Augmenting Prognostication: Utilizing Activity Trackers to Enhance Survival Prediction in Metastatic Non-Small Cell...

Area of Science:

• Oncology
• Digital Health
• Wearable Technology

Background:

• Performance status (PS) assessment is crucial for cancer treatment decisions but is subjective.
• Activity trackers can collect objective patient-generated data to improve PS assessment.
• Current PS assessment methods are prone to miscommunication between patients, physicians, and families.

Purpose of the Study:

• To evaluate the prognostic ability of activity tracker data in patients with metastatic non-small cell lung cancer (mNSCLC).
• To explore novel health indices derived from activity tracker data.
• To compare the prognostic performance of wearable sensor data with traditional PS assessment.

Main Methods:

• Prospective, observational feasibility study of 55 mNSCLC patients using Fitbit® activity trackers.
• Collected data included steps, distance, heart rate, and activity intensity.
• Defined and analyzed novel metrics: Heart rate Activity zone Mismatch (HAM) and excessive Sedentary Heart Rate (eSHR).
• Used multivariable Cox proportional hazards models and Harrell's concordance index (C-index) for survival analysis.

Main Results:

• Activity tracker metrics, including daily steps and novel indices (HAM, eSHR), were significantly associated with overall survival (OS).
• The prognostic model incorporating PS-physician assessment showed the highest concordance (C-index = 0.75).
• Models using daily distance (C-index = 0.74) and steps (C-index = 0.73) also demonstrated strong prognostic ability.

Conclusions:

• Wearable activity tracker data are prognostic for survival in mNSCLC patients.
• These objective measures may serve as valuable supplements or alternatives to subjective PS assessments.
• Activity trackers hold potential for enhancing clinical trial design and patient management in oncology.