We introduce Feasibility-Guided Fair Adaptive Reinforcement Learning (FG-FARL), an offline RL procedure that calibrates per-group safety thresholds to reduce harm while equalizing a chosen fairness target (coverage or harm) across protected subgroups. Using de-identified longitudinal trajectories from a Medicaid population health management program, we evaluate FG-FARL against behavior cloning (BC) and HACO (Hybrid Adaptive Conformal Offline RL; a global conformal safety baseline). We report off-policy value estimates with bootstrap 95% confidence intervals and subgroup disparity analyses with p-values. FG-FARL achieves comparable value to baselines while improving fairness metrics, demonstrating a practical path to safer and more equitable decision support.
@article{arxiv.2509.09655,
title = {Feasibility-Guided Fair Adaptive Offline Reinforcement Learning for Medicaid Care Management},
author = {Sanjay Basu and Sadiq Y. Patel and Parth Sheth and Bhairavi Muralidharan and Namrata Elamaran and Aakriti Kinra and Rajaie Batniji},
journal= {arXiv preprint arXiv:2509.09655},
year = {2025}
}