기타학회
| Comparison of First Order Pharmacokinetic Equation and Bayesian Approach in Estimating Vancomycin AUC in Critically Ill Patients | ||
|---|---|---|
Department of Pharmacy, Asan Medical Center, South Korea
Purpose As the AUC/MIC ratio was presented as a Vancomycin TDM index, discussions regarding accurate and clinically applicable AUC estimation have dominated research in recent years. In AUC estimation, there is a First order equation method that requires two blood collections (trough and peak concentrations), and a Bayesian approach has recently been introduced which can be analyzed with one blood collection. This study evaluates the clinical applicability of Bayesian approach in adult critically ill patients, by comparing AUC and recommended dose of two methods.
Method Medication profile and blood concentration records were retrospectively collected through electronic medical records of Asan medical center. Targeted patients were those in adult Intensive Care Unit of Asan Medical Center and whose trough and peak concentration samples were collected for Vancomycin monitoring. For estimating Vancomycin AUC with First order equation method, both trough and peak concentrations were used. Bayesian approach only used trough levels.
Results Estimated AUC with First order equation and that of Bayesian method were compared. Recommended dose was aimed to reach AUC 400-600 mg•h/L, assuming all bacterial MIC was 1 mg/L. With patients whose second blood concentrations were collected after the first monitoring, estimations from first sampling and actual values of second sampling were compared. Compared indicators were trough level and AUC. Patients with continuous renal replacement therapy and those without RRT were analyzed separately.
Conclusion For critically ill patients without renal replacement therapy, this study corroborates Bayesian approach using trough level is applicable in estimating Vancomycin AUC/MIC ratio. In estimating figures of patients with CRRT, however, some values were dramatically deviated from the estimations. The data suggests that it is necessary to verify that the Bayesian population model of critically ill patients with RRT can represent adult ICU patients of each medical center. |
||
