Explainable machine learning models for identifying mild cognitive impairment in older patients with chronic pain
Abstract Background Mild cognitive impairment (MCI) is prevalent in older adults with chronic pain, making early detection crucial for dementia prevention and healthy aging. This study aimed to determine MCI risk factors in older patients with chronic pain and to develop 9 machine learning models to...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-01-01
|
| Series: | BMC Nursing |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12912-025-02723-8 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Background Mild cognitive impairment (MCI) is prevalent in older adults with chronic pain, making early detection crucial for dementia prevention and healthy aging. This study aimed to determine MCI risk factors in older patients with chronic pain and to develop 9 machine learning models to identify MCI risk. Methods A total of 612 older patients with chronic pain were recruited between October 2023 and July 2024. Data collected included patients’ general information, cognitive function, pain level, depression, and sleep quality. The dataset was randomly divided into training set and testing set, and processed by Min-Max Normalization and SMOTETomek comprehensive sampling. SVM-RFE and LASSO regression were used for variable selection. We then developed machine learning models and interpreted them by SHAP. Results Age, education level, number of pain sites, pain duration, pain level, depression and sleep quality were risk factors of MCI in older patients with chronic pain. The Extreme Gradient Boosting (XGBoost) model performed best (AUC 0.925), with pain level, age, and depression as the most important variables. Conclusions We successfully developed 9 machine learning models to identify MCI risk. These models provide a tool for nurses to detect MCI risk early. We recommend that nurses integrate machine learning techniques into clinical nursing practice for managing MCI. However, these findings require validation with longitudinal data to confirm predictive validity for MCI progression. |
|---|---|
| ISSN: | 1472-6955 |