Is Trough Concentration of Vancomycin Predictive of the Area Under the Curve? A Commentary
Recently, Bel Kamel et al1 presented work in this journal concerning the correlation between vancomycin trough levels (Cmin) and area under the concentration–time curve over 24 hours (AUC24) in elderly patients (https://paperpile.com/c/EsidMG/fy4J). In their report, they convincingly spell out the merits of dosing patients based on a specific target AUC rather than Cmin. We agree with the authors that Bayesian dose adaptation software could and should be used to facilitate the calculation of AUC and render nontrough samples useful.
The main result discussed in the article by Bel Kamel et al is the correlation between measured trough concentrations and calculated AUC values in a dataset of elderly patients, which we feel is a valuable addition to the literature. In the final part of their statistical analysis of this dataset, however, the authors construct a receiver operating characteristic curve using the vancomycin trough concentration as a predictor for the classification of patients between AUC24 ≥ 400 and AUC <400 (mg∣h−1∣L−1). The authors argue that, based on their dataset and analysis, a concentration of 10.8 mg/L is the “optimal” trough concentration to decide whether a patient is above or below an AUC24 of 400. This finding is presented as one of the main conclusions of the article. Although we agree with most of the analysis and conclusions in their article, we respectfully object to the use of this classification rule and the selection of Cmin of 10.