Objectives This study aimed to determine whether gout flare resolution is associated with physical activity and sleep. These findings were intended to inform the future development of a definition for flare resolution. Methods Data from 81 participants with gout who experienced > 1 flare during a 6‐month prospective observational study were included. Daily activity (step count) and sleep (total sleep time and deep sleep percentage) were collected remotely using wrist‐worn trackers. Daily flare occurrence was captured using smartphone diaries. Activity/sleep outcomes during each flare episode were compared to those during three resolution timeframes: T1: the first day following the flare; T2: the consecutive window immediately following the flare and equal in length to the flare; and T3: the consecutive window equal in length to the flare and immediately following a washout window (equal in length to the flare). Adjusted differences in medians between flare and resolution timeframes were estimated using linear quantile mixed models. Results Daily step count was lower during flares compared with all resolution timeframes (all P < 0.033). T3 provided the best overall model fit, suggesting that resolution is most clearly distinguished from the flare when using this timeframe. No significant differences between flare and resolution were observed for the sleep outcomes for any resolution timeframe. Conclusions Physical activity improves soon after flare cessation, with the clearest distinction between flare and resolution observed over a longer post‐flare recovery period, whereas sleep metrics do not differentiate flare from resolution states. These findings support physical activity as a useful objective domain for gout flare resolution.
Objective Our objective was to evaluate the ability of different large language models (LLMs) to translate and analyze qualitative interviews describing pregnancy experiences among women with systemic lupus erythematosus (SLE). An additional objective was to compare themes identified by LLMs with those identified by human researchers. Methods We analyzed 26 semi‐structured interviews originally conducted in Portuguese by Rodrigues et al. among women with SLE exploring their perinatal care experiences. Three prompting strategies (zero‐shot, few‐shot, and chain‐of‐thought) were used on the transcripts across four state‐of‐the‐art LLMs (GPT‐4o, Claude Sonnet 4.5, Gemini 2.5 Flash, and DeepSeek R) to generate key themes. Model outputs were evaluated using an LLM‐as‐a‐judge framework assessing thematic coverage, thematic unity, and semantic similarity. Results Across models and prompting strategies, LLMs generated similar themes that closely aligned with those identified by human researchers. LLM‐as‐a‐judge evaluations demonstrated high thematic unity and semantic similarity. Claude Sonnet 4.5 showed the most consistent performance. Few‐shot prompting consistently outperformed zero‐shot and chain‐of‐thought approaches. Conclusion LLMs show promise as supportive tools for analyzing qualitative research by preserving thematic coherence and capturing the complexity of pregnancy experiences in women with SLE. Although zero‐shot prompting is widely used, it produced lower thematic results in our study, whereas few‐shot prompting achieved the most comprehensive and coherent theme extraction. LLM‐assisted qualitative analysis may complement human interpretation and has the potential to support the inclusion of non‐English language data and diverse patient perspectives in lupus research. image
Generative artificial intelligence (AI) is increasingly being used in rheumatology practice, to assist with tasks such as documentation, record summarization, and clinical decision support. Rheumatologists need a practical approach to determining what evidence and oversight each application requires. This narrative review proposes a framework based on two dimensions: contextual complexity and influence on patient care. Greater contextual complexity can make errors harder to detect, whereas greater influence on care increases the potential consequences of an undetected error. We examine evidence for workflow benefits and clinical effectiveness within this framework and consider safety, equity, privacy, governance, and implementation. Existing evidence on the benefits, accuracy and reliability of AI applications remains mixed. For example, studies of ambient AI scribes suggest reduced documentation burden, but evidence of improved patient outcomes from applications supporting clinical decisions is more limited. For applications with limited clinical influence, evaluation has initially focused on workflow, usability, and documentation quality. Tools that synthesize longitudinal information require additional assessment of completeness and source traceability. Those influencing diagnosis or treatment require prospective evaluation of clinical benefit and safety. Relevant vendor evidence should be assessed alongside proportionate local testing and ongoing monitoring. Safeguards should address risks such as hallucinations, automation bias, and propagating documentation errors, with attention to performance across patient groups. Changes in functionality or clinical use should prompt reassessment. The framework provides a structured basis for adoption decisions and identifies priorities for research on patient outcomes and feasible oversight in routine rheumatology care. image
Objectives This study synthesised the content and reporting completeness of published physical therapist‐led interventions for glenohumeral osteoarthritis (GHOA). Methods This systematic review undertook electronic database searches in PubMed, EMBASE, Cochrane, CINAHL and PEDro (until June 2025) for any peer‐reviewed articles/studies describing physical therapist‐led interventions across non‐surgical and perioperative settings for adults aged 45 years and over with GHOA. The content and reporting completeness were synthesized using the 22‐item Template for Intervention Description and Replication for rehabilitation interventions (TIDIeR_Rehab). Results Overall, 134 articles/studies describing 156 physical therapist‐led interventions were included, comprising mostly case series designs (n=47, 35.1%) or narrative reviews (n=38, 28.4%). Of the remaining 49 articles/studies, only 11 were studies reporting RCT results and three were protocols of RCTs in progress. Intervention content varied considerably, and the overall completeness of intervention item reporting was poor (only 18%). Interventions in non‐surgical contexts commonly included strengthening exercises (n=33, 53.2%), range of motion (ROM) exercises (n=23, 37.1%), and electrotherapy modalities (n=16, 25.8%). Sling immobilisation was most reported in postoperative interventions (n=64, 72.7%). Many different proposed mechanisms of action were reported (n=49, 31.4%); but all were biomechanically orientated. Conclusion Very few high‐quality empirical studies evaluating physical therapist‐led interventions for GHOA have been published. We found substantial heterogeneity in the content of published physical therapist‐led interventions, and the reporting completeness was generally poor. This suggests considerable uncertainty regarding best practice physical therapy for GHOA in terms of intervention content. Future high‐quality empirical research is needed, preferably using RCT designs that include clear descriptions of interventions.
Background Extent of uveitis risk in juvenile idiopathic arthritis (JIA) is based on clinical factors. We aim to compare the trajectories of tear‐based biomarker levels of patients with JIA with and without uveitis. Methods This prospective multicenter study included children with JIA at risk for uveitis (JIA‐no‐U), and with JIA‐associated uveitis (JIA‐U). Serial tear samples were collected from each eye by Schirmer strips at enrollment (E) and follow‐up (FU1, FU2). S100 proteins, cytokines, and chemokines were assayed by Luminex or ELISA and compared between groups. Findings Fifty patients (37 JIA‐no‐U, 13 JIA‐U) contributed 222 samples (150 JIA‐no‐U, 73 JIA‐U). At E there were no significant differences in biomarker levels between JIA‐U and JIA‐no‐U. During follow‐up of JIA‐U, monocyte chemoattractant protein‐1 (MCP‐1) levels increased longitudinally from E to FU1 and FU2 (1.63±0.31 to 2.46±0.31, p=0.005; and 2.89±0.32, p<0.0001), similar in Chemokine (C‐X‐C motif) Ligand (CXCL)‐10 from E to FU1 and FU2 (4.39±0.41 to 5.26±0.41, p = 0.030; and 6.51±0.43, p < 0.0001) and in IL‐8 levels from E to FU1 and FU2 (3.09±0.30 to 4.25±0.30, p = 0.006; and 4.65±0.32, p ≤ 0.001). In both groups, S100A12 and S100A8/A9 increased from E to FU. Conclusion Longitudinal changes in MCP‐1, CXCL10 and IL‐8 levels in tear fluid in JIA‐U may reflect inflammatory ocular processes and support the role of monocytes and interferon related pathways in uveitis. Lack of differences in S100 proteins may reflect systemic inflammation from arthritis. Tear‐based biomarkers may support clinical factors for identifying children with JIA who develop uveitis.
