Methods: This was a retrospective, opportunistic cohort study conducted at two pediatric hospitals in patients receiving vancomycin from 2003 to 2015. ARC was defined as a CLVANCO of ≥ 130 mL/min/1.73m2. Using NONMEM 7.2, a one-compartmental model with first-order kinetics and Bayesian analysis was used to estimate CLVANCO. Internal model validation was performed using the bootstrap technique.
Results: ARC was identified in 29 of 250 (12%) total subjects with 658 vancomycin serum concentrations. Male gender and weight were similar between these groups, at 49% and 30 (interquartile range [IQR] 15-50) kg, respectively. Median age and baseline serum creatinine (SCr) were different between those with and without ARC (11.3 [IQR 8.7-13.8]) vs 9.0 [3.0-14.2] yr, p=0.037 and 0.33 [0.30-0.40] vs 0.40 [0.30-0.60] mg/dL, p=0.013, respectively). The final model for CLVANCO (L/hr) was 0.118 * Weight (e (-1.13*(SCr – 0.40)). Mean CLVANCO in those with vs without ARC were 144 and 90 mL/min/1.73m2 (p < 0.001). The CL was similar to the median bootstrap analysis values, and was within the 95% confidence intervals. CLVANCO was weakly correlated to the glomerular filtration rate estimated by the Modified Schwartz method (Spearman R2= 0.083).
Conclusion: ARC was identified in one of ten critically-ill pediatric patients using CLVANCO, with an increase of ~ 50 mL/min/1.73m2 in those with ARC. Children with life-threatening invasive MRSA infection and ARC may not receive adequate vancomycin exposure. Since SCr was very low in those with ARC, it may be useful as an initial marker for identifying ARC.
J. S. Bradley, None
L. Nazer, None
T. Tran, None
A. Nguyen, None
E. Bradley, None
J. Le, None
See more of: Poster Abstract Session