• Marand Blog

healthcare, OPENeP, paperless hospital, precision dosing

Towards the Right Dose of Precision Medicine

June 07, 2018 Emil Plesnik
0 Comments

Determining the optimal medication dose for an individual patient is challenging, as no two patients are the same in the sense of drug absorption, distribution, and clearance. The dosing decision is a complex one, especially for medications with narrow therapeutic ranges, but also for patients with impaired physiological functions and comorbidities. To add to the challenge, clinicians now face this dilemma on a daily basis in busy health environments. 

The conventional dosing strategy used for most medications is either a fixed method (for example, X mg every 12 hours), or a covariate-based dose adjustment (for example, a dose based on weight). These methods are often not precise enough to achieve the optimal therapeutic effect or prevent adverse effects. To optimise dosing, model-informed precision dosing is recommended. This approach determines the right dose for each patient based on multiple factors such as age, weight, gender, genetics, lifestyle, diet, and metabolism. Once implemented, its vital role is referred to as therapeutic drug monitoring (TDM). TDM is the feedback strategy of adjusting the dosing regimen based on plasma drug concentration measurements, which represent additional information to guide the medication therapy. The main advantage of applying TDM is the ability to use the right dose with an increased drug effect, but reduced adverse effects, all of which improves patient outcome and safety. 

Therapeutic drug monitoring can be further enhanced by applying Bayesian statistical methods to predict patient plasma drug concentrations. The required values for a Bayesian prediction of plasma drug concentrations are: the known population pharmacokinetic model for the applied drug, a patient’s demographics, and at least one measured plasma drug concentration which is sampled at a random time within a dosing interval. With additional plasma drug concentration samples, the accuracy of predictions improves. 

Consequently, Bayesian TDM enables the simulation and optimisation of dosing strategies in less time, with improved accuracy, and without the need for regular sampling of plasma drug concentrations. When applying Bayesian TDM, a clinician starts by selecting the correct drug and drug model for the patient being treated. Then, patient demographics relevant to the selected pharmacokinetic model are entered into the system along with dosing details and at least one measured plasma drug concentration. An optimal drug dosage strategy is then selected from among the simulations. The dosing strategy is then further improved with the feedback the system receives from additional plasma drug concentration samples. This method is successful as it allows the patient’s doctor to learn about the causes of an individual patient’s response. 

Individualised precision dosing using Bayesian TDM is becoming a current practice for anticoagulants, aminoglycosides, chemotherapeutics, immunosuppressants, and antiepileptics. In connection with Bayesian dosing, clinical studies have reported that:

Despite all the advantages, the implementation of precision dosing which has been enhanced with Bayesian TDM into modern medical practices has been hindered due to the complex, computationally intensive mathematics involved, something which requires specialised software. Limited access to the data in proprietary monolithic IT solutions is also slowing down the uptake of this precision dosing solution because of the need to manually copy the required data between systems. 

Establishing an open data and open-standards-based platform is vital to enable reliable and secure access to the required data, and to encourage the implementation of patient-centred applications for medication management and precision dosing. This is a crucial facilitator for applying optimal dosing regimens, a key milestone for achieving the optimal treatment for each patient, and a decisive step towards precision medicine.

Written by Emil Plesnik

Data scientist, holding a PhD degree in biomedical data analysis. Curious about data analytics, statistics, machine learning and programming and their use to develop data-driven solutions to challenges in healthcare.

Leave a comment