# The hidden risks of temporal resampling in clinical reinforcement learning

arXiv: https://arxiv.org/abs/2602.06603 · HTML: https://mlsanity.com/paper/2602.06603/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Safe offline RL, one of 13 directions in Offline RL and imitation learning from data.
- Builds on 40 of its 124 references that are in the map.
- First posted on arXiv 6 Feb 2026.

## Placed in

- Offline RL and imitation learning from data / [Safe offline RL](https://mlsanity.com/offlinerl/safe-offline-rl/)
- Deep reinforcement learning methods / [RL for clinical treatment and surgery](https://mlsanity.com/deeprl/rl-for-clinical-treatment-and-surgery/)

## Builds on (most cited first, 30 of 40)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems](https://mlsanity.com/paper/2005.01643/) (2020)
- [Conservative Q-Learning for Offline Reinforcement Learning](https://mlsanity.com/paper/2006.04779/) (2020)
- [Offline Reinforcement Learning with Implicit Q-Learning](https://mlsanity.com/paper/2110.06169/) (2021)
- [Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction](https://mlsanity.com/paper/1906.00949/) (2019)
- [Deep Reinforcement Learning at the Edge of the Statistical Precipice](https://mlsanity.com/paper/2108.13264/) (2021)
- [Recurrent Neural Networks for Multivariate Time Series with Missing Values](https://mlsanity.com/paper/1606.01865/) (2016)
- [Neural Controlled Differential Equations for Irregular Time Series](https://mlsanity.com/paper/2005.08926/) (2020)
- [Minigrid & Miniworld: Modular & Customizable Reinforcement Learning Environments for Goal-Oriented Tasks](https://mlsanity.com/paper/2306.13831/) (2023)
- [Batch Policy Learning under Constraints](https://mlsanity.com/paper/1903.08738/) (2019)
- [Multi-Time Attention Networks for Irregularly Sampled Time Series](https://mlsanity.com/paper/2101.10318/) (2021)
- [Deep Reinforcement Learning for Robotics: A Survey of Real-World Successes](https://mlsanity.com/paper/2408.03539/) (2024)
- [Supervised Reinforcement Learning with Recurrent Neural Network for Dynamic Treatment Recommendation](https://mlsanity.com/paper/1807.01473/) (2018)
- [Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences](https://mlsanity.com/paper/1610.09513/) (2016)
- [Self-Supervised Transformer for Sparse and Irregularly Sampled Multivariate Clinical Time-Series](https://mlsanity.com/paper/2107.14293/) (2021)
- [When Should We Prefer Offline Reinforcement Learning Over Behavioral Cloning?](https://mlsanity.com/paper/2204.05618/) (2022)
- [Continuous State-Space Models for Optimal Sepsis Treatment - a Deep Reinforcement Learning Approach](https://mlsanity.com/paper/1705.08422/) (2017)
- [A Reinforcement Learning Approach to Weaning of Mechanical Ventilation in Intensive Care Units](https://mlsanity.com/paper/1704.06300/) (2017)
- [A Survey on Principles, Models and Methods for Learning from Irregularly Sampled Time Series](https://mlsanity.com/paper/2012.00168/) (2020)
- [Bootstrapping Fitted Q-Evaluation for Off-Policy Inference](https://mlsanity.com/paper/2102.03607/) (2021)
- [Improving Sepsis Treatment Strategies by Combining Deep and Kernel-Based Reinforcement Learning](https://mlsanity.com/paper/1901.04670/) (2019)
- [An Empirical Study of Representation Learning for Reinforcement Learning in Healthcare](https://mlsanity.com/paper/2011.11235/) (2020)
- [Medical Dead-ends and Learning to Identify High-risk States and Treatments](https://mlsanity.com/paper/2110.04186/) (2021)
- [Clinician-in-the-Loop Decision Making: Reinforcement Learning with Near-Optimal Set-Valued Policies](https://mlsanity.com/paper/2007.12678/) (2020)
- [Deep Reinforcement Learning for Dynamic Treatment Regimes on Medical Registry Data](https://mlsanity.com/paper/1801.09271/) (2018)
- [Multi-Objective SPIBB: Seldonian Offline Policy Improvement with Safety Constraints in Finite MDPs](https://mlsanity.com/paper/2106.00099/) (2021)
- [Semi-Markov Offline Reinforcement Learning for Healthcare](https://mlsanity.com/paper/2203.09365/) (2022)
- [Unifying Cardiovascular Modelling with Deep Reinforcement Learning for Uncertainty Aware Control of Sepsis Treatment](https://mlsanity.com/paper/2101.08477/) (2021)

## Built on by (most cited first, 2 of 2)

- [Insulin4RL: Real-Time Insulin Management in the Intensive Care Unit for Offline Reinforcement Learning](https://mlsanity.com/paper/2606.19481/) (2026)
- [What a Data Release Must Preserve for Off-Policy Evaluation](https://mlsanity.com/paper/2610.15283/) (2026)

## Abstract (arXiv)

Reinforcement learning (RL) is a type of artificial intelligence for making optimal choices. In healthcare, researchers generally use offline RL (ORL), where models are trained and evaluated from retrospective observational data. To accommodate naturally irregular clinical records, researchers often resample the data into uniform time intervals before training (known as binning). However, binning may create counterfactual trajectories that do not represent clinical practice. We first quantified these distortions in patients receiving continuous insulin infusions in the MIMIC-IV intensive care database. Binning reversed the apparent causal order for over 70% of paired glucose measurements and insulin adjustments within the same interval, and frequently masked significant dysglycaemic events and merged distinct treatment decisions together. We then conducted an in silico clinical trial using 30 virtual patients with type 1 diabetes from the FDA-approved UVA/Padova simulator. The simulator was modified to include stochastic intervals between decisions and used to generate a training dataset for offline RL. We trained three ORL algorithms on both the unprocessed dataset and equivalent datasets resampled at 10-minute, 2-hour, and 4-hour intervals. When deployed back into the simulated environment, temporal resampling was found to reduce model performance by up to 60% relative to unprocessed data, with 4-hour binning causing all agents to perform worse than the dataset's baseline. Retrospective evaluation on resampled data actively obscured this effect, predicting 1.5-3x better returns than agents achieved in practice. We recommend that future research in this area prioritises datasets with natural clinical timings between decisions, which may be a necessary step before these models can be safely deployed into patient care.