorat is similar to the strain found in Ratin, a discontinued European product that caused human illness. Both strains were the same phage type, were indistinguishable by plasmid profile typing, and were different from 97% of salmonellae (8) in that they did not decarboxylate lysine. Although differences were noted between PFGE patterns of the Biorat strain and the Ratin strain, the similarities suggest that they may have originated from a common strain. Because the strains are similar and no evidence shows that the Biorat strain has decreased virulence, the Biorat strain is likely as pathogenic to humans as the Ratin strain.

In a retrospective study in Denmark from 1926 through 1956, Martin Kristensen identified 122 patients infected with the Ratin strain (12 and unpub. data), including 5 (4%) who died, 3 of whom were children. Twenty-two (18%) of 122 patients were reported to have eaten food items contaminated with Ratin, while 43 (35%) had handled the rodenticide. In 1956, Taylor (13) also reported several outbreaks of food poisoning associated with S. Enteritidis rodenticides and concluded that the use of bacterial rodenticides should be stopped.

Rodenticides containing salmonellae were evaluated during a plague outbreak in San Francisco in 1895 (2); they were found to have no definable impact on the rodent population, but they caused illness and death in humans who prepared and handled them. In 1921, Willfuhr and Wendtland (14) reported several outbreaks of human Salmonella infections from rodenticides. In one of these outbreaks, Russian prisoners of war who ate a large number of Ratin potato baits became ill, and two died. In another outbreak in 1918, two persons died and approximately 35 became ill after eating a cake that had been intentionally contaminated with Ratin (14,15). From 1920 through 1940, other outbreaks associated with Salmonella-based rodenticides were reported, and several of these outbreaks included deaths (15–18).

The Biorat product insert, as well as information available on the Internet (6,7), claims that the product is not harmful to humans and does not contaminate the environment. Recent newspaper articles have generated interest in using Salmonella-based rodenticides as an alternative to chemical rodenticides. Advocates for the use of Biorat claim, “[Biorat] has absolutely no secondary effects on other animals, on the environment, or on humans…. It contains a strain of Salmonella that only affects rats” (19).

S. Enteritidis causes severe diarrheal illness, which can be life-threatening, especially among children, the elderly, and immunocompromised persons. We have not identified any peer-reviewed, scientific data on the safety of Salmonella-based rodenticides, and to our knowledge, all strains of S. Enteritidis are capable of causing human illness. Noting the hazards of Salmonella-based rodenticides, many countries have banned their use, and WHO has repeatedly recommended against use of salmonellae in rodenticides.

Figure 2

Figure 2. A sample of rodenticide that resembles grains of rice but contains pathogenic Salmonella enterica serotype Enteritidis.

Current concerns about bioterrorism suggest an additional public health threat posed by a commercially available strain of S. Enteritidis. Salmonella-based rodenticides have already been used intentionally to cause human illness (14,15); however, human illness may more commonly be caused by inadvertent exposure to Salmonella-based rodenticides. These rodenticides are generally mixed with grains to form baits (Figure 2). Biorat, for example, is made with whole rice and can easily be mistaken for food. Ingesting a few grams of bait, with at least 200,000 CFU per gram, could easily cause a severe case of salmonellosis.

To determine why Salmonella-based rodenticides are still used despite information about the public health hazards, we conducted a literature search with the keywords “Salmonella” and “rodenticide” (National Library of Medicine, http://www.ncbi.nlm.nih.gov). We found 10 articles, in addition to the recent public health publications discussed above (2,5); none addressed the public health hazards of Salmonella-based rodenticides. Many of the reference materials we used to prepare the present article were not written in English or were not retrievable from current electronic databases. The continued use of Salmonella-based rodenticides may likely be related to the fact that the content of important but dated scientific papers is unlikely to be known to current decision-makers.

Salmonella-based rodenticides may contain an approved rodenticide, such as warfarin, in concentrations high enough to kill rats, and the addition of S. Enteritidis has not been shown to increase the effectiveness of the poison (2). Extensive use of Salmonella-based rodenticides in the past may have increased the prevalence of Salmonella in rodents (1) and consequently increased the potential for human salmonellosis by transmission from rodents to food or food animals. Unfortunately, a misperception exists that some strains of S. Enteritidis are not pathogenic to humans. We recommend informing rodent-control authorities and the public that S. Enteritidis is a known human pathogen and that use of Salmonella-based rodenticides has had severe public health consequences. Effective and safe alternatives to Salmonella-based rodenticides are available worldwide.