Explore how patient information, a published population model, and measured vancomycin levels change estimated exposure over time.
Educational use only. This prototype has software verification but no independent clinical validation. Do not use it to make patient-care decisions.
Patient
Enter a current clinical snapshot
Regimen
Describe the planned dosing schedule
Levels
Add measured concentrations if available
Exposure
Review estimated AUC and concentration
This controls model eligibility and auto-selection. It does not diagnose the patient.
Model selection
General adult intermittent infusion selects Thomson 2009.
Fit for this patient profile
Within published population ranges
No encoded age, weight, renal-range, obesity, or dialysis warning was triggered.
The model uses
The model cannot see
Current patient summary: 55 years, 70.0 kg, SCr 1.00 mg/dL. Eligibility flags describe published ranges; they are not a clinical suitability determination.
Add measured concentrations and their timing to update the population estimate. This explorer assumes the regimen above was unchanged and administered on schedule.
No measured levels added
Add a concentration and its timing to update the population estimate
First 24 hours
AUC₀–₂₄311mg·h/L
Begins with no modeled vancomycin present
Second 24 hours
AUC₂₄–₄₈421mg·h/L
Second modeled day on the unchanged regimen
At steady state
AUC₂₄,ss532mg·h/L
Steady-state average from daily dose ÷ CL
Early exposure and steady-state exposure answer different questions and should not be used interchangeably.
Preparing interactive chart…
| Dose (mg) | Interval | End-infusion Cmax,ss | Pre-dose Cmin,ss | AUC₂₄,ss | Rationale |
|---|---|---|---|---|---|
1000Nearest target match | q12h | 33.7 mg/L | 14.6 mg/L | 533 mg·h/L | AUC₂₄ 532 mg·h/L within target 400-600 |
2000 | q24h | 49.7 mg/L | 10.3 mg/L | 533 mg·h/L | AUC₂₄ 532 mg·h/L within target 400-600 |
1750 | q24h | 43.5 mg/L | 9.0 mg/L | 466 mg·h/L | AUC₂₄ 466 mg·h/L within target 400-600 |
750 | q8h | 32.7 mg/L | 18.9 mg/L | 599 mg·h/L | AUC₂₄ 599 mg·h/L within target 400-600 |
Plain text for an EHR note or pharmacist handoff. It includes the entered clinical inputs, model provenance, fit warnings, levels, PK estimates, and modeled exposure.
Patient name, date of birth, medical-record number, and encounter identifiers are not collected.
Review before adding to the medical record. This export is a calculation summary and is not a medication order.
Published equation translation
Verified locally
Source equations reproduced with reference cases
Structural PK mathematics
Verified locally
Infusion, superposition, steady-state, and mass-balance invariants
Bayesian implementation
Verified locally
Synthetic recovery, residual error, and optimizer-boundary checks
External retrospective validation
Not performed
Requires timestamped deidentified patient data
Prospective clinical validation
Not performed
Requires protocol, governance, and independent review
Clinical decision support status
Not authorized
Educational demonstration only
“Verified locally” means the implementation passed the repository’s deterministic software tests. It does not establish predictive performance in an external population.
Deterministic log-grid MAP estimation uses each model’s published CL/volume variability and residual error. A local log-scale Laplace approximation makes parameter uncertainty visible without presenting it as exact.
One-, two-, and three-compartment IV models cover intermittent and continuous infusion. Dialysis clearance is switched on only during entered Westra HDF/HD sessions.
Auto-selection considers clinical population and administration structure. Every model remains labeled Match, Caution, Not applicable, or Source data required; override cannot bypass structural incompatibility.