Related Experiment Video
Updated: Jun 7, 2025
09:16
Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli
Published on: April 5, 2019
10.6K
Predicting frailty in older patients with chronic pain using explainable machine learning: A cross-sectional study
Xiaoang Zhang1, Yuping Liao1, Daying Zhang2
1School of Nursing, Jiangxi Medical College, Nanchang University, Nanchang, China; Department of Pain Medicine, the 1(st) affiliated hospital, Jiangxi Medical College, Nanchang University, Nanchang, China.
Geriatric Nursing (New York, N.Y.)
|November 9, 2024
Summary
Identifying frailty in older adults with chronic pain is vital. A machine learning model using factors like pain level and depression showed high accuracy for early frailty detection.
Area of Science:
- Gerontology
- Computational Medicine
- Pain Management
Background:
- Frailty is prevalent in older adults experiencing chronic pain, increasing risks of falls, disability, and cognitive decline.
- Current methods for identifying frailty in this demographic are insufficient.
- Early detection of frailty is critical for timely intervention and improved health outcomes.
Purpose of the Study:
- To investigate frailty risk factors in elderly individuals with chronic pain.
- To develop and evaluate machine learning models for accurate frailty identification.
- To enhance clinical decision-making for geriatric patients with persistent pain.
Main Methods:
- Developed nine distinct machine learning models to predict frailty.
- Utilized the Shapley Additive Explanations (SHAP) method for model interpretability.
- The Random Forest (RF) model was selected based on performance metrics.
Main Results:
- The Random Forest model achieved high performance: 0.822 accuracy, 0.797 precision, and 0.881 AUC.
- Key predictive variables included age, BMI, education, pain duration, number of pain sites, pain level, depression, and Activity of Daily Living (ADL).
- Pain level, depression, and ADL were identified as the most significant factors in the RF model.
Conclusions:
- A robust machine learning model, particularly Random Forest, can effectively identify frailty in older adults with chronic pain.
- The model's reliance on clinical and self-reported data facilitates practical application in healthcare settings.
- Early identification via this model can lead to targeted interventions, promoting healthy aging and reducing adverse health events.
