Nonparametric instrumental regression with competing risks data
Principal Investigator
Name
Ingrid Van Keilegom
Degrees
Ph.D.
Institution
Katholieke Universiteit Leuven
Position Title
Professor
Email
About this CDAS Project
Study
HIPB
(Learn more about this study)
Project ID
HIPB-2
Initial CDAS Request Approval
Mar 1, 2021
Title
Nonparametric instrumental regression with competing risks data
Summary
This project aims at developing a new nonparametric methodology to solve endogeneity issues with competing risks data. The proposed technique uses an instrumental variable to estimate the causal effect of an endogenous treatment, when the outcome of interest are competing durations. Such an instrumental variable is available when there is noncompliance in the treatment arm of a randomized trial, as is the case in the HIPB trial, which is therefore uniquely suited to our research needs. In such a situation, "naively" comparing treatment and control groups only identifies the effect of belonging to the treatment group rather than the actual effect of the treatment. In contrast, our approach identifies treatment effects on the usual quantities of interest with competing risks data: cause-specific hazards, subdistribution hazards and cause-specific cumulative incidence functions. Random right censoring on the competing durations is allowed. The project is funded by the European Research Council (2016-2021, Horizon 2020 / ERC grant agreement No.\ 694409).
Aims
- Study identification of the effect of an endogenous treatment on competing risks outcomes
- Develop a methodology to estimate treatment effects using an instrumental variable
- Use the technique to uncover new findings on real datasets
Collaborators
Professor Ingrid Van Keilegom, ORSTAT, KU Leuven (Belgium)
Professor Jean-Pierre Florens, Toulouse School of Economics, Université Toulouse Capitole (France)
Dr. Jad Beyhum, ORSTAT, KU Leuven (Belgium)