Determining hypertensive patients' beliefs towards medication and associations with medication adherence using...

Area of Science:

• Biomedical Informatics
• Machine Learning in Healthcare
• Cardiovascular Health

Background:

• Hypertension management requires understanding medication adherence.
• Conventional methods for assessing adherence have limitations.
• Machine learning applications for predicting adherence in hypertensive patients are novel.

Purpose of the Study:

• To assess the feasibility of machine learning (ML) models for identifying factors associated with medication adherence in hypertensive patients.
• To develop and compare ML models (RF, ANN, SVR) for predicting adherence levels.
• To identify key variables influencing adherence using ML techniques.

Main Methods:

• Utilized data from 160 hypertensive patients.
• Employed Random Forests (RF) for variable importance ranking and backward elimination.
• Developed RF, Support Vector Regression (SVR), and Artificial Neural Networks (ANN) models to predict adherence.
• Used Self-Organizing Feature Maps (SOM) for data visualization.

Main Results:

• RF, SVR, and ANN models achieved high accuracy in predicting adherence (78%, 79%, 65% respectively).
• Identified significant predictors of adherence: educational level, marital status, General Overuse, monthly income, and Specific Concern.
• ML model predictions showed no significant difference from actual adherence scores.

Conclusions:

• ML offers an effective alternative to conventional methods for identifying adherence factors in hypertensive patients.
• Findings can inform targeted patient education strategies to improve medication adherence.
• This approach enhances understanding of patient adherence for better hypertension management.