Does thermal imaging correlate with musculoskeletal examination in the identification of inflamed joints in children and young people with juvenile idiopathic arthritis? A prospective diagnostic accuracy study

Background: Evidence demonstrates early, aggressive treatment improves outcomes in juvenile idiopathic arthritis (JIA). Early diagnosis of joint inflammation rests largely on expert clinical examination, supported by radiological investigations. Thermal imaging (TI) is easily accessible and is useful in detecting joint inflammation in adult rheumatoid arthritis. This pilot study aimed to assess if high-resolution TI correlates with musculoskeletal examination in identification of changes in skin temperature and joint inflammation. Methods: Participants aged 4-16 years with a diagnosis of JIA and suspected knee joint inflammation (uni- or bilateral) were included. Exclusion criteria comprised presence of other conditions such as infection, which could confound the skin temperature signal. Participants were permitted to continue with their prescribed medicines. TI data and musculoskeletal examination findings were recorded on the same occasion (within a one hour time period) for both knee joints, prior to the participant receiving intra-articular steroid injections as part of their routine care. Participants acclimatised to room temperature for 10 minutes whilst sitting, ensuring no contact over the knees. TI data of the front and back of both knees was subsequently obtained in standing. A member of the clinical team then performed a detailed examination of knee joints documenting presence/absence of warmth, swelling and restriction. The research and clinical teams were blinded to suspicion. This study received full ethical approval, and all participants gave their written informed assent or consent before taking part. Results: Eleven participants were included (Mage = 10.6 years, range 5-15 years; 6 females). Correlation between TI (front and back) and musculoskeletal examination data was seen in participants 3, 5 and 10. Partial correlation between TI (front or back) and musculoskeletal examination was seen in participants 1, 4, 9 and 11. Participant 8 had a fixed flexion knee deformity and abnormal gait pattern; TI and musculoskeletal examination data did not correlate for this individual. Conclusion: We describe novel data comparing musculoskeletal examination alongside TI recordings. Skin temperature did not always correlate with examination findings in this small cohort and we plan to compare TI with established imaging techniques (contrast-enhanced MRI and ultrasound) in clinically active JIA to explore this further.