Exploration of once-daily dosing of aminoglycosides through Bayesian simulation

To simulate peak and trough concentrations, using Bayesian forecasting, achieved with a variety of once-daily dosing (ODD) regimens; to evaluate dosing regimens required to produce target peak and trough concentrations; to compare the peak-to-MIC (minimum inhibitory concentration) ratios and time above MIC for various dosing regimens and MICs; to stratify the information based on renal function estimates; and to use the results from these simulations to make recommendations regarding optimum ODD of aminoglycosides.

Simulation of ODD using a Bayesian technique and existing patient data.

A tertiary referral, community teaching hospital.

One hundred consecutive adults from the author's data base, with a wide variety of infections and underlying illnesses, who met strict inclusion criteria.

Two methods of dosing, weight-based (4-7 mg/kg) and target concentration-based (peak 10, 15, or 20 mu g/ml, trough < or = 0.3 mu g/ml), were evaluated. Each patient had a known dosing-sampling history, stable renal function, and at least two measured serum concentrations, and were being treated with either gentamicin or tobramycin.

A wide range of peak and trough serum concentrations are achieved when dosages are chosen based on patient weight. Even with large dosages, some patients had very low peak-to-MIC ratios and time above the MIC, and vice versa. Variations in MIC had a much greater effect on dosing target values than did variations in dosage. A large degree of variability was also noted in doses and dosing intervals when using a target serum concentration approach. For both methods, an inverse relationship existed between calculated creatinine clearance and time above MIC, although there was little change over the range of 60-119 ml/minute.

Bayesian simulation showed that weight-based ODD of aminoglycosides did not produce clinically acceptable serum concentrations or target values in many patients. Young and elderly patients, and any patient with a creatinine clearance below 60 or above 119 ml/minute, are especially likely not to achieve an optimum serum concentration profile. Aminoglycoside ODD should be individualized by evaluating the peak-to-MIC ratio, time above MIC, and patient response.