![[ Visit Client Website ]](images/banner.gif)
Objectives: To develop a predictive pharmacokinetic model for accurately dosing gentamicin in newborns across broad ranges of gestational and post natal ages.
Study Design: Population pharmacokinetic analysis.
Methods: Study approval was obtained from our institution’s IRB. Demographic, clinical and gentamicin dosage regimen and serum concentration information was gathered on all NICU patients known to have received gentamicin from September 2009, through July 2014. Data were subjected to cross-checking, error correction, and validation. After validation in Excel, gentamicin dosage regimen information and clinical covariates were placed in Pmetrics format. We used the Pmetrics iterative 2-stage Bayesian population pharmacokinetic modeling program to estimate population parameter values for the one compartment model with multiple infusion inputs.
Results: We collected and analyzed data for 163 neonates (105 males). Their gestational ages ranged from 22 to 42 weeks (73 of 163 were less than 27 weeks). At the time of gentamicin administration ages ranged from day of life (DOL) 2 through DOL 125 (76 of 163 were less than 29 days) while weights were between 395 and 4,930 grams. Only 9 had serum creatinines exceeding 0.8 mg/dl. Good estimates of each patient’s individual parameters were obtained; regression of observed gentamicin serum concentrations on those predicted from the individual’s estimated distribution volume and elimination rate constant yielded an r2 of 0.998. Our predictive model, based on dosing weight, DOL, and serum creatinine yielded an r2of 0.885.
Conclusions: Our predictive model for serum gentamicin levels in neonates appears useful in neonates of GA greater than 23 weeks who have serum creatinines that are age appropriate and who are DOL up to 70 days.