Autonomous Delirium Monitoring and Adaptive Prevention
Delirium is a common acute brain dysfunction syndrome affecting up to 50% of patients in the intensive care unit (ICU). Delirium is associated with prolonged ICU stay, increased mortality, long-term cognitive impairments, and an annual cost of $164 billion in the United States. Non-pharmacological approaches remain the cornerstone of delirium prevention, but the strategy of observing patient mobility and circadian desynchrony due to sleep disruption and light and noise exposure often falls short due to sporadic human observations.
The overall objective of this study is to develop ADAPT, the Autonomous Delirium Monitoring and Adaptive Prevention system, using novel pervasive sensing and deep learning (DL) techniques. ADAPT will autonomously monitor and quantify patient’s mobility and circadian desynchrony in terms of nightly disruptions, light intensity, and sound pressure level. These risk factors will be integrated into a dynamic model to predict delirium trajectories. From the monitored factors and/or the predicted delirium trajectory, ADAPT will prompt adaptive actions in real-time to increase patient mobility, reduce nightly disruptions, optimize ambient light, and reduce noise.
Our central hypothesis is that ADAPT will be superior to existing approaches for identifying high-risk patients by autonomously monitoring uncaptured delirium risk factors, and will ultimately provide an effective, non-pharmacological adaptive prevention strategy for delirium in the ICU.
This tool will be developed and validated in terms of a) predictions of delirium trajectory transitions that are precise and interpretable, b) autonomous monitoring and quantification of mobility and circadian cues that correlate to mobility assessments and circadian biomarkers, and c) adaptive prevention with real-time action prompts that ICU physicians are satisfied with.
ADAPT’s successful application would augment clinical decision-making by streamlining clinical processes and overall improve patient outcomes in the ICU.
Select ADAPT Publications
- “Potentials and Challenges of Pervasive Sensing in the Intensive Care Unit,” Davoudi A, Shickel B, Tighe PJ, Bihorac A, Rashidi P. Frontiers in Digital Health. 2022;4:773387. Published 2022 May 17.
- “The effect of non-pharmacologic strategies on prevention or management of intensive care unit delirium: a systematic review,” Cupka JS, Hashemighouchani H, Lipori J, Ruppert MM, Bhaskar R, Ozrazgat-Baslanti T, Rashidi P, and Bihorac A. F1000 Research. 2020;9:1178. Published 2020 Sep 28.
More ADAPT Information
- ADAPT: Autonomous Delirium Monitoring and Adaptive Prevention, National Institutes of Health (NIH) RePORTER