A systematic review and empirical analysis of the relation between dose and duration of drug action.

Journal article

There is a log-linear relation between the dose and duration of action of drugs with single-compartment pharmacokinetics and direct, reversible mechanisms of action. However, it has been suggested that this relation does not extend to drugs whose metabolites are active or slowly eliminated, drugs with saturable kinetics, and drugs with hit-and-run effects. The purpose of this study is to test this hypothesis and to quantify the relationship by way of a systematic review coupled to an empirical analysis. All issues of 4 clinical pharmacology journals from 1980 to 2005 are hand-searched for articles that present pharmacodynamic response versus time curves for 4 or more different doses. Data on duration of action, dose, and area under the plasma concentration versus time curve from zero to infinity (AUC) are abstracted and analyzed by panel data regression modeling, with within-study fixed effects. Duration of drug action is defined as the time during which a pharmacodynamic effect (or response) exceeds a nominal threshold. The generalized models of all observations from 33 publications, with duration of action as the dependent variable and the logarithm of the dose (or AUC) as the explanatory variable, yield significant log-linear relationships. The regressions for individual studies are correctly specified in 27 cases; there are insufficient data for analysis in 10 studies, and a log-linear specification is deemed inappropriate in 6. Analysis of published dose-ranging studies shows that the duration of action of a drug is directly proportional to the logarithm of dose across a wide range of different drugs, extending a result that was previously documented for very few compounds.

Authors: Hughes, D; Aronson, J

• Publication status: Published
• Journal: Journal of clinical pharmacology
• Volume: 50
• Issue: 1
• Pages: 17-26
• Publication date: 2010-01-01
• DOI: 10.1177/0091270008329555
• EISSN: 1552-4604
• ISSN: 0091-2700
• Language: English
• Keywords: Area Under Curve; Prescription Drugs; Models, Statistical; Time Factors; Humans; Dose-Response Relationship, Drug