The growing availability of patient-generated health data is reshaping how rheumatoid arthritis (RA) can be monitored beyond the clinic. This narrative review examines how artificial intelligence (AI) supports remote RA monitoring through digital biomarkers, patient-reported tools, and connected technologies, with particular attention to patient engagement and implementation barriers. A structured PubMed-based search identified literature on RA, AI, machine learning, remote monitoring, wearables, smartphone applications, telemedicine, and digital health. Across the reviewed studies, AI was used mainly to interpret longitudinal patient-reported outcomes, activity-tracker data, and daily smartphone symptom reports. Collectively, these approaches suggest that remote monitoring can enrich between-visit assessment, support flare detection, and provide a more continuous picture of disease experience than episodic clinic review alone. Yet the evidence base remains uneven, with many studies limited by small samples, short follow-up, exploratory modeling, and heterogeneous outcomes. The literature also makes clear that the practical success of remote monitoring depends on more than algorithmic performance; usability, adherence, digital literacy, trust, privacy, workflow integration, and equitable access all shape whether these tools can function in routine care. The most credible near-term model is therefore not fully automated monitoring, but hybrid, clinician-linked systems co-designed with users. AI-supported remote monitoring is a promising extension of RA care, but its long-term value will depend on whether it can be validated rigorously and implemented in ways that are patient-centered, clinically meaningful, and socially inclusive.
Szymczyk et al. (Mon,) studied this question.