ORIGINAL RESEARCH article
Front. Neurol.
Sec. Neurocritical and Neurohospitalist Care
Volume 16 - 2025 | doi: 10.3389/fneur.2025.1700064
This article is part of the Research TopicPrecision Medicine in Neurocritical CareView all 12 articles
Can One-Step Reinforcement Learning Guide Optimal Timing for PEG and Tracheostomy in Severe TBI? Insights from a 2016-2023 Retrospective Cohort Study at a Single Academic Institution
Provisionally accepted
- 1University of South Florida, Morsani College of Medicine, Florida, United States
- 2Tampa General Hospital, Tampa, United States
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Background: Acute management of Traumatic brain injury (TBI) presents several challenges in hospital resource planning. While early Trach (Trach) and percutaneous endoscopic gastrostomy (PEG) tube placement may improve patient outcomes, the optimal timing and selection criteria for these interventions remain unclear. This study evaluates the impact of PEG and Trach timing on key clinical outcomes and applies one-step reinforcement learning (RL) to recommend intervention timing. Methods: This retrospective cohort study included 263 adult intensive care unit inpatients (194 males, 69 females, age range 18-87) diagnosed with severe TBI requiring Trach and/or PEG between January 1, 2016, and December 31, 2023, at a single academic institution. Key outcomes included ICU and hospital length of stay, complications, time to oral feeding/decannulation, readmission, and mortality. One-step Temporal Difference (TD) Learning and Q-learning were used to predict the expected value of interventions and recommend optimal timing based on patient states, respectively. Results: Early PEG and Trach interventions were linked to significantly shorter ICU and hospital LOS with fewer complications. Delayed PEG placement, however, was associated with a 67% odds reduction of mortality (OR: 0.33, p = 0.033) compared to early placement despite having more total complications. One-step RL suggested greater cumulative rewards with earlier intervention and successfully recommended the ideal day of PEG/Trach intervention using initial patient presentation. Conclusion: Early interventions are generally linked with improved outcomes, although delaying PEG or Trach placement may be advantageous in select situations toward reducing mortality. RL techniques, such as TD and Q-learning, can support decisions for intervening.
Keywords: Traumatic Brain Injury, Tracheostomy, PEG, reinforcement learning, Clinical decision support
Received: 05 Sep 2025; Accepted: 21 Oct 2025.
Copyright: © 2025 Babel, Vanderpool, Inkel and Behbahaninia. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Milad Behbahaninia, behbahaninia@usf.edu
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.