A comparison of ultrasound and clinical examination in the detection of flexor tenosynovitis in early arthritis

BACKGROUND Tenosynovitis is widely accepted to be common in rheumatoid arthritis (RA) and postulated to be the first manifestation of RA, but its true prevalence in early disease and in particular the hand has not been firmly established. The aims of this study were first to investigate the frequency and distribution of finger flexor tenosynovitis using ultrasound in early arthritis, second to compare clinical examination with ultrasound (US) using the latter as the gold standard. METHODS 33 consecutive patients who had who were initially diagnosed with polyarthritis and suspected of polyarthritis and clinical suspicion of inflammatory arthritis of the hands and wrists were assessed during consecutive, routine presentations to the rheumatology outpatient clinic. We scanned a total of 165 finger tendons and subsequent comparisons were made using clinical examination. RESULTS Flexor tenosynovitis was found in 17 patients (51.5%) on ultrasound compared with 16 (48.4%) of all patients on clinical examination. Most commonly damaged joint involved on US was the second finger followed by the third, fifth, and fourth. Both modalities demonstrated more pathology on the second and third metacarpophalangeal (MCP) compared with the fourth and fifth MCP. A joint-by-joint comparison of US and clinical examination demonstrated that although the sensitivity, specificities and positive predictive values of clinical examination were relatively high, negative predictive value of clinical examination was low (0.23). CONCLUSIONS Our study suggest that clinical examination can be a valuable tool for detecting flexor disease in view of its high specificity and positive predictive values, but a negative clinical examination does not exclude inflammation and an US should be considered. Further work is recommended to standardize definitions and image acquisition for peritendinous inflammation for ultrasound.