Utility of the Berlin Initiative Study-1 equation for the prediction of serum vancomycin concentration in elderly patients aged 75 years and older.

Abstract

Accurate assessment of renal function is essential for determining serum vancomycin (VCM) concentration. Creatinine clearance (Ccr)-calculated using the Cockcroft and Gault (CG) equation-can be used to evaluate renal function for determining VCM dosage. However, Ccr-based evaluation may not be an accurate representation of the renal function in the elderly. Herein, we examine the effectiveness of estimated glomerular filtration rate (eGFR) calculated using the Berlin Initiative Study-1 (BIS1) equation, for predicting the serum VCM concentration. Herein, we retrospectively analyzed patients (aged ≥ 75 years) who had received VCM. Serum VCM concentration was predicted based on Ccr and eGFR. eGFR was calculated using the Japanese equation for eGFR (eGFR_JAP), Modification of Diet in Renal Disease (MDRD) equation (eGFR_MDRD), chronic kidney disease epidemiology collaboration (CKD-EPI) equation (eGFR_CKD-EPI), and BIS1 equation (eGFR_BIS1). The predicted serum VCM concentration was compared with the measured values. Prediction bias, accuracy, and precision were evaluated by calculating the mean prediction error (ME), mean absolute prediction error (MAE), and root mean squared prediction error (RMSE). Our results showed that the ME between the measured and the predicted values calculated using Ccr and each eGFR was the largest and smallest when calculated based on Ccr and eGFR_MDRD, respectively. MAE and RMSE were the largest and smallest when calculated based on Ccr and eGFR_BIS1, respectively. A significant difference was observed in the MAE associated with eGFR_JAP, eGFR_MDRD, and eGFR_CKD-EPI compared to that associated with eGFR_BIS1. In conclusion, our results suggest that the BIS1 equation might be useful for determining the VCM dosage in the elderly.