Abstract
Objective: To assess the clinical factors affecting the reliability of a pharmacokinetic simulation model for vancomycin (VCM) in critically ill children. Materials and methods: We enrolled 84 patients admitted to our pediatric intensive care unit and treated with VCM between March 2010 and November 2011. The mean pediatric logistic organ dysfunction (PELOD) score was 10.7±10.6. The prediction error was calculated from the observed and predicted concentrations by using the pharmacokinetic simulation software. We analyzed which clinical factor affected the prediction error by using the univariate analysis. Results: The mean prediction error was −16±64%. We found that the PELOD score, percent fluid overload (%FO), and estimated creatinine clearance (estCCr) affected the prediction error. The prediction error was higher in patients whose PELOD score was ≥20 (12.4±68.0%, P < 0.05), %FO was ≥10 (13.1±42.7%, P < 0.05), and estCCr was < 50 (−44.8±80.6%, P < 0.05). Conclusion: The PELOD score, %FO, and estCCr affected the reliability of the pharmacokinetic simulation model for VCM in critically ill children.
