Methods. During September 1998 through August 1999, we obtained weekly surveillance cultures from consenting liver transplant patients and surfaces in their rooms. Pooled handwash specimens from personnel also were obtained. Specimens were processed on selective media to detect VRE, and isolates were typed by pulsed field gel electrophoresis. Information was collected from patient records concerning in-hospital treatment and clinical course.

Results. Serial cultures were obtained during 33 admissions of 29 patients. VRE were detected in initial specimens from 6 admissions, and nosocomial acquisition of VRE occurred in 12 (44%) of the remaining 27 admissions. Seven different strain types of VRE were detected. The initial site of acquisition was stool in all cases; bile became culture-positive in only two patients. Overall, 16 (55%) of the 29 patients became colonized, usually after transplantation. VRE were detected in environmental cultures during 10 admissions and in 2 of 21 pooled handwashes. No statistically significant differences in clinical status or treatment were found when colonized patients were compared to noncolonized controls. The only VRE infection resulted from a choledochojejunostomy anastomotic leak.

Conclusion. Alimentary tract colonization by VRE occurred commonly in liver transplant patients, probably by cross-transmission. The clinical consequences were modest in the patients studied, but colonized transplant patients provide a substantial reservoir for continued VRE transmission in hospitals.