For the high-inoculum experiments, the stock was used without further dilution.
For the low-inoculum experiments, a further 1-in-4,000 dilution was made in Ca-MH broth to achieve a final inoculum of ∼4 ×10 CFU/ml.
The simulations suggested a dose-dependent change in bacterial densities based on findings from the high- and low-bacterial-inoculum experiments (Fig. 7).
Clearance, Vc, Kgmax,s, Kgmax,r, and POPMAX were assumed to be the same in both the high- and low-inoculum experiments.
For the low-inoculum experiments, the r 2 values of Bayesian prior predictions for the bacterial densities were 0.884 and 0.885 for the total population and resistant subpopulation, respectively.
Equation 5, 6, and 7 would take the same form as equations 2, 3, and 4 but describe the effect of drug in low-inoculum experiments and are not shown here.
In the low-bacterial-inoculum experiments, trough plasma total piperacillin concentration/MIC ratios of 3.4 and 10.4, from the bolus and extended-infusion regimens, respectively, were required to suppress the emergence of antimicrobial resistance.
For the high-inoculum experiments, the r 2 values of Bayesian prior predictions versus observations for the bacterial densities were 0.329 and 0.782 for the total population and resistant subpopulation, respectively.
Working conditions, in terms of substrate and inoculum proportions (experiment 1, EXP 1) and the liquid replacement schedule (experiment 2, EXP 2) were initially set.
The pyramiding of the two R-genes proved to be the only management practice which totally suppressed the emergence of virulent isolates, both in laboratory with high inoculum density experiments and in the 3-year field experiment.
