In mid-1997, fecal specimens from 1,085 healthy volunteers were collected and frozen at -70°C as baseline specimens for a water quality study in Melbourne, Australia. These previously described (9) volunteers were from 294 families with young children who lived in a lower to middle-class suburb. They were specifically chosen because they were representative of young Australian families in eating habits and medical care. Thus, the elderly, the unmarried, and the very poor were not represented. Study participants had no history of diarrhea or antibiotic use at the time of specimen collection.

For VRE screening, frozen samples were thawed and spread onto enterococcosal agar (BBL Microbiology Systems, Cockeysville, MD) containing 6 µg/ml vancomycin and incubating at 35°C for 48-72 hours. Each sample was also incubated in enterococcosal broth (BBL) without vancomycin for 48 hours, prior being spread onto enterococcosal agar containing 6 µg/ml vancomycin. From esculin-positive colonies, three of each morphologic appearance were selected from the direct agar and the subcultured enterococcosal broth cultures of each specimen and provisionally identified as either Enterococcus faecium or E. faecalis by routine criteria (gram-positive cocci, L-pyrrolidonyl-β-naphthylamide hydrolase-positive, nonmotile, catalase-negative, and pigment-negative) (10,11). All isolates fulfilling these criteria were assessed for susceptibility to vancomycin and teicoplanin by the E-test method (AB Biodisk, Dalvagen, Sweden). Isolates with an MIC to vancomycin ≥2 µg/ml were analyzed for the presence of vanA, B, C1, or C2/3 genes by polymerase chain reaction; if vanA- or vanB-positive, they were then confirmed as either E. faecium or E. faecalis by automated biochemical analysis (BioMerieux Vitek Inc., MO) and by polymerase chain reaction, and were assessed by pulsed-field gel electrophoresis (PFGE) (8,12,13).

To be certain that this analysis of frozen specimens was a valid assessment of the rate of fecal VRE carriage in this population, two additional studies were conducted. First, to verify that -70°C storage did not affect the overall recovery of enterococci, a randomly selected subset of 112 of the 1,085 specimens were cultured on enterococcosal agar (BBL) without antibiotic and assessed for enterococci by routine identification techniques (10,11). In addition, to assess whether -70°C storage could selectively impair recovery of VRE, fresh fecal specimens were spiked with concentrations (10⁴-10⁸/ml) of isolates of E. faecalis (vanA, n=1, vanB, n=2) or E. faecium (vanA, n=2; vanB, n=2), then stored at -70°C for 2 weeks before being thawed and cultured.

Enterococci were identified in 106 of the subset of 112 specimens, a rate consistent with previous reports (14). VRE were also consistently recovered from our frozen spiked specimens. These findings suggest that -70°C storage of specimens does not affect the recovery of enterococci in general, or VRE in particular.

In the community study, VRE was isolated from 2 (0.2%) of the 1,085 (95% confidence intervals 0%-0.4%) specimens. Both isolates were vanB E. faecium and were detected in both agar and broth cultures. E-test MIC results at 24 hours for both isolates were 12 µg/ml for vancomycin and 0.125 µg/ml for teicoplanin. Although they were identical by PFGE, they displayed a PFGE pattern that is uncommon among VRE isolates from Melbourne hospitals. The two participants who were vanB E.faecium positive were not related. One was a 34-year-old man and the other a 30-year-old woman; both were married, and each had two children. None of the other family members had detectable fecal VRE colonization. The man had a history of stable ulcerative colitis and was receiving sulfasalazine.

This study suggests that fecal colonization with VRE is present but uncommon in Australia. Despite widespread agricultural use of avoparcin and high community rates of antibiotic use, vanA E. faecium was not identified. vanB E. faecium, which is the most common clinical VRE strain in Australia (8) and whose nosocomial transmission remains a concern (15), was identified in healthy Australians.