How Vancomyzer calculates a regimen, and what has been checked.
Vancomyzer is a Bayesian vancomycin dosing calculator for clinicians. This page explains the model it uses, how measured levels change an estimate, and which checks have been run. Vancomyzer has not yet been validated in real patients.
A dose recommendation is only reviewable if the prior, the fit and the evidence behind the model are visible. Some tools show a number without them. Commercial platforms add EHR integration, population-specific models and implementation support; compare those against your institution’s needs. Vancomyzer shows the model, the assumptions and the evidence behind each estimate, and the core calculator is free for individual clinicians.
The Colin 2019 prior is documented with its primary citation. A model-based 90% credible band (not yet validated against patient levels) shows how much a fit can and cannot say. The Bayesian step is explained in plain language inside the calculator (turn on Teaching mode in the calculator’s settings).
Population-model equations
The Colin 2019 equations describe how patient characteristics affect the starting estimates. The calculator then adjusts those estimates to measured blood levels. Reviewing that adjustment requires more than checking one equation.
Fit to measured levels
Bayesian estimates combine the population model with measured levels, fitted to the doses actually given, including a loading dose when one is entered. The calculator shows measured-versus-predicted differences and flags a poor fit for clinical review.
90% credible band
The shaded band shows the middle 90% of concentrations from 400 simulated sets of pharmacokinetic values. It reflects uncertainty in the model estimates, updated with blood levels when available. It excludes measurement error, so it is not a range for predicting a future blood test result. Its reliability in patients has not been established.
Published starting model
Colin PJ, et al. Vancomycin pharmacokinetics throughout life: results from a pooled population analysis and evaluation of current dosing recommendations. Clin Pharmacokinet. 2019;58(6):767-780. Pooled population analysis of data from 14 studies (2,554 individuals), from neonates to elderly adults. Published model evidence does not establish clinical validation of Vancomyzer.
Scope
Vancomyzer supports adults receiving intermittent intravenous vancomycin. Pediatric dosing, dialysis and continuous infusion are outside its scope. Independent clinical validation is pending.
Access
Access to Vancomyzer is currently free. All plans use the same calculation method. Contact us about account access or institutional requirements.
Vancomycin clearance, Colin 2019
Clearance is the product of a typical value and four patient-related factors. Source: Clin Pharmacokinet. 2019;58(6):767-780, Table 3.
CL = θCL × (weight / 70)^0.75 size scaling
× F·maturation ≈ 1.0 in adults
× F·age-decline 50% lower by age 61.6 y
× F·creatinine serum creatinine effectWith measured levels, the individual estimate starts from this population value and is adjusted to fit the patient’s levels. The adjustment is bounded: one unusual level cannot override the population data; it moves the estimate.
The population model is Colin 2019, a pooled population pharmacokinetic analysis of data from 14 studies (2,554 individuals) spanning neonates through elderly adults. It is used for every adult, at every body size. Published evaluation at BMI 40 or more is limited (Colin 2021: 15 of 49 obese adults), so review measured levels early in heavier patients.
Colin 2019 two-compartment population PK model
Colin PJ, et al. Vancomycin pharmacokinetics throughout life: results from a pooled population analysis and evaluation of current dosing recommendations. Clin Pharmacokinet. 2019;58(6):767-780.
The population model, used for every adult at every body size. CC BY-NC.
doi:10.1007/s40262-018-0727-5Janmahasatian S et al. — Quantification of lean bodyweight (FFM equations)
Clin Pharmacokinet. 2005;44(10):1051-1065.
Fat-free mass is displayed for context at BMI 40 or more; it does not change the Colin 2019 calculation.
doi:10.2165/00003088-200544100-00004Rybak MJ et al. — Therapeutic monitoring of vancomycin (ASHP/IDSA/PIDS/SIDP 2020)
Am J Health Syst Pharm. 2020;77(11):835-864.
Source of the AUC₂₄ 400–600 mg·h/L target.
doi:10.1093/ajhp/zxaa036| Page | What it covers | Status |
|---|---|---|
| Equations and derivations | Every equation and constant the calculator uses, with the reference check for a typical adult. | Published model |
| Literature reproducibility | Published vancomycin cases run through the calculator. Selected population predictions are pass/fail checks; cases from other models provide context. | Developer-run |
| Predictive performance | 200 simulated ICU patients using a model adapted from published research. Each prediction is checked against a concentration not used to calculate it. | Developer-run |
| Comparison with Tucuxi | The same starting assumptions and simulated blood levels compared in two programs. Their approaches to blood-level uncertainty differ. Repeating the Vancomyzer calculations gave the same estimates for these cases. | Developer-run |
| Independent reference calculation | Concentration, AUC and dose-history calculations compared with a separate mathematical calculation, including 40 random loading-dose schedules; agreement within one part in a million. | Developer-run |
| Loading-dose handling | Levels drawn after a loading dose are fitted to the doses actually given, removing the overestimate that comes from recording the load as a maintenance dose. | Developer-run |
| Uncertainty band coverage | Synthetic patients: how often the true concentration falls inside the 90% band, under the calculator's assumptions and selected published variability estimates. | Not yet validated |
| Independent clinical validation | Retrospective and prospective evaluation with patient data at a participating institution. | Pending |
“Developer-run” identifies checks designed and performed by the software’s author using published values or simulated patients. These are not independent clinical validation.
- Vancomyzer is not FDA-cleared or approved. Its intended regulatory basis and limitations are described in the medical disclaimer.
- A poor fit to measured levels calls for review of timing, data quality and the clinical situation. It does not establish that a proposed regimen is appropriate.
- The developer uses published cases and synthetic comparisons to test the software. This work is not independent clinical validation.
- The clinician must review a recommendation against the patient’s condition, local protocol and therapeutic drug monitoring.
Vancomyzer™ is a clinical decision-support tool for qualified healthcare professionals only. Not FDA-cleared or approved. Designed to meet the non-device clinical decision support criteria of FD&C Act §520(o)(1)(E); not reviewed by the FDA. Engineered by Dōsys™.