Associations Between Smartphone-Derived Behavioral Data and Rheumatoid Arthritis Flares

Area of Science:

• Rheumatology
• Digital Health
• Data Science

Background:

• Rheumatoid arthritis (RA) flares significantly impact patient quality of life.
• Early detection of RA flares is crucial for effective disease management.

Purpose of the Study:

• To characterize rheumatoid arthritis (RA) flare frequency and severity.
• To assess if short-term changes in patient-reported outcomes (PROs) can signal RA flares.
• To examine the relationship between passive smartphone data and self-reported RA flares.

Main Methods:

• PROs were collected in conditional and fixed phases, alongside passive smartphone data (mobility, communication).
• Logistic generalized estimating equation models and multivariate analyses were used to assess associations.
• Demographic and seasonal confounders were adjusted for in the analyses.

Main Results:

• Flares were linked to worse PROs, including patient global assessment and pain.
• Increased mobility and longer text messages were associated with fewer flares.
• Slower reaction times and shorter text messages correlated with increased flare occurrence.

Conclusions:

• Integrating PROs with passive smartphone data reveals novel associations with RA flare occurrence.
• These findings suggest potential for developing predictive models for earlier RA flare recognition.
• Personalized algorithms could improve future rheumatoid arthritis disease management.