GENERAL VERSUS SUBPOPULATION VALUES IN BAYESIAN PREDICTION OF AMINOGLYCOSIDE PHARMACOKINETICS IN HEMATOLOGY-ONCOLOGY PATIENTS

The predictive performance of Bayesian estimates incorporating pharmac okinetic values for hematology-oncology patients was compared with tha t of Bayesian estimates incorporating general population values. In st udy phase 1, medical records were reviewed for 50 adult patients with a hematologic or oncologic diagnosis who had received i.v. gentamicin or tobramycin. Aminoglycoside pharmacokinetic values were calculated f or the patients by using a modified two-point Sawchuk-Zaske method, an d the subpopulation mean for each variable was determined. In phase 2, data for 10 other hematology-oncology patients receiving aminoglycosi des were entered into the Abbottbase Bayesian pharmacokinetics program . Aminoglycoside pharmacokinetic values and serum concentrations for e ach of these 10 patients were estimated, first using the program's gen eral population values and then repeating the analysis using the subpo pulation means for Volume of distribution and renal clearance slope ob tained in phase 1. The serum aminoglycoside concentrations predicted b y each Bayesian method were compared with the actual peaks and troughs . Both the peak and trough predictions of the Abbottbase program using the subpopulation values for volume of distribution and renal clearan ce slope were significantly less biased than those predicted by the Ab bottbase program incorporating the general population values. The meth ods did not differ significantly in precision. Use of subpopulation ph armacokinetic values in Bayesian predictions of serum aminoglycoside c oncentrations in hematology-oncology patients reduced bias significant ly but had no significant effect on precision.