Vancomycin Prescribing Practices in Hospitalized Chronic Hemodialysis Patients Kevin Green, MD, Gerald Schulman, MD, David W. Haas, MD, William Schaffner, MD, and Erika M.C. D’Agata, MD, MPH ● To determine the prevalence of and indications for vancomycin administration among hospitalized chronic hemodialysis patients, we performed a 3-month prospective cohort study at a tertiary care center. Modified guidelines for vancomycin use from the Hospital Infections Control Practices Advisory Committee of the Centers for Disease Control and Prevention were used. Vancomycin was administered during 56 of 144 admissions (39%) requiring chronic hemodialysis compared with 336 of 7,212 admissions (5%) not requiring hemodialysis (relative risk, 11; 95% confidence interval, 8 to 15; P F 0.001). Among chronic hemodialysis patients, vancomycin use was judged appropriate for 131 of the 164 vancomycin doses (80%). The most common appropriate indication was empiric therapy in a febrile patient before culture or susceptibility results. Of 32 infections identified in patients who received empiric vancomycin, 15 infections (47%) were caused by -lactam–resistant pathogens. Among the 33 doses (20%) judged inappropriate, continued therapy for a presumed infection despite failure to identify a -lactam–resistant pathogen was the most common indication. Although vancomycin administration was frequent among hospitalized chronic hemodialysis patients, its use was justified in the majority of cases. Efforts should focus on limiting vancomycin administration for treating infections caused by -lactam–sensitive pathogens.  2000 by the National Kidney Foundation, Inc. INDEX WORDS: Vancomycin; enterococcus; hemodialysis; hospital; chronic; antibiotic. T HE INCIDENCE of infections caused by vancomycin-resistant enterococci (VRE) is increasing in the dialysis population. In 1995, 11% of the hemodialysis centers in the United States reported at least one patient colonized or infected with VRE. By 1996, the number of centers reporting VRE had doubled to 21%. 1 Because identified risk factors for VRE coloniza- tion and/or infection include severity of underly- ing illness, frequent hospitalizations, and antibi- otic exposure, particularly to vancomycin, 2-5 the increase in VRE cases among dialysis patients was inevitable. Vancomycin exposure is a potentially modifi- able risk factor for VRE acquisition. To deter- mine the indications for vancomycin administra- tion, we performed a cohort study among hospitalized patients requiring chronic hemodi- alysis. We sought to identify the frequency and reasons for inappropriate administration of van- comycin based on modified guidelines of the Hospital Infections Control Practices Advisory Committee (HICPAC) of the Centers for Disease Control and Prevention. 6 METHODS Vanderbilt University Medical Center (Nashville, TN) is a 663-bed tertiary care hospital. Approximately 14,000 pa- tients are admitted per year, of whom an average of 400 (3%) require chronic hemodialysis. From August 8 through November 11, 1998, a prospective cohort study was performed on all patients admitted to Vanderbilt University Medical Center who required chronic hemodialysis. Approval for this study was obtained from the institutional review board. A computerized pharmacy data- base was used to obtain information on vancomycin expo- sure for all patients admitted to the medical center. Chronic hemodialysis was defined as dialysis initiated before the current hospitalization. Medical records for pa- tients receiving chronic hemodialysis were reviewed for demographic and clinical data. Indications for vancomycin use were classified as appro- priate or inappropriate based on modified HICPAC criteria. 6 Because of a high proportion of methicillin-resistant Staphy- lococcus aureus isolates (37%) at our institution and fre- quent infections caused by coagulase-negative staphylo- cocci among the dialysis population, we modified the HICPAC criteria to include as appropriate the empiric admin- istration of a single dose of vancomycin in a febrile chronic hemodialysis patient pending culture results or pending susceptibility data if gram-positive cocci in clusters were identified. Subsequent vancomycin doses administered to the same patient were considered inappropriate if a -lactam– resistant gram-positive pathogen was not identified, per HICPAC guidelines. Infections were classified according to From the Department of Medicine, Divisions of Infectious Diseases and Nephrology, and the Departments of Microbi- ology and Immunology and Preventive Medicine, Vanderbilt University School of Medicine, Nashville, TN. Received April 30, 1999; accepted in revised form August 10, 1999. Address reprint requests to Erika M.C. D’Agata, MD, Vanderbilt University Medical Center, Division of Infectious Diseases, Medical Center North A-3310, 21st and Garland St, Nashville, TN 37232. E-mail: erika.d’agata@mcmail. vanderbilt.edu  2000 by the National Kidney Foundation, Inc. 0272-6386/00/3501-0011$3.00/0 64 American Journal of Kidney Diseases, Vol 35, No 1 (January), 2000: pp 64-68