LiDS 2019

The 2_nd Conference on Lifetime Data Science

The Lifetime Data Science Section of American Statistical Association is pleased to announce an exciting upcoming conference on Lifetime Data Science: Foundations and Frontiers which will be held at the University of Pittsburgh May 29-31, 2019. The event will begin with short courses by experts in topics of current interest on May 29 and will be followed by a two-day conference featuring keynote addresses by internationally renowned statisticians, a student paper competition, a poster session and many stimulating talks. A banquet will be held on May 30, 2019.

Keynote Speakers

Causal inference for survival data, with emphasis on mediation analysis

Odd O. Aalen

Thursday, May 30, 9:00-10:00 EST

Causal inference is becoming an important theme in survival analysis. We discuss causal mediation analyses for survival data and propose an approach based on the additive hazards model. The emphasis is on a dynamic point of view, that is, understanding how the direct and indirect effects develop over time. To define direct and indirect effects in a longitudinal survival setting we take an interventional approach (Didelez, 2018) where treatment is separated into one aspect affecting the mediator and a different aspect affecting survival. In general, this leads to a version of the non-parametric g-formula (Robins, 1986). In the present talk, we demonstrate that combining the g-formula with the additive hazards model and a linear structural equation model for the mediator process results in simple and interpretable expressions for direct and indirect effects in terms of relative survival as well as cumulative hazards. Our results generalise and formalise the method of dynamic path analysis (Fosen et al, 2006; Strohmaier et al, 2015) and also work by Lange and Hansen (2011). An application will be given.

Regression Models and Multivariate Life Tables

Ross Prentice

Friday, May 31, 8:00-9:00 EST

Regression methods that adapt Cox regression to multivariate failure times,on the same or different failure time axes,will be presented. These methods specify Cox –type semiparametric regression models for marginal single and double failure rates, and use estimating functionsand empirical process methods,like those developed by Danyu Lin, L.J. Weiand colleagues for marginal single failure hazard rates,forhazard ratio parameter and for baseline hazard rate estimation. Sandwich –type variance process estimatorsare developed for all model parameters, along with a perturbation resampling procedure for complex constructs of modeled parameters. As a byproduct semiparametric estimators of pairwise survivor functions,given covariates that may be evolving in time, are readily obtained from Peano series representations of these survivor functions in terms of marginal single and double failure rates, and corresponding semiparametric estimators of cross ratio and concordance functions are also readily obtainedto characterize pairwise dependenciesbetween failure timesgiven covariates. An application to clinical outcome data from a large low-fat dietary intervention trial among postmenopausal women will be presented, and some contrast between these approaches and thosebased on counting process intensity modeling, as well asonfrailty and copula modeling,will be provided. This is joint work with Dr. Shanshan Zhao of NIEHS.

Semiparametric Regression Analysis of Interval-Censored Data

Danyu Lin

Friday, May 31, 9:00-10:00 EST

Interval censoring arises frequently in clinical, epidemiological, financial, and sociologicalstudies, where the event or failure of interest is not observed at an exact time point but is rather known to occur within a time interval induced by periodic monitoring. We formulate the effects of potentially time-dependent covariates on the interval-censored failure time through semiparametric regression models, such as the Cox proportional hazards model. We study nonparametric maximum likelihood estimation with an arbitrary number of monitoring times for each study subject. We develop an EM algorithm that involves very simple calculationsand converges stably for any dataset, even in the presence of time-dependent covariates. We show that theestimators for the regression parameters areconsistent, asymptotically normal, and asymptotically efficient with an easily estimated covariance matrix. In addition, we extend the EM algorithm and asymptotic theory to competing risks and multivariate failure time data. Finally, we demonstrate the desirable performance of the proposed numerical and inferential procedures through simulation studies and applications to real medical studies.

Short Courses

May 29, 2019, 8:30-16:30 EST

Two-Phase Studies for Lifetime Data

Speakers: Ørnulf Borgan (University of Oslo, Norway)
                     Sven Ove Samuelsen (University of Oslo, Norway)

Dynamic Prediction in Survival Analysis

Speaker: Hein Putter (Leiden University Medical Center, The Netherlands)

Biased Sampling, Left Truncation and Survival Analysis

Speaker: Jing Qin (NIH/NIAID)

Important Dates

Submissions

Conference Registration

To register click Here!