A multiphase study protocol of identifying, and predicting cancer-related symptom clusters: applying a mixed-method...

Area of Science:

• Oncology
• Palliative Care
• Health Informatics

Background:

• Symptom management in advanced cancer presents challenges, particularly in identifying and predicting symptom clusters (SCs).
• The cluster approach to symptom management is gaining attention for its potential to improve patient care.
• Accurate identification and prediction of SCs are crucial for effective interventions.

Purpose of the Study:

• To identify symptom clusters (SCs) in patients with advanced cancer.
• To develop machine learning algorithms for predicting identified SCs.
• To enhance clinical decision-making and guide symptom management clinical trials.

Main Methods:

• A two-phase study employing a parallel mixed-method design (quantitative and qualitative) for SC identification.
• Quantitative data analyzed using descriptive-analytical methods; qualitative data using content analysis.
• Machine learning, specifically a tree-based method, used to develop predictive algorithms for SCs.

Main Results:

• The study successfully identified symptom clusters in advanced cancer patients.
• Machine learning algorithms were developed to predict these identified SCs.
• Integration of quantitative and qualitative data provided a comprehensive understanding of SCs.

Conclusions:

• The findings facilitate more effective symptom management for cancer patients through SC prediction.
• Predictive models enhance clinical decision-making, leading to improved patient care.
• Results offer valuable guidance for future clinical trials focused on symptom management.