Discrepancies in Vancomycin Clearance Estimation Among Renal Function Models in Patients With Low Serum Creatinine Levels: A Pilot Study.

Vancomycin dosing is guided by population pharmacokinetic models that consider renal function. However, low serum creatinine (Scr) levels may reflect heterogeneous conditions and lead to inaccurate vancomycin clearance. This study evaluated the predictive ability of different population pharmacokinetic models for vancomycin clearance in low Scr settings. This single-centre retrospective study included hospitalised patients aged ≥18 years who received intravenous vancomycin. Predicted vancomycin clearance was calculated using population pharmacokinetic models across six scenarios incorporating creatinine clearance (Ccr), serum albumin, reduced muscle mass, Scr rounding and Japanese cystatin C-based estimated glomerular filtration rate (eGFRJapanese, Cys-C). Agreement was evaluated using concordance correlation coefficients (CCC) and Bland-Altman analysis. Among the six scenarios (n = 24), the Ccr-based model exhibited the best agreement (CCC = 0.79). Models based on eGFRJapanese, Cys-C or rounded-up Scr systematically underestimated clearance under augmented renal clearance. In those with a body mass index ≥25 kg/m2, clearance increased more substantially, and a similar trend was observed in patients aged <70 years. In patients with low Scr, Ccr-based models showed better agreement with two-point estimated vancomycin clearance. In obese or younger patients with low Scr, two-point estimated vancomycin clearance may be a practical approach for vancomycin dosing.