Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
July 26, 2026Psychological MethodsOpen Access

Adjustment set selection for estimating optimal treatment rules under confounding.

View Full Paper
Ask AI
Bookmark
Share

Authors

NGNina GalanterSSSusan M. ShortreedEMErica E. M. Moodie

Discussion

Loading...

Member takes

Overview

Randomized trial investigates variable selection for optimal treatment rules in unipolar depression, indicating treatment tailoring effects.

Key Points

  • This research aims to compare various variable selection strategies for estimating optimal individualized treatment rules in depression care.
  • Compared outcome adaptive lasso, group lasso, doubly robust estimation, double-index propensity score, high-dimensional balancing propensity score, and causal ball lasso for variable selection.
  • Analyzed data from 74,058 Kaiser Permanente Washington patients with new antidepressant dispensation between 2008 and 2018.
  • Utilized electronic health records to assess effect on symptom severity after treatment.
  • All variable selection methods provided similar unbiased estimates.
  • Differed in excluding extraneous variables and computational efficiency.
  • Tailoring treatment based on baseline symptom severity showed no significant impact on symptom severity after 6 months (p-value not specified).

Cite This Study

Galanter et al. (2026) studied this question.

synapsesocial.com/papers/6a65a49fd3aea3239cd771cdhttps://doi.org/10.1037/met0000812
View Full Paper
Ask AI
Bookmark
Share