Personalised Risk Modelling for Older Adult Cancer Survivors: Combining Wearable Data and Self-Reported Measures to...

Area of Science:

• Digital Health
• Oncology
• Geriatric Medicine

Background:

• Electronic health records (EHRs) lack daily life context for remote patient monitoring.
• Wearable devices offer objective data but need integration with patient-reported outcomes.
• Older adult cancer survivors require personalized care considering evolving risk factors.

Purpose of the Study:

• To develop dynamic risk models for vulnerability and anxiety in older adult cancer survivors.
• To integrate wearable sensor data with self-reported outcomes for personalized cancer care.
• To analyze the influence of activity and sleep patterns on patient-reported vulnerability and anxiety.

Main Methods:

• Utilized process mining techniques to analyze real-world data from wearable devices and self-reported outcomes.
• Combined objective data from wearables with subjective patient data (vulnerability, anxiety, depression).
• Developed two dynamic risk models in collaboration with healthcare professionals using data from the LifeChamps study.

Main Results:

• Prolonged sedentary activity correlated with a higher risk of vulnerability.
• Highly dynamic sleep patterns were associated with increased reports of anxiety and depression.
• Prostate-metastatic patients exhibited a higher risk of vulnerability than other cancer types.

Conclusions:

• Wearable sensors and AI can enhance the analysis of dynamic risk factors in cancer care.
• Personalized risk models integrating wearable and self-reported data are crucial for evolving patient needs.
• Understanding the interplay between lifestyle patterns and psychological outcomes improves supportive cancer care.