Detection of flares by decrease in physical activity, collected using wearable activity trackers, in rheumatoid arthritis or axial spondyloarthritis: an application of Machine-Learning analyses in rheumatology.

BACKGROUND: Flares in rheumatoid arthritis (RA) and axial spondyloarthritis (axSpA) may influence physical activity. The objective was to assess longitudinally the association between patient-reported flares and activity-tracker-provided steps per minute, using machine-learning.

METHODS: This prospective observational study (ActConnect) included patients with definite RA or axSpA. During 3 months, physical activity was assessed continuously by number of steps/minute, using a consumer grade activity tracker, and flares were self-assessed weekly. Machine-learning techniques were applied to the dataset. After intra-patient normalization of the physical activity data, using multiclass Bayesian methods, sensitivities, specificities and predictive values of the machine-generated models of physical activity to predict patient-reported flares were calculated.

RESULTS: In all, 155 patients (1339 weekly flare assessments and 224,952 hours of physical activity assessment) were analyzed: for RA (N=82) and axSpA (N=73) patients respectively, mean age was 48.9±12.6 and 41.2±10.3 years; mean disease duration was 10.5±8.8 and 10.8±9.1 years; 14 (17.1%) and 41 (56.2%) were males. Disease was well-controlled (mean DAS28: 2.2±1.2; mean BASDAI: 3.1±2.0) but flares were frequent (22.7% of all weekly assessments). The model generated by machine-learning performed well against patient-reported flares (mean sensitivity: 96% [95% confidence interval 94-97%], mean specificity: 97% [96-97%], mean positive and negative predictive value: 91% [88-96and 99% [98-100%]). Sensitivity analyses were confirmatory.

CONCLUSION: Although these pilot findings will have to be confirmed, the correct detection of flares by a machine learning processing of activity trackers data opens the way for future studies of remote-control monitoring of disease activity, with great precision and minimal patient burden. This article is protected by copyright. All rights reserved.