The increasing body of data available in biomedicine makes it crucial to develop and to use adequate biological and statistical models. This is particularly true for clinical investigation of antimicrobials, where one follows the change of several biomarkers, and explains the success of pharmacometrics. Pharmacometrics, the “science of quantitative clinical pharmacology”, is based on the development of semi-physiological models of drug response(s) in order to improve drug development and drug personalization.
We perform research on methods for nonlinear mixed-effects models, which are used in pharmacometrics to analyse variability in the longitudinal data obtained in clinical trials. Our goals are to develop new statistical approaches, evaluate them by simulation, apply them to real data and show their utility and limitations. More specifically we focus on optimal adaptive designs, model evaluation, tests in pharmacogenetics, joint modelling, and model averaging approaches for model-based analysis of pivotal trials. We develop reference R-tools for pharmacometrics, with harmonised libraries of models.
We also provide better insight for the use of antimicrobial agents, especially combined therapy and personalized treatment. We design clinical trials or cohorts to better understand response to anti-infective agents and its variability, especially in HIV, HCV, infectious endocarditis, and community respiratory tract infections. We analyze the longitudinal data of the clinical studies via modeling to derive meaningful knowledge from all measured data. We diversify our expertise to other new or changing infectious agents, as for instance hepatitis Delta, influenza and bacteria resistant to antibiotics. We also focus on the description and understanding of the determinants of emergence of new infectious diseases, develop and evaluate new therapeutic or prevention strategies, and evaluate their psycho-social impact.