Predicting Six-Minute-Walking-Distance in COPD Patients From Wearable Devices in Real-World Setting.

Area of Science:

• Biomedical Engineering
• Digital Health
• Pulmonology

Background:

• Gait speed is a key health indicator, especially for older adults and those with chronic diseases.
• Consistent real-world gait speed measurement is lacking, posing a challenge for continuous health monitoring.
• Previous research established gait speed models in controlled environments.

Purpose of the Study:

• To test and validate gait speed prediction models in Chronic Obstructive Pulmonary Disease (COPD) patients using real-world data.
• To investigate the prediction of Six-Minute-Walking-Distance (6MWD) using gait speed data from wearable devices.
• To evaluate the effectiveness of integrating data from multiple wearable devices for gait speed assessment.

Main Methods:

• Collected real-world gait speed data from COPD patients using a smartphone, a smartwatch, and sensorized shoes.
• Applied established gait speed models to daily activity data from these wearable devices.
• Evaluated various combinations of wearable devices to determine their impact on prediction accuracy.

Main Results:

• Combining multiple wearable devices consistently enhanced gait speed predictions compared to single devices.
• The highest correlation (r = 0.71) and statistical significance (p = 0.0029) were achieved when using data from all three device types.
• Models demonstrated applicability to COPD patients during everyday activities, extending previous findings.

Conclusions:

• Multi-device integration is crucial for improving the accuracy of real-world gait speed assessment.
• Wearable sensor data holds significant potential for continuous and reliable monitoring of gait in chronic conditions like COPD.
• This approach offers a promising avenue for remote patient monitoring and personalized healthcare interventions.