Clinical Pharmacokinetic Dosing Application

Precision dosing for Vancomycin and Aminoglycosides using Bayesian PK modeling

Hello, Stranger!

Please log in or register to access the application.

About This Application

This free application provides clinical decision support for therapeutic drug monitoring and dosing optimization of:

  • Vancomycin — Area Under the Curve (AUC₂₄) based dosing targeting 400-600 mg·hr/L
  • Aminoglycosides (Gentamicin, Tobramycin, Amikacin) — Peak and trough based dosing with extended-interval (pulse) or traditional dosing support
HIPAA Compliance & Patient Privacy

This application is designed for de-identified patient data entry. All patient records use unique healthcare identifiers (HCID format: SSSS-NNNNNN) rather than names or medical record numbers, supporting HIPAA-compliant workflows and protecting patient privacy. Site-based access controls ensure data isolation between institutions and users.

Key Features:
  • Patient Management — Track demographics, weights, creatinine clearance, and clinical history
  • Initial Dosing Recommendations — Population-based dosing using patient-specific parameters
  • Bayesian PK Fitting — Individualized parameter estimation using drug levels
  • 1-Compartment (MP/Matzke-Pai) — Standard model for both drug classes
  • 2-Compartment Models — Goti and Carreno models for vancomycin when indicated
  • Target Attainment Analysis — Assess if current regimen achieves therapeutic goals
  • Dosing Adjustments — Recommendations to optimize peak, trough, or AUC₂₄
  • Population Analytics — Dashboard with model performance metrics and quality indicators
  • Multi-User Collaboration — Site-based access control for team workflows
  • Data Export — CSV export of PK fits and de-identified patient data

New to the App? Practice with a Test Patient

You can learn the entire workflow without entering real patient data. From the Initial Dosing page, use the two yellow/blue practice buttons:

  1. Create Test Patient (top of Initial Dosing page) — generates a fully populated test patient with randomized demographics (age, sex, height, weight), a baseline serum creatinine, and an HCID prefixed with TEST-. Test patients are private to your account and are auto-deleted after 5 days, so they will not clutter your dashboard or analytics.
  2. Populate Practice Levels & Go to PK Fitting (appears at the bottom of Initial Dosing once you have run Calculate Initial Doses and Calculate Peak/Trough/AUC) — uses the loading dose, maintenance dose, interval, and infusion period that you just calculated to seed the test patient with:
    • 1 loading dose + 5 maintenance doses leading up to steady state
    • A steady-state peak and trough level (±10% jitter around the predicted values)
    • Several serum creatinine values across the dosing window
    You are then taken directly to the PK Fitting page with the test patient pre-loaded so you can run a Bayesian fit and see how the individualized parameters compare to the population estimates.

Practice patients are excluded from analytics (Dashboard, MP/2-Compartment Model Performance, Algorithm Comparison) so they will not affect any of your real-population statistics.

Getting Started — Workflow Guide

Step 1: Initial Dosing
  1. Navigate to Initial Dosing
  2. Select Drug from dropdown
  3. Select Dosing Method:
    • Traditional — Q8H, Q12H, Q24H dosing
    • Pulse (aminoglycosides only) — Extended-interval dosing (24-48 hr)
  4. Select Patient or enter patient data directly in the Patient Data — Edit Values card
  5. Update Conditions (if applicable)
  6. Update Indications (if applicable)
  7. Click Calculate Initial Doses to generate population-based recommendations
  8. Review and adjust dosing parameters as needed:
    • Edit Loading Dose
    • Edit Maintenance Dose
    • Edit Dosing Interval
    • Edit Infusion Period
  9. Press Calculate Peak / Trough / AUC to project steady-state values
  10. Save Loading Dose and Save Dosing Regimen once satisfied with calculated values
Step 2: Record Administered Doses
  1. Navigate to Doses → Add Dose
  2. Enter actual administered dose:
    • Dose (mg)
    • Infusion Period (hr)
    • Date/Time of administration
    • Drug
  3. Record all doses administered to the patient
Step 3: Record Drug Levels
  1. Go to Levels → Add Level
  2. Enter measured drug concentration:
    • Level (mg/L) — Lab result
    • Level Type — Peak, Trough, Random, Pre-dose, Post-dose
    • Date/Time of sample collection
  3. Ensure levels are drawn at appropriate times:
    • Trough — Just before next dose
    • Vancomycin Peak — 2 hr after infusion end
    • Aminoglycoside Peak — 1 hr after infusion end
Step 4: Record Serum Creatinines
  1. Navigate to Creatinines → Add Creatinine
  2. Enter serum creatinine measurements:
    • Creatinine (mg/dL) — Lab result
    • Date/Time of sample collection
  3. Record baseline creatinine before starting therapy and monitor regularly
  4. System automatically calculates creatinine clearance (CrCl) using Cockcroft-Gault equation
Step 5: PK Fitting & Individualized Dosing
  1. Navigate to PK Fitting
  2. Select the patient and drug
  3. System automatically loads:
    • Patient demographics and renal function
    • Dose history prior to drug levels
    • All measured drug levels
  4. Choose PK Model:
    • MP/Matzke-Pai — 1-compartment (default for all drugs)
    • Goti — 2-compartment for vancomycin
    • Carreno — 2-compartment for vancomycin
  5. Click Fit Model to perform Bayesian optimization
  6. Review fitted parameters (Vd, CL, Ke, T½) and prediction error metrics
  7. Adjust dose/frequency in the "Dosing Adjustment Recommendation" section
  8. Calculate steady-state projections (peak, trough, AUC₂₄)
  9. Save Current Regimen if dose adjustment is needed
Step 6: Ongoing Monitoring
  1. Use Patient Dashboard for individual patient overview:
    • View all PK fittings and parameter trends
    • Check target attainment (AUC₂₄ for vancomycin, trough for aminoglycosides)
    • Review dose history and level history
  2. Repeat Steps 2-5 as needed when new levels are drawn
  3. Monitor for:
    • Vancomycin: AUC₂₄ 400-600 mg·hr/L
      Note: AUC monitoring requires both peak and trough levels for accuracy. Trough-only monitoring is less precise.
    • Gentamicin/Tobramycin: Trough <2 mg/L, Peak 8-10 mg/L (traditional) or >15 mg/L (pulse)
    • Amikacin: Trough <10 mg/L, Peak 20-30 mg/L (traditional) or >45 mg/L (pulse)

Population Analytics Dashboard

Access MP Model Performance from the navigation menu to view institution-wide analytics:

  • Patient Demographics — Age, weight, BMI, renal function distributions
  • PK Parameter Distributions — Fitted vs population parameters (Vd, CL, Ke, T½)
  • Clearance vs Renal Function — Linear regressions with BSA normalization
  • Target Attainment — Percentage of patients achieving therapeutic goals
  • Model Quality Metrics — Fit convergence rates, residual analysis, prediction errors
  • Dosing Accuracy — Predicted vs achieved targets for model validation
  • Body-Size Diagnostics — Assess model bias across BMI categories
  • Multiple Linear Regression — Identify covariates affecting drug clearance

Use the drug filter dropdown to switch between Vancomycin and Aminoglycosides analytics.

Additional Information

Dosing Strategy Definitions:
  • Traditional Dosing — Standard intermittent dosing (Q8H, Q12H, Q24H) targeting conventional peak and trough ranges
  • Pulse Dosing (Aminoglycosides) — Extended-interval dosing (Q24-48H) with higher doses, targeting prolonged post-antibiotic effect and minimizing toxicity
PK Model Selection:
  • MP/Matzke-Pai (1-compartment) — First-line model for all drugs; sufficient for most clinical scenarios
  • 2-Compartment Models (Goti, Carreno) — Consider for vancomycin when:
    • Early post-dose samples (<4 hr after infusion) show poor fit with 1-comp model
    • Distribution phase is clinically relevant
    • Note: Requires more data points in a dosage interval for reliable parameter estimation
Best Practices:
  • Update creatinine regularly — Renal function affects clearance; recent SCr improves dose recommendations
  • Record dose times accurately — PK calculations depend on precise timing
  • Draw levels at appropriate times — Both a peak and trough are recommended: Peaks aminoglycosides: 1 hr after end of infusion, vancomycin: 2 hr after end of infusion. Troughs before a dose except pulse aminoglycoside dosing when a 14 hour post dose level is recommended.
  • Save regimens — Document dosing recommendations for tracking and validation
  • Review model fit quality — Check convergence status and prediction error before accepting dosing changes

Ready to Get Started?

Register for an account to begin using the Clinical PK Dosing Application.

Register Log In