The Random Forest Model Has the Best Accuracy Among the Four Pressure Ulcer Prediction Models Using Machine Learning...

Area of Science:

• Nursing
• Medical Informatics
• Machine Learning

Background:

• Pressure ulcers represent a significant nursing adverse event.
• Accurate prediction of pressure ulcer development is crucial for patient care.
• Big data and machine learning offer potential for improved predictive modeling.

Purpose of the Study:

• To develop and compare machine learning models for predicting pressure ulcer nursing adverse events.
• To identify the optimal model for accurate pressure ulcer occurrence prediction.
• To assess the feasibility of a big data-driven management system for pressure ulcer prediction.

Main Methods:

• Retrospective analysis of 5814 patients, with 1673 experiencing pressure ulcer events.
• Development of prediction models using Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), and Artificial Neural Network (ANN).
• Evaluation of 19 screening variables and comparison of model performance based on accuracy and AUC values.

Main Results:

• All four developed models demonstrated good predictive performance.
• Area Under the Curve (AUC) values for all models exceeded 0.95.
• The Random Forest (RF) model exhibited higher accuracy in predicting pressure ulcer occurrence compared to other models.

Conclusions:

• Machine learning and big data technologies are feasible for developing predictive systems for nursing adverse events.
• Random Forest and Decision Tree models are particularly suitable for constructing pressure ulcer prediction models.
• This study provides a foundation for future big data-based pressure ulcer risk warning systems.