NextDose: A web-based Bayesian dose forecasting tool

Last updated 29 July 2021

Target Concentration

The target concentration principle is the central dogma of rational dose individualization. It forms the basis of the Target Concentration Intervention (TCI) approach to concentration controlled dosing. The principle is based on the assumption that drug effects are determined by the unbound drug concentration at the receptor. Observed responses involve a chain of events after receptor activation. These responses are secondary to the drug binding to the receptor.

For all practical purposes the unbound concentration in blood is a useful predictor of the receptor concentration. There may be delays in reaching and binding to the receptor but once steady state is reached the time related dissociation between unbound concentration in blood and the consequent effect are nearly always negligible.

The target concentration is the concentration which produces the desired target drug effect.

Deciding on the Target Concentration

Once the target drug effect has been decided then the relationship between effect and concentration (pharmacodynamics) can be used to predict the target concentration required to achieve the target effect.

It is however common to find that the pharmacodynamics are not well known. In that case the target concentration can be defined by the steady state concentration associated with the target drug effect. This can be calculated from the dose and dosing interval known to be safe and effect for typical patient and the population drug clearance. The simplest reasonable guess at the target concentration is thus the average steady state concentration.

In clinical practice the concentrations used to individualize drug dose are defined by tradition using empirically based suggestions made in the literature. These concentrations are often described as a therapeutic range typically without a clear pharmacodynamic basis. This range can give some guidance into the distribution of concentrations that are safe (not too high) and effective (not too low). In that case it is better described as the acceptable range to distinguish it from the therapeutic window. The acceptable range is similar to the reference range concept used for other laboratory measurements such as serum sodium or plasma glucose. The therapeutic window mistake occurs when the acceptable range, derived from a population sample, is used to interpret the concentration measured in an individual. Unfortunately, if the measurement is within the therapeutic window it is commonly interpreted to mean that no dose change is required. Dose adjustment only happens if the concentration is outside of the window. This is the hallmark of the therapeutic drug monitoring approach (TDM) which is demonstrably inferior to target concentration intervention (TCI) which uses concentration measurements to predict the dose that will achieve the target and achieves better clinical outcomes (Metz, Holford et al. 2019, Holford, Ma et al. 2020).

The lack of a scientific rationale for choosing a target concentration means measured concentrations are based on convenience e.g. concentrations taken just before the next dose (the trough concentration). The trough concentration is arguably the worst time to measure a concentration in order to estimate key pharmacokinetic parameters such as clearance. To make things even worse trough concentration measurements are confused with the target concentration and clinicians attempt to empirically adjust the dose so try to obtain a measured concentration the same as the target concentration. While this will eventually converge on the right dose it is an inefficient and time wasting process that means the patient is sub-optimally treated.

Deciding on the Concentration Measurement Time

The steady state concentration (Css) is exactly linked to the area under the concentration time curve over a dosing interval (DI) at steady state (Css=AUC0-DI/DI). Surrogates for the steady state target concentration have been devised based on simple algorithms to calculate the associated AUC from several concentration measurements. These algorithms, most commonly the trapezoidal method, are biased and will always underestimate the true AUC. This is because the highest measured concentration is, except by very rare chance, always less than the actual highest concentration. In addition, the trapezoidal approximation assumes linear changes of concentration with time during absorption which further underestimates the true AUC. All trapezoidal AUC methods will therefore lead to an underestimate of the steady state concentration associated with the effect. The AUC based methods get worse when limited sampling strategies are used with just one or two concentrations used to predict the AUC. When these two approximations are combined the results become ever more dubious.

The solution to the dosing challenge is to clearly separate the target concentration from the measurement. The measurement is used, along with a PK model and Bayesian prior information, to estimate the PK parameters for the individual. The PK parameters are then used with the target concentration to predict the dose that will achieve the target. This can be done with any number of measurements including no measurements at all. If only a few concentrations are measured it becomes more important to pay attention to the optimal timing of the concentrations in order to improve the estimation of the PK parameters.

In order to estimate clearance the ideal time is when the concentration measurement is determined only by clearance. In principle this occurs during a constant rate infusion when concentration has approached steady state. Because drug infusion long enough to approach steady state are uncommon this is not a useful method. As an approximation, the concentration in the middle of the dosing interval at steady state may be close to the average steady state concentration. This is certainly a better time than either a single peak or trough to estimate clearance.

When concentrations change a lot during the dosing interval (e.g. gentamicin with a 24 h dosing interval) then two concentration measurements are recommended. The first 30 minutes after the peak (end of infusion) and the second at 8 h after the infusion start. It is not a good idea to take a trough concentration at 24 h after the infusion start because the concentration will often be less than the limit of detection.

Copyright All rights reserved | Developed by Sam Holford & Nick Holford 2012-2021


Holford, N., G. Ma and D. Metz (2020). "TDM is dead. Long live TCI!" British Journal of Clinical Pharmacology Early View(doi:10.1111/bcp.14434).

Metz, D. K., N. Holford, J. Y. Kausman, A. Walker, N. Cranswick, C. E. Staatz, K. A. Barraclough and F. Ierino (2019). "Optimizing Mycophenolic Acid Exposure in Kidney Transplant Recipients: Time for Target Concentration Intervention." Transplantation 103(10): 2012-2030.