Keywords: survival analysis, time-to-event, counterfactual inference, causal survival analysis
Abstract: Balanced representation learning methods have been applied successfully to counterfactual
inference from observational data. However, approaches that account for
survival outcomes are relatively limited. Survival data are frequently encountered
across diverse medical applications, i.e., drug development, risk profiling, and clinical
trials, and such data are also relevant in fields like manufacturing (for equipment
monitoring). When the outcome of interest is time-to-event, special precautions
for handling censored events need to be taken, as ignoring censored outcomes may
lead to biased estimates. We propose a theoretically grounded unified framework
for counterfactual inference applicable to survival outcomes. Further, we formulate
a nonparametric hazard ratio metric for evaluating average and individualized
treatment effects. Experimental results on real-world and semi-synthetic datasets,
the latter which we introduce, demonstrate that the proposed approach significantly
outperforms competitive alternatives in both survival-outcome predictions and
treatment-effect estimation.
One-sentence Summary: We propose a counterfactual survival analysis framework for adjusting for bias from two unknown sources, namely, confounding due to covariate dependent selection bias and censoring mechanism (informative or non-informative).
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Supplementary Material: zip
Reviewed Version (pdf): https://openreview.net/references/pdf?id=YsI-sbZoa
12 Replies
Loading