To evaluate the accuracy of a smartphone-based joint self-assessment app in measuring active rheumatoid arthritis compared to physician assessments.
Key Findings:
The app overestimated active disease compared to physician assessments.
Patient-reported joint counts showed a 95% positive predictive value (PPV) for low disease activity but only a 50% negative predictive value (NPV) for active disease, indicating limited reliability in detecting active disease.
Agreement between patient and physician assessments was modest, with correlation coefficients of 0.5 for tender joints and 0.33 for swollen joints.
Patients reported higher counts of tender and swollen joints than physicians, particularly for swollen joints.
Interpretation:
Patient-derived joint counts may be useful for confirming low disease activity or remission but are less reliable for identifying active disease.
Limitations:
Modest agreement may reflect real-world differences in patient environments and timing.
Difficulty in distinguishing pain from tenderness may affect reporting accuracy.
Comorbid conditions like fibromyalgia may influence patient-reported counts.
Limitations of app-based training without in-person or video instruction should be noted.
Conclusion:
While the app shows potential for confirming low disease activity, it is not reliable for accurately assessing disease activity in less well-controlled rheumatoid arthritis, highlighting the need for caution in its use.
