Schedule

Regular meeting time: Every other Thursday 12:15 - 1:15 pm, Fishbowl Conference Room (UOP 2-254) or online

October 2 (12:15 - 1:15 PM, UOP 2-254 Fishbowl Conf Rm, hybrid)

Presenter: Han Lu

Topic: Restricted Mean Survival Time (RMST), which is the area under the survival curve from time 0 to a restricted time tau, is gaining research popularity as it offers an alternative summary to proportional hazard-based models in comparing treatment effects for clinical trials. However, how to choose the restricted time tau? This paper explores a data-driven way to find the optimal restricted time considering effect size and estimation precision at the same time.

Reference:  Beyond Fixed Restriction Time: Adaptive Restricted Mean Survival Time Methods in Clinical Trials by Drs. Jinghao Sun, Douglas Schaubel, and Eric Tchetgen Tchetgen

October 16 Postponed to October 23 (12:15 - 1:15 PM, UOP 2-254 Fishbowl Conf Rm, hybrid)

Presenter: Dr. Anne Eaton 

Topic: How to test and conclude the treatment effects when two survival curves are expected to cross at a certain time? This paper proposes statistical methods for comparing treatments when survival curves cross, focusing on detecting differences in long-term survival after a prespecified time point. 

Reference: Comparing Treatments in the Presence of Crossing Survival Curves: An Application to Bone Marrow Transplantation, Logan, Klein, and Zhang (2008)

October 30 (12:15 - 1:15 PM, UOP 2-254 Fishbowl Conf Rm, hybrid)

Presenter: Caleb Griffiths

Topic: Our division's Ph.D. student, Caleb, will present an overview of the pseudo-observations. 

Reference: Pseudo-observations in survival analysis, Andersen and Perme (2009)

November 14 (11 AM -1 PM, UOP 2-274 Student Collaboration Rm)

Note: this is NOT our usual meeting time! We moved the usual meeting to Friday to watch a webinar together. 

Title: Statistical analysis with the occurrence of a terminal event

Speaker: Bin Nan, Ph.D., Chancellor's Professor, Department of Statistics, UC Irvine 

Sponsor: Lifetime Data Science Section (ASA)

Abstract: In cohort studies, time plays important roles in data collection and data analysis. Two distinct sets of statistical tools have been developed primarily for analyzing data in cohort studies: one is longitudinal data analysis that focuses on properly handling temporal dependence among repeatedly collected measurements over time, one is survival analysis that deals with time-to-event data when censoring occurs. In this webinar, I will present some recently developed more interpretable work on analyzing data in cohort studies in a particular situation: a terminal event occurs.

For analyzing longitudinal data with the occurrence of a terminal event that is subject to right censoring, I will first introduce the idea of conditional modeling given the terminal event time (usually death time) via a parametric model, then extend it to a more robust nonparametric bivariate time-varying coefficient model in which the time-varying coefficients capture the longitudinal trajectories of covariate effects along with both the follow-up time and the residual lifetime.

For analyzing time-to-event data, e.g. disease onset, with the occurrence of a terminal event, in contrast to the commonly used semi-competing risks or illness-death model, I will introduce a nonparametric estimation of the distribution of the disease onset time conditional on the death time, which consists of two components that provide straightforward interpretations of the disease onset during the lifespan. Regression models are constructed for these two components when evaluating risk factors of disease onset is of interest.

All these statistical developments are motivated by real studies. In addition to presenting statistical properties and simulation results, I will also present the results of several data examples analyzed by the proposed methods.

November 27

No meeting on Thanksgiving - Happy Holiday!

December 11 (12:15 - 1:15 PM, UOP 2-254 Fishbowl Conf Rm, hybrid)

Topic: end-of-semester social event.

Spring 2026, event in scheduling

Presenter: Dr. Tianmeng Lyu

Topic (tentative): Restricted mean survival time (RMST) estimation in a clinical trial with a composite endpoint of an interval-censored event (e.g., progression) and right-censored death, benefits by converting the composite endpoint to interval censoring. 

Reference:  Gao et al. (2025), RMST for Interval-Censored Data in Oncology Clinical Trials