Jul 2026
Assessing penalized approaches for estimating causal treatment effects under extremely limited overlap in oncology.
In oncology subgroups where low treatment prevalence and limited effective overlap lead to unstable or non-convergent LRMs, penalized regression approaches provide a practical strategy for improving ATO estimation.
Sangwon Lee, H. Chae, Dong-woo Choi et al.
· Cancer Epidemiology · 0 citations