Grant: Alliance Grants
McGill University
$44,550.00
Statistical methods for design and analysis of precision clinical trials
EF 2023-24
Résumé
Traditionally clinical trials aim to establish effectiveness of new drugs and therapeutics by showing that the average effect over the patient population is greater than a minimum clinically important level. However, the average treatment effect in most cases is not representative of how any particular patient is going to respond to the treatment. The heterogeneity in patient response to therapeutics is high in some disease settings where patients of certain characteristics respond positively to the treatment but the treatment effect on other patients may be small or even negative. Precision trials aim to better understand the relationship between patient characteristics (covariates) and their response to the treatment and develop treatment rules that are tailored to patients according to their characteristics. The problem of identifying important covariates that can be used to tailor the treatment and the combination of covariate values that correspond to large treatment effects is statistically challenging. In this research we develop a Bayesian modelling frameworks and adaptive designs to simultaneously establish treatment effectiveness and identify the prescriptive sub-set of the covariate space. We will consider these problems in late phase randomized clinical trials where the treatments are dichotomous, i.e., either patients receive the treatment or not, as well as dose optimization trials where an optimal dose needs to be identified for a patient with a given set of covariate values. Particularly, we propose flexible non-parametric models to describe the varying patient response across the covariate space where we borrow ideas from spatial statistics. In addition, we propose a family of Bayesian adaptive designs where the probability of receiving the treatment for each patient is determined based on their characteristics as well as the characteristics and outcome of patients enrolled in the trial thus far. The design will include a variety of adaptive decisions including stopping rules that are defined based on the effect heterogeneity as well as the average treatment effect. An important aspect of the research will focus on development of efficient computational tools that facilitate adoption of the proposed methods in practice.
Programme
Alliance Grants
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
586660-2023
Name nom
Golchi, Shirin S
Department département
Epidemiology, Biostatistics and Occupational Health
Organization id
61
Institution établissement
McGill University
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
44550
Program id
ALLRP
Program name en
Alliance Grants
Program name fr
Subventions Alliance
Group en
RESEARCH PARTNERSHIPS
Group fr
PARTENARIATS DE RECHERCHE
Committee code
96
Committee name en
RPP Internal Decision Cttee
Committee name fr
Cté de décision interne - PPR
Area of application code
1101
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Human health (including medically-related psychological research)
Area of application fr
Santé publique (y compris recherche psychologique liée à la médecine)
Research subject code
3004
Research subject group en
Statistics and probability
Research subject group fr
Applied statistics
Research subject en
Statistique et probabilité
Research subject fr
Statistique appliquée
Application title
Statistical methods for design and analysis of precision clinical trials
Keyword
Bayesian adaptive clinical trials; Dose-response modelling; Effect heterogeneity; Gaussian process; Patient subgroups; Precision dose optimization; Precision medicine
Application summary
Traditionally clinical trials aim to establish effectiveness of new drugs and therapeutics by showing that the average effect over the patient population is greater than a minimum clinically important level. However, the average treatment effect in most cases is not representative of how any particular patient is going to respond to the treatment. The heterogeneity in patient response to therapeutics is high in some disease settings where patients of certain characteristics respond positively to the treatment but the treatment effect on other patients may be small or even negative. Precision trials aim to better understand the relationship between patient characteristics (covariates) and their response to the treatment and develop treatment rules that are tailored to patients according to their characteristics. The problem of identifying important covariates that can be used to tailor the treatment and the combination of covariate values that correspond to large treatment effects is statistically challenging. In this research we develop a Bayesian modelling frameworks and adaptive designs to simultaneously establish treatment effectiveness and identify the prescriptive sub-set of the covariate space. We will consider these problems in late phase randomized clinical trials where the treatments are dichotomous, i.e., either patients receive the treatment or not, as well as dose optimization trials where an optimal dose needs to be identified for a patient with a given set of covariate values. Particularly, we propose flexible non-parametric models to describe the varying patient response across the covariate space where we borrow ideas from spatial statistics. In addition, we propose a family of Bayesian adaptive designs where the probability of receiving the treatment for each patient is determined based on their characteristics as well as the characteristics and outcome of patients enrolled in the trial thus far. The design will include a variety of adaptive decisions including stopping rules that are defined based on the effect heterogeneity as well as the average treatment effect. An important aspect of the research will focus on development of efficient computational tools that facilitate adoption of the proposed methods in practice.