Vancomycin volume of distribution estimation in adults with class III obesity

Abstract

To compare methods of estimating vancomycin volume of distribution (V) in adults with class III obesity. A retrospective, multicenter pharmacokinetic analysis of adults treated with vancomycin and monitored through measurement of peak and trough concentrations was performed. Individual pharmacokinetic parameter estimates were obtained via maximum a posteriori Bayesian analysis. The relationship between V and body weight was assessed using linear regression. Mean bias and root-mean-square error (RMSE) were calculated to assess the precision of multiple methods of estimating V. Of 241 patients included in the study sample, 159 (66.0%) had a body mass index (BMI) of 40.0-49.9 kg/m2, and 82 (34.0%) had a BMI of ≥50.0 kg/m2. The median (5th, 95th percentile) weight of patients was 136 (103, 204) kg, and baseline characteristics were similar between BMI groups. The mean ± S.D. V was lower in patients with a BMI of 40.0-49.9 kg/m2 than in those with a BMI of ≥50.0 kg/m2 (72.4 ± 19.6 L versus 79.3 ± 20.6 L, p = 0.009); however, body size poorly predicted V in regression analyses (R2 < 0.20). A fixed estimate of V (75 L) and use of a weight-based value (0.52 L/kg by total body weight [TBW]) yielded similar bias and error in this population. Results of the largest analysis of vancomycin V in class III obesity to date indicated that use of a fixed V value (75 L) and use of a TBW-based estimate (0.52 L/kg) for estimation of vancomycin V in patients with a BMI of ≥40.0 kg/m2 have similar bias. Two postdistribution vancomycin concentrations are needed to accurately determine patient-specific pharmacokinetic parameters, estimate area under the curve, and improve the precision of vancomycin dosing in this patient population.