Feasibility, Safety, and Preliminary Effect Estimates of an Ultrasound‐Informed Hybrid Joint‐Sparing Rehabilitation Strategy in Patients With Rheumatoid Arthritis: A Prospective Non‐Randomized Proof‐of‐Concept Study

BACKGROUND: and PurposePhysical rehabilitation is recommended as an adjunct to pharmacological treatment in rheumatoid arthritis (RA); however, exercise prescription remains challenging when synovitis is active. Musculoskeletal ultrasonography (MSK-US), particularly power Doppler (PD), provides a sensitive assessment of synovial inflammatory activity and may help inform rehabilitation planning. The primary objectives were to describe the feasibility and safety of a 12-week ultrasound-informed hybrid joint-sparing rehabilitation protocol in patients with RA. Pain intensity (VAS) was the designated primary outcome for generating preliminary effect estimates, with ultrasonographic indices as secondary outcomes. No feasibility criteria were prespecified. METHODS: This prospective non-randomized proof-of-concept study, reported according to the TREND statement, enroled 60 consecutive patients fulfiling the 2010 ACR/EULAR criteria between February 2024 and January 2025. Participants were allocated by clinic attendance order (sequential recruitment) into a rehabilitation group (n = 30) or a control group (n = 30). The rehabilitation group received a 12-week ultrasound-informed hybrid protocol (supervised outpatient sessions plus a structured home programme) comprising weekly paraffin wax baths and baseline ultrasound-stratified exercise progression, alongside stable pharmacological therapy. Joint regions of the index (most affected) hand with baseline PD > 1 (68.9% of examined regions) were managed conservatively at programme initiation, with resistance loading deferred in favour of pain-free range-of-motion exercises. The control group received stable pharmacological therapy plus a single educational session. Pain was assessed using a 10-cm visual analogue scale (VAS). MSK-US of wrist, MCP, and PIP regions was performed at baseline and Week 12 using OMERACT semi-quantitative scoring. The distributions of individual ΔVAS values showed no overlap between groups: the smallest improvement in the rehabilitation group (1.4 cm) exceeded the largest improvement in the control group (1.1 cm); control-group changes were unusually uniform and the effect size was probably overestimated. Between-group changes in total grey-scale synovitis (p = 0.327) and total PD (p = 0.340) were not statistically significant; no erosion progression was detected. No serious adverse events occurred; safety conclusions were limited to 12 weeks. Baseline DAS28 and VAS were comparable, although the rehabilitation group was older with longer disease duration and higher ESR. DISCUSSION: In this small, single-centre, non-randomized study, the protocol could be delivered as planned and no serious adverse events were observed over 12 weeks (no feasibility criteria were prespecified). Pain reduction was greater than with minimal education, but the difference is probably overestimated, may reflect non-specific effects of much greater therapist contact, and cannot be attributed to ultrasound-informed prescription; ultrasonographic synovitis and vascularity did not differ between groups. These preliminary, hypothesis-generating findings should inform, cautiously, the design of future randomized trials with active comparators.