Household contamination with Salmonella enterica - 1 - Dispatches

Emerging Infectious Diseases, Jan, 2003 by Daniel H. Rice, Dale D. Hancock, Paivi M. Roozen, Maryanne H. Szymanski, Beth C. Scheenstra, Kirsten M. Cady, Thomas E. Besser, Paul A. Chudek

Household contamination with Salmonella enterica increases when occupational exposure exists (cattle farms with known salmonellosis in cattle, a salmonella research laboratory, or a veterinary clinic experiencing an outbreak of salmonellosis). Fifteen of 55 (27.2%) vacuum cleaner bags from households with occupational exposure to S. enterica were positive versus 1 of 24 (4.2%) without known exposure. Use of a carpet cleaner and several cleaners/disinfectants reduced, but failed to eliminate, S. enterica from artificially contaminated carpet.

**********

Although most cases of nontyphoid salmonellosis in humans are foodborne, a significant number appear to be acquired from households contaminated with Salmonella enterica (1-3). Sources and sites of contamination include house members with clinical disease, pets with sub-clinical infection, contaminated items brought into the home, toilet bowls, carpet, floors, refrigerators, and kitchen sinks and counter tops (1-6). Culture of vacuum cleaner bag contents has been used as a tool to screen households for contamination with S. enterica (1-3). The purpose of the present study was to determine the frequency of contamination with S. enterica, as indicated by culture of vacuum cleaner bag contents, in homes in which the residents had differing levels of occupational exposure.

The contents of vacuum cleaner bags (N=79 bags), collected from household vacuum cleaners, were cultured from five groups: 1) occupants had no known exposure to livestock or S. enterica in the workplace (n=12), 2) one or more occupants had direct contact with livestock with no known recent salmonellosis cases (n=12), 3) one or more occupants had direct contact with cattle salmonellosis cases associated with the serovar Typhimurium (n=26), 4) occupants were exposed to cats involved in a veterinary clinic outbreak of feline salmonellosis associated with the serovar Typhimurium (n=16), and 5) one or more occupants were laboratory or field workers engaged in research on S. enterica (n=13).

Vacuum cleaner bags were stored at room temperature and cultured for S. enterica within one week of collection. The contents of each bag were cultured in duplicate. Twenty-five g of bag contents was added to 225 mL buffered peptone water (BPW, Remel Inc., Lenexa, KS), and incubated overnight at 37[degrees]C. Preenriched samples were mixed, and 1 mL of BPW was transferred to 9 mL of tetrathionate broth (Tet, Remel Inc.), incubated overnight at 37[degrees]C, and then 100 [micro]L of Tet was transferred to 10 mL Rappaport-Vassillladis broth (R10, Difco, Detroit, MI). The Tet tubes were incubated an additional 24 h with the R10 tubes at 37[degrees]C, and then plated onto brilliant green agar containing sulfadiazine (BGS, Difco, Detroit, MI). BGS plates were incubated for 48 h at 37[degrees]C, examined at 24 h and 48 h, and suspect colonies were biochemically screened. All S. enterica isolates were serotyped by the National Veterinary Services Laboratory, Ames, Iowa.

Salmonella organisms from all groups were found in household vacuum cleaner bags, except those from homes in which occupants had no contact with livestock or exposure to S. enterica (Table 1) in the workplace. S. enterica serovar Dublin was found in 1 of 12 (8.3%) vacuum bags collected from households with direct contact with livestock having no known recent cases of salmonellosis. Eight of 26 (30.8%) vacuum bags from households with occupants who had contact with Salmonella-infected cattle were positive. One of the positive vacuum bag samples came from a home in which an infant developed salmonellosis concurrent with an outbreak on the dairy farm where his father was employed. From households where occupants were exposed to an outbreak of feline salmonellosis, 3 of 16 (18.8%) of bags were positive, and from households of personnel engaged in field and laboratory-based research on S. enterica, 4 of 13 (30.8%) bags were positive. All S. enterica isolates from households with known occupational exposure belonged to the serovar Typhimurium; as might be expected, given that all known contact exposures were with this serovar.

Since vacuum cleaners are primarily used to clean floors, the floors were likely the primary site of household contamination in this study. To ascertain the best way to remove S. enterica from carpeted floors (to advise affected persons), we began a study to identify a means of decontaminating carpet that was artificially contaminated with S. enterica. In this experiment, nine carpet segments (40 cm x 80 cm) were attached to separate sections of plywood. Each carpet segment was subdivided into four quadrants and a 15-cm x 25-cm rectangle was marked in each quadrant with indelible ink. Five serovars of S. enterica (Typhimurium, Dublin, Infantis, Heidelberg, and Newport), chosen for their resistance to the antibotics ampicillin, chloramphenicol, and streptomycin were mixed in fresh bovine feces (106 cfu/g). Approximately 500 g of this feces was evenly distributed onto each carpet segment by vigorous rubbing with a sponge mop. Feces-coated carpet segments were allowed to dry overnight at room temperature. Pretreatment samples were collected from the upper right and lower left quadrant of each segment by wiping a sponge (Specisponge, Nasco International Inc., Fort Atkinson, WI), saturated with BPW, over the surface of the carpet (10 times in one direction and 10 times perpendicular to the initial direction). The sponge was placed into a Whirl-pak bag (Nasco International Inc.) containing 25 mL BPW and mixed by using a Stomacher laboratory blender for 60 s. Serial dilutions of the BPW/carpet content suspension were spread onto MacConkey agar plates (Remel Inc.) containing ampicillin (256 [micro]g/mL), chloramphenicol (8 [micro]g/mL), and streptomycin (32 [micro]g/mL) (MacACS) and incubated at 37[degrees]C overnight. Non-lactose-fermenting colonies on MacACS were counted and a subset assayed biochemically and serologically for S. enterica. One carpet segment was used as a control (not cleaned) and the remaining eight segments were cleaned, until free of visible soiling, with a commercial wet-vacuum carpet cleaning system along with the proprietary carpet-cleaning agent. After cleaning, two carpet segments were treated, by using the wet vacuum system, with each of; chlorhexidine (Virosan Bio-Ceutic, Boehringer Ingelheim Vetmedica Inc., St Joseph, MO, 8 oz/ gal), a quaternary ammonium disinfectant (Lysol all purpose cleaner, Reckitt Benckiser Inc., Wayne, NJ, 8 oz/gal) and a phenolic disinfectant (LpH Ag, STERIS Corp., St. Louis, MO, 0.5 oz/gal). Two cleaned segments were not treated with a disinfectant. Carpet segments were allowed to dry overnight at room temperature, after which sponge samples were collected from the upper left and lower right quadrants and cultured as described.