Grant: Discovery Grants Program - Individual
University of Saskatchewan
$25,000.00
Joint Models for Follow-Up Data: Extensions, Diagnostics and Remedies
EF 2024-25
Résumé
In follow-up studies, different types of variables are collected for each unit under study. In Engineering, for example, it is often of interest to study the time until a building is no longer useable. To achieve this, a follow-up study can be used to collect data on age of the structure, several indicators of structural integrity (e.g., structural cracks) for which measurements are collected repeatedly over time, wind speed, and time to failure of the structure; the aim is to evaluate if age, the indicators of structural integrity, and wind speed are strong predictors (called covariates) for the risk of failure of the structure in question (the response variable is time to failure of a structure or time-to-event). Here, some covariates are assumed fixed for the duration of the study (e.g., age of structures), and some are time-dependent (e.g., structural cracks). The focus of this proposal is on follow-up studies where both fixed and time-dependent covariates are available along with an observation on the time to an event of interest. A time-dependent covariate can also be internal or external. An internal covariate is generated by the unit under study (e.g., an indicator of structural integrity, where the unit is a structure), whereas an external covariate changes its value due to external characteristics (e.g., wind speed). The joint modeling approach works in two steps: (a) modeling the internal covariate taking into account measurement error to construct its complete history, and (b) quantifying the effects of the covariates on time-to-event taking into account the history as constructed in (a). Combining the two processes enables us to mutually borrow information from each process and gain efficiency in statistical inference. I will emphasize four topics which the standard joint modeling approach cannot properly handle and requires methodological development. The first topic considers developing methods to check whether the effects of the covariates are constant over time. The second topic considers situations where the precise values of the covariates are not available, with the aim to develop methods to estimate true covariate effects using surrogate data. The third topic involves developing an approach to understand the effect of a binary internal covariate out of a particular state (e.g., working condition/failure) on time-to-event. The fourth topic involves dynamic predictions, where survival probabilities are dynamically updated as additional information is recorded. Overall, the proposal deals with novel statistical methods, which involve both theory and computation. The long-term objective is to establish a statistical package to analyze follow-up data with wide ranging characteristics, which includes joint modeling. The participation of HQP in this project is quite important. They will be involved by contributing new ideas and implementing methodologies. All HQP will be trained for a research career or inspiring employment.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Saskatchewan, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2023-03969
Name nom
Khan, Shahedul
Department département
Mathematics and Statistics - Mathematics and Statistics
Organization id
16
Institution établissement
University of Saskatchewan
Province en
Saskatchewan
Province fr
Saskatchewan
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
25000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1508
Committee name en
Mathematics and Statistics
Committee name fr
Mathématiques et statistique
Area of application code
1206
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Engineering
Area of application fr
Génie
Research subject code
3005
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Biostatistics
Research subject fr
Biostatistiques
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Joint Models for Follow-Up Data: Extensions, Diagnostics and Remedies
Keywords
survival analysis; time-to-event data; follow-up studies; longitudinal data; joint modeling
Application summary
In follow-up studies, different types of variables are collected for each unit under study. In Engineering, for example, it is often of interest to study the time until a building is no longer useable. To achieve this, a follow-up study can be used to collect data on age of the structure, several indicators of structural integrity (e.g., structural cracks) for which measurements are collected repeatedly over time, wind speed, and time to failure of the structure; the aim is to evaluate if age, the indicators of structural integrity, and wind speed are strong predictors (called covariates) for the risk of failure of the structure in question (the response variable is time to failure of a structure or time-to-event). Here, some covariates are assumed fixed for the duration of the study (e.g., age of structures), and some are time-dependent (e.g., structural cracks). The focus of this proposal is on follow-up studies where both fixed and time-dependent covariates are available along with an observation on the time to an event of interest. A time-dependent covariate can also be internal or external. An internal covariate is generated by the unit under study (e.g., an indicator of structural integrity, where the unit is a structure), whereas an external covariate changes its value due to external characteristics (e.g., wind speed). The joint modeling approach works in two steps: (a) modeling the internal covariate taking into account measurement error to construct its complete history, and (b) quantifying the effects of the covariates on time-to-event taking into account the history as constructed in (a). Combining the two processes enables us to mutually borrow information from each process and gain efficiency in statistical inference. I will emphasize four topics which the standard joint modeling approach cannot properly handle and requires methodological development. The first topic considers developing methods to check whether the effects of the covariates are constant over time. The second topic considers situations where the precise values of the covariates are not available, with the aim to develop methods to estimate true covariate effects using surrogate data. The third topic involves developing an approach to understand the effect of a binary internal covariate out of a particular state (e.g., working condition/failure) on time-to-event. The fourth topic involves dynamic predictions, where survival probabilities are dynamically updated as additional information is recorded. Overall, the proposal deals with novel statistical methods, which involve both theory and computation. The long-term objective is to establish a statistical package to analyze follow-up data with wide ranging characteristics, which includes joint modeling. The participation of HQP in this project is quite important. They will be involved by contributing new ideas and implementing methodologies. All HQP will be trained for a research career or inspiring employment.