BACKGROUND: Colonization and infection with vancomycin-resistant Enterococcus faecium (VREF) has been associated with the use of vancomycin and other antibiotics in individual patients. The objective of this study was to determine the association of VREF with the aggregate usage of antibiotics on nursing units in a hospital. METHODS: This was a retrospective correlation study. A usage ratio was calculated for each parenteral antibiotic on each nursing unit as the per-bed usage by weight of that antibiotic divided by its average usage throughout the hospital. An average usage ratio (AUR) for each nursing unit was calculated as the mean of usage ratios of individual antibiotics. The AUR was used to compare the usage of antibiotics among nursing units in the hospital. The incidence of VREF infections on individual nursing units in a Veterans Affairs Medical Center was correlated with the usage of parenteral antibiotics separately and in aggregate in univariate and multivariate regression analyses. RESULTS: The AUR was strongly and positively correlated with the recovery of VREF on individual nursing units. By univariate analyses, increasing use of each antibiotic tested was associated with isolation of VREF but only clindamycin remained significant in the multivariate model. However, usage of various antibiotics was highly interrelated, and only clindamycin usage was significantly correlated with usage of all other antibiotics studied. Intensive care and acute care units and units with fewer patient beds were more likely to have patients with VREF infection than were subacute care units (p < 0.003) or larger units (p < 0.01). CONCLUSIONS: VREF infections were associated with greater aggregate antibiotic use on nursing units. Determination of antibiotic usage ratios may provide a convenient and useful tool for examining the association of antibiotic usage with other nosocomial infections.
BACKGROUND: Colonization and infection with vancomycin-resistant Enterococcus faecium (VREF) has been associated with the use of vancomycin and other antibiotics in individual patients. The objective of this study was to determine the association of VREF with the aggregate usage of antibiotics on nursing units in a hospital. METHODS: This was a retrospective correlation study. A usage ratio was calculated for each parenteral antibiotic on each nursing unit as the per-bed usage by weight of that antibiotic divided by its average usage throughout the hospital. An average usage ratio (AUR) for each nursing unit was calculated as the mean of usage ratios of individual antibiotics. The AUR was used to compare the usage of antibiotics among nursing units in the hospital. The incidence of VREF infections on individual nursing units in a Veterans Affairs Medical Center was correlated with the usage of parenteral antibiotics separately and in aggregate in univariate and multivariate regression analyses. RESULTS: The AUR was strongly and positively correlated with the recovery of VREF on individual nursing units. By univariate analyses, increasing use of each antibiotic tested was associated with isolation of VREF but only clindamycin remained significant in the multivariate model. However, usage of various antibiotics was highly interrelated, and only clindamycin usage was significantly correlated with usage of all other antibiotics studied. Intensive care and acute care units and units with fewer patient beds were more likely to have patients with VREF infection than were subacute care units (p < 0.003) or larger units (p < 0.01). CONCLUSIONS:VREF infections were associated with greater aggregate antibiotic use on nursing units. Determination of antibiotic usage ratios may provide a convenient and useful tool for examining the association of antibiotic usage with other nosocomial infections.