TDMx: A web-application for therapeutic drug monitoring enhanced by pharmacometrics
Sebastian G. Wicha (1), Martin G. Kees (1,2), Alexander Solms (3), Iris K. Minichmayr (1), Alexander Kratzer (4), Charlotte Kloft (1)
(1) Dept. of Clinical Pharmacy and Biochemistry, Institute of Pharmacy, Freie Universitaet Berlin, Germany, (2) Dept. of Anaesthesiology and Intensive Care, Charité Universitaetsmedizin Berlin - Campus Benjamin Franklin, Berlin, Germany, (3) Institute of Mathematics, University of Potsdam, Germany, (4) Hospital Pharmacy, University Hospital Regensburg, Germany
Background: Pharmacometric models have evolved as useful tools to quantify and explain pharmacokinetic (PK) variability between patients and to explore its resulting pharmacodynamic (PD) consequences e.g. on probabilities of target attainment (PTA) of anti-infectives. The application of pharmacometric PK/PD models in clinical practice is yet limited, as available software is either difficult to use or lacks functionality. Hence, we aimed to develop ‘TDMx’ (www.tdmx.eu), an easy-to-use, but powerful modular software tool for bedside dosing decisions, making use of state-of-the-art pharmacometric techniques such as (i) probability of target attainment analysis without requiring drug measurements, (ii) Bayesian PK estimation and definite PK/PD target attainment analysis if drug measurements are available and (iii) (adaptive) optimal design to sample at the most informative time points. As a starting point, population PK models for the anti-infectives gentamicin, amikacin, meropenem and piperacillin have been implemented.
- TDMx is freely accessible via a user-friendly web interface from any common (modern) browser and quickly provides answers to the clinically relevant TDM-related questions.
- Patient module: Input module for all available patient data (e.g. demographics, laboratory measurements) as well as the dosing schedule provided to ‘TDMx’.
- Probabilistic dosing module: Computes the PTA to select a likely effective dosing regimen solely using patient covariates, e.g. to initiate empiric therapy, or to guide institutions where no drug measurements are available.
- Bayesian dosing module: Estimates individual PK parameters for definite target attainment analyses if drug measurements are available.
- Optimal design module: Predicts the most informative sampling time points for precise estimation of either PK parameters (e.g. in a research setting) or PD surrogates (e.g. T>MIC as a relevant PK/PD index in clinical routine).
- ‘TDMx’ was successfully validated against NONMENTM.
Conclusion: With ‘TDMx’, we provide a user-friendly and powerful web-application, making use of state-of-the art pharmacometric techniques to support bedside dosing decision-making. In comparison to other currently available tools, ‘TDMx’ offers broader functionality and can entirely be used in a web browser.