ABSTRACT
<p><b>INTRODUCTION</b>To assess the efficacy of screening stools sent for Clostridium difficile cytotoxin assay (CDTA) for surveillance of vancomycin-resistant enterococci (VRE).</p><p><b>MATERIALS AND METHODS</b>From April to May 2005, all stools submitted for CDTA were also cultured for VRE using vancomycin containing culture media. Isolates were identified to species level and vancomycin resistance confirmed, followed by polymerase chain reaction (PCR) for detection of vancomycin resistance genes and DNA fingerprinting. Over 2 consecutive days during that period, stool specimens or rectal swabs were also obtained from all patients in high-risk units (haematology, oncology, renal and intensive care). Fifty-one patients in each group were compared in terms of VRE risk factors previously identified.</p><p><b>RESULTS AND DISCUSSION</b>The prevalence of VRE in both groups was similar [3/204 (1.5%) in the CDTA arm and 1/97 (1.0%) in the high-risk arm; P = 1.0, Fisher's exact test]. Prevalence of risk factors for VRE colonisation, including age, duration of hospitalisation, exposure to antibiotics, exposure to surgical procedures, presence of malignancy and diabetes mellitus was similar in both groups (P > 0.05). Only renal failure (P < 0.05) was more common in the high-risk group. All 4 isolates of VRE identified were genetically distinct by variable number tandem repeat (VNTR) typing; 3 were Enterococcus faecium (2 with the vanB gene, 1 with vanA) and one E. faecalis.</p><p><b>CONCLUSION</b>Less than 2% of our high-risk patients are VRE carriers. In-hospital VRE screening using stools sent for CDTA is a simple, reasonable surrogate for screening individual high-risk patients as the patient risk profile is similar and the yield comparable in a low-prevalence setting.</p>