Clinical Decision Support
An advanced pharmacokinetic calculator implementing Bayesian Maximum A Posteriori (MAP) estimation for individualized vancomycin dosing. Supports multiple validated population PK models, AUC-guided dosing per ASHP/IDSA 2020 consensus guidelines, and real-time concentration-time profiling.
Methodology: Five published population PK models with Bayesian MAP estimation. AUC-guided dosing per ASHP/IDSA/SIDP 2020 consensus guidelines. All calculations run client-side with no patient data transmitted.
Enter measured vancomycin serum levels to refine PK estimates via Bayesian MAP estimation. The algorithm updates population priors with patient-specific data to generate individualized parameters.
No observed levels entered
Add drug levels to enable Bayesian parameter estimation
Blue line: predicted concentration curve | Red dots: observed levels | Green band: target trough range
| Dose (mg) | Interval | Pred Peak | Pred Trough | AUC₂₄/MIC | Rationale |
|---|---|---|---|---|---|
1750Recommended | q12h | 40.9 mg/L | 8.1 mg/L | 487 | AUC₂₄/MIC 487 within target 400-600 |
1250 | q8h | 35.8 mg/L | 12.4 mg/L | 522 | AUC₂₄/MIC 522 within target 400-600 |
2000 | q12h | 46.7 mg/L | 9.3 mg/L | 557 | AUC₂₄/MIC 557 within target 400-600 |
1000 | q8h | 28.7 mg/L | 9.9 mg/L | 418 | AUC₂₄/MIC 418 within target 400-600 |
1500 | q12h | 35.1 mg/L | 7.0 mg/L | 418 | AUC₂₄/MIC 418 within target 400-600 |
Grid-search optimization minimizing the MAP objective function with log-normal priors. Two-phase approach: coarse grid (30×20) followed by fine refinement (20×20) around the optimum.
One-compartment IV infusion model with multi-dose superposition. Steady-state peak/trough via accumulation factor. AUC₂₄ calculated both analytically (Dose₂₄/CL) and via trapezoidal integration.
Five validated models spanning 1984-2018: Matzke, Bauer, Thomson, Crass, and Empiric Trough. Each provides population priors for CL and Vd with inter-individual variability (BSV) estimates.