Developing a Wearable Sensor-Based Digital Biomarker of Opioid Dependence
Authors' conclusion
Wearable sensor-derived digital biomarkers can be used to predict opioid use status (naïve versus chronic) and the differentiating features may be detecting opioid dependence.
Study context
- Population
- Hospitalised patients with acute pain receiving opioid analgesia, monitored with wearable sensors through hospitalisation and up to 7 days post-discharge169 opioid administrations
- Participants
- 41 patients
- Exposure
- Opioid-naïve vs chronic opioid users with acute pain
- Outcome
- Prediction of opioid use status (naïve vs chronic) by wearable sensor-based machine-learning models
Methodological detailsmachine-learning model development, cross-validation and holdout validation
- Machine-learning model development
- Cross-validation and holdout validation
Main finding
As reported in the abstract
Participant-level performance of 4 decision tree-based models (14 sensor-based, 5 clinical features)
Accuracy (range)
70% to 90%
Sensitivity (range)
67% to 100%
Specificity (range)
67% to 100%
Observational data: associations, not effects.
Original publication
Developing a Wearable Sensor-Based Digital Biomarker of Opioid Dependence. Anesth Analg. Published 16 October 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →