Abstract Number:
1162
Submission Type:
Invited Paper Session
Participants:
Fangya Mao (1), Li Cheung (1), Ran Tao (2), Jonathan Schildcrout (2), Qingning Zhou (3), Fangya Mao (1)
Institutions:
(1) National Cancer Institute, Bethesda, MD, USA, (2) Vanderbilt University, Nashville, TN, USA, (3) University of North Carolina at Charlotte, Charlotte, NC, USA
Chair:
Session Organizer:
Speaker(s):
Session Description:
In modern epidemiological and biomedical studies, a significant challenge arises from budget constraints related to costly covariates such as genotyping, biomarker assays, and medical tests. Two-phase designs provide a valuable solution to this challenge. Generally, Phase I collects outcome data and cost-effective covariates for the entire cohort. In Phase II, a sub-sample is chosen for acquiring the costly covariates, a process that optimizes resource allocation with respect to the target inference. This session features four speakers at various career stages from universities or federal agencies, each actively engaged in advancing this field. They will share insights into recent advancements in two-phase study designs for contemporary biomedical research.
Dr. Ran Tao (Tenure-Track Assistant Professor, Vanderbilt University) will start the session with a presentation on "Efficient designs and analysis of two-Phase studies with longitudinal binary data". He will introduce a novel class of residual-dependent sampling (RDS) designs that select informative individuals using data available on the longitudinal outcome and inexpensive covariates. Together with the RDS designs, he will present a semiparametric analysis approach that efficiently utilizes all available data for parameter estimation implemented by a numerically stable and computationally efficient EM algorithm.
Dr. Jonathan Schildcrout (Professor, Vanderbilt University) will present "A two-phase study focused on longitudinal ordinal outcome data in critically ill sepsis patients". The primary focus is on the design and analysis for a secondary study concerning the association of baseline syndecan-1 concentrations (a biomarker capturing glycocalyx degradation) and ordered states (discharged, hospitalized, intensive care unit, or not alive) over time. He will discuss strategies for identifying the appropriate individuals to sample based on available data and subsequently detail the methods for analyzing this carefully selected sample.
Dr. Qingning Zhou (Associate Professor, University of North Carolina at Charlotte) will present "Improving estimation efficiency of case-cohort study with interval-censored failure time data". She will introduce a novel sieve maximum weighted likelihood estimator developed within the framework of the Cox model based on the case-cohort sample. Additionally, she will present a procedure designed to update this estimator by leveraging information from the full cohort. This proposed method outperforms the commonly-used inverse-probability weighting estimator, showcasing greater efficiency. Moreover, it demonstrates flexibility in its ability to incorporate auxiliary variables, further enhancing estimation efficiency.
Dr. Fangya Mao (Postdoctoral Fellow, National Cancer Institutes) will end this session with a presentation on "Two-phase designs with data from cancer screening cohorts". Such data encompasses both interval-censored incident disease and undiagnosed prevalent disease and mixture models are employed for analysis. The motivation is to evaluate new, expensive tests for US risk-based consensus management guidelines for cervical screening. She will introduce an innovative class of bivariate residual-dependent designs, incorporating modified score-type residuals. This approach enhances the precision of parameter estimation, particularly in relation to the potential utility of the biomarker for diagnostics and prediction.
Sponsors:
No Additional Sponsor 3
ENAR 1
International Society for Clinical Biostatistics 2
Theme:
Statistics and Data Science: Informing Policy and Countering Misinformation
Yes
Applied
Yes
Estimated Audience Size
Small (<80)
I have read and understand that JSM participants must abide by the Participant Guidelines.
Yes
I understand and have communicated to my proposed speakers that JSM participants must register and pay the appropriate registration fee by June 1, 2024. The registration fee is nonrefundable.
I understand