Control of glycopeptide-resistant enterococci in an oncology unit

Enterococci are responsible for 10-12% of nosocomial infections. Since 1987 the incidence of glycopeptide-resistant enterococci (GRE; termed vancomycin-resistant enterococci in the United States) has increased dramatically. The mechanism of GRE is well understood and involves mutation of a single terminal amino acid in a peptidoglycan precursor leading to reduced vancomycin affinity. Studies implicated antibiotic selection pressure as a major risk factor for GRE infection and colonization. In the hematology unit of a London hospital, GRE emerged in December 1993, with 38% of patients positive in a point prevalence study. Between December 1993 and June 1995, 13 patients acquired GRE bacteremia, five of whom died. Between 1995 and 1996 a prospective sequential study was undertaken to determine the effect of changing antibiotic treatment of febrile neutropenia (FN). All patients admitted to the hematology unit were enrolled in the study. In phase 1, treatment continued with ceftazidime monotherapy. In phase 2, piperacillin-tazobactam replaced ceftazidime, and enhanced infection control measures were encouraged. In phase 3, therapy returned to ceftazidime. In phase 1, 57% of patients became colonized or infected with GRE. Phase 2, which was divided into two 4-month cohorts, showed a significant reduction in GRE acquisition, falling to 8% colonization and no clinical infections in the second 4 months. In phase 3, despite heightened infection control procedures, the acquisition rate rose to 36%.