Heterogeneous survivor average causal effects beyond monotonicity: Applications to a clinical trial evaluating mechanical ventilation strategies
Abstract
Clinical trials in critical care often evaluate outcomes that are truncated by death, such as time to discharge alive, for which treatment effects are not well defined among patients who would not survive under one or both treatment strategies. Principal stratification provides a natural framework for defining survivor causal effects, but existing approaches often rely on monotonicity assumptions that may be implausible in settings where treatment can affect survival in competing directions. This concern is illustrated by the ARDS Network trial of lower versus higher positive end-expiratory pressure (PEEP), in which higher PEEP may benefit some patients by improving alveolar recruitment while harming others through over-distention or hemodynamic compromise. Moreover, the largely null average findings of the original trial do not rule out the possibility of clinically meaningful subgroups that may benefit from or be harmed by higher PEEP. We propose a monotonicity-free framework for identifying conditional survivor average causal effects (CSACE) using an interpretable sensitivity parameter that characterizes latent principal-stratum membership. We then develop a flexible Bayesian estimation strategy based on BART, a posterior-mean-based variable importance measure, and a distribution-aware conditional inference tree that uses the full posterior draws of individualized CSACEs. In simulation studies, the proposed BART-based approach improves individualized CSACE estimation and variable-importance recovery compared with linear modeling, especially under nonlinear treatment-effect heterogeneity. Applied to the ARDS PEEP trial, our method reveals clinically interpretable heterogeneity among estimated always-survivors, identifying subgroups with posterior evidence of benefit or harm that are obscured by the overall null trial results.
Related
- Optimal bang-bang ventilation under mechanical cycle closure: a reduced lung gas-exchange model
- Monotonicity and bounded variation for harmonic mappings with applications to special functions
- Monotonicity formulas in positively curved settings with applications to two-phase free boundary problems
- Validation of a Computational Respiratory System Model for Mechanical Ventilation
- Evaluating multi-season occupancy models with autocorrelation fitted to heterogeneous datasets
BibTeX
@misc{zhu2026heterogeneous,
title = {{Heterogeneous survivor average causal effects beyond monotonicity: Applications to a clinical trial evaluating mechanical ventilation strategies}},
author = {Zihan Zhu and Guangyu Tong and Fernando Godinho Zampieri and Snigdha Jain and Michael O. Harhay and Fan Li},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.23211},
eprint = {2609.23211},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.23211},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Heterogeneous survivor average causal effects beyond monotonicity: Applications to a clinical trial evaluating mechanical ventilation strategies
AU - Zihan Zhu
AU - Guangyu Tong
AU - Fernando Godinho Zampieri
AU - Snigdha Jain
AU - Michael O. Harhay
AU - Fan Li
PY - 2026
DA - 2026/09/19
PB - arXiv
DO - 10.48550/arxiv.2609.23211
UR - https://arxiv.org/abs/2609.23211
AB - Clinical trials in critical care often evaluate outcomes that are truncated by death, such as time to discharge alive, for which treatment effects are not well defined among patients who would not survive under one or both treatment strategies. Principal stratification provides a natural framework for defining survivor causal effects, but existing approaches often rely on monotonicity assumptions that may be implausible in settings where treatment can affect survival in competing directions. This concern is illustrated by the ARDS Network trial of lower versus higher positive end-expiratory pressure (PEEP), in which higher PEEP may benefit some patients by improving alveolar recruitment while harming others through over-distention or hemodynamic compromise. Moreover, the largely null average findings of the original trial do not rule out the possibility of clinically meaningful subgroups that may benefit from or be harmed by higher PEEP. We propose a monotonicity-free framework for identifying conditional survivor average causal effects (CSACE) using an interpretable sensitivity parameter that characterizes latent principal-stratum membership. We then develop a flexible Bayesian estimation strategy based on BART, a posterior-mean-based variable importance measure, and a distribution-aware conditional inference tree that uses the full posterior draws of individualized CSACEs. In simulation studies, the proposed BART-based approach improves individualized CSACE estimation and variable-importance recovery compared with linear modeling, especially under nonlinear treatment-effect heterogeneity. Applied to the ARDS PEEP trial, our method reveals clinically interpretable heterogeneity among estimated always-survivors, identifying subgroups with posterior evidence of benefit or harm that are obscured by the overall null trial results.
ER -