A Smartphone-Based Decision Support Tool for Predicting Patients at Risk of Chemotherapy-Induced Nausea and Vomiting:...

Area of Science:

• Oncology
• Data Science
• Digital Health

Background:

• Chemotherapy-induced nausea and vomiting (CINV) significantly impacts patient quality of life and treatment outcomes.
• Effective management of CINV is crucial for patient well-being and treatment adherence.

Purpose of the Study:

• To compare data mining models for predicting CINV risk.
• To develop a smartphone application for clinical decision support in CINV risk assessment.

Main Methods:

• Retrospective review of electronic medical records from 2010-2012.
• Utilized five data mining algorithms (naive Bayes, logistic regression, neural network, support vector machine, decision tree) on six patient datasets based on emetogenicity and CINV phase.
• Developed an iOS smartphone app using the decision tree algorithm with both bulk and adaptive questionnaire approaches.

Main Results:

• The decision tree algorithm demonstrated superior performance in predicting CINV risk for high emetogenic chemotherapies.
• Accuracy ranged from 79.3% to 94.8% across six patient groups.
• The developed smartphone app utilizes the decision tree for its consistent performance and explainable nature, offering an adaptive approach for point-of-care use.

Conclusions:

• The study provides a personalized, evidence-based solution for CINV management.
• The developed application can enhance patient quality of life and potentially reduce healthcare costs.
• The findings support the integration of data mining and mobile health technology in oncology care.