Purpose: The feasibility of using a limited-sampling algorithm for administration of vancomycin for treatment of vascular-access-related bacteremia in outpatient high flux hemodialysis was investigated.
Methods: The original vancomycin-dosing algorithm used at our hemodialysis unit required stat orders for serum vancomycin concentrations before each hemodialysis session to determine the dose of vancomycin to be administered posthemodialysis. Vancomycin concentration data obtained using this original algorithm from January through September 2001 were retrospectively analyzed to determine how many vancomycin concentrations measured 5-20 microg/mL and identify potential clinical predictors of vancomycin removal.
Results: A total of 409 serum vancomycin concentrations were drawn during the study period. Ninety-seven percent of concentrations drawn were within 5-20 microg/mL. Twenty-eight patients had data evaluable to determine pharmacokinetic parameters. Mean +/- S.D. vancomycin removal was 39% +/- 13%. Body weight and duration of dialysis alone, blood flow rate, and dialysate flow rate were not predictive of vancomycin removal. Based on these data, a revised algorithm with limited vancomycin sampling data was initiated in December 2002. Retrospective analysis of concentrations obtained and achieved by this algorithm demonstrated a 70% reduction in the number of vancomycin concentration determinations, with 93% of these concentrations within 5-20 microg/mL. The estimated annual cost saving to the hemodialysis unit with the revised algorithm was 7552 dollars.
Conclusion: A vancomycin-dosing algorithm using limited concentration monitoring for hemodialysis patients achieved comparable vancomycin concentrations to those found with more frequent monitoring and resulted in significant cost savings.