Identifying the Risk Factors Associated with Nursing Home Residents' Pressure Ulcers Using Machine Learning Methods.

Area of Science:

• Computational biology
• Health informatics
• Machine learning applications in healthcare

Background:

• Machine learning (ML) offers advanced capabilities for improving predictive accuracy and automating knowledge discovery through data-driven insights.
• Predictive modeling is essential in healthcare for identifying at-risk populations and informing preventative strategies.

Purpose of the Study:

• To evaluate and compare the performance of various machine learning algorithms for predicting pressure ulcers (PUs).
• To assess the accuracy, sensitivity, specificity, and predictive values of different ML models using real-world datasets.

Main Methods:

• Application of representative ML algorithms: random forest, logistic regression, and various Support Vector Machine (SVM) models (linear, polynomial, radial, sigmoid).
• Development of a predictive model utilizing a dataset of N = 60.
• Validation of models based on key performance metrics including accuracy, sensitivity, specificity, and predictive values.

Main Results:

• The random forest model achieved the highest prediction accuracy (0.814).
• Logistic regression (0.782), polynomial SVM (0.779), radial SVM (0.770), and linear SVM (0.767) showed comparable performance.
• Sigmoid SVM yielded the lowest accuracy (0.674).

Conclusions:

• The random forest model is the most effective for predicting pressure ulcers (PUs) in nursing home settings.
• Identified diverse factors, including nursing home and resident characteristics, that are predictive of PUs.
• Emphasized the importance of considering these identified factors to mitigate PU development in nursing home residents.