The role of ultrasonography in the investigation and management of rheumatic conditions

Aims: The importance of inflammation in the development of joint damage and subsequent functional limitation, together with the increasing awareness that clinical assessment is insensitive at detecting synovitis, has led to the use of modern imaging modalities such as ultrasonography (US) in rheumatology. This with an emphasis on improved detection of synovitis and earlier, more effective therapeutic intervention. The aim of this thesis was to investigate the role of US in routine practice, by validating its role in diagnosis, prognosis and therapeutic intervention, across a range ofcommon rheumatological clinical scenarios. Methods: Construct validity for US-detected synovitis in knee arthritides was examined by comparison with clinical assessment and arthroscopy. In order to examine the benefits of US-guided joint injections, the accuracy of guided shoulder injections was observed. As well, the efficacy of US-guided corticosteroid injections in hip osteoarthritis and predictors of outcome were also assessed. The sensitivity of US over clinical examination was assessed in a cohort of rheumatoid arthritis patients with low disease activity levels. The diagnostic and therapeutic predictive value of US-detected synovitis in inflammatory hand pain was examined in a large cohort. Finally the clinical utility of US at altering diagnosis and management in clinical rheumatology practice was prospectively examined. Results: For detection of knee synovitis, using arthroscopy as the gold-standard, US had higher sensitivity (98 vs 85%) and specificity (75 vs 25%) than clinical assessment. Kappa values for inter- and intra-reader reliability of US were 0.71 and 0.85 respectively. In the shoulder US-guided injections were 100% accurate, but 55% had evidence of extrabursal contrast limited to the tissue planes adjacent to the bursa. Outcome following hip injection was poor, but synovitis without osteophytes on US was the best predictor of short term benefit. In rheumatoid arthritis patients in clinical remission, the majority satisfying established remission criteria, US detected gray-scale and power Doppler (PD) synovitis in 85% and 60% patients respectively. The kappa value for inter-reader reliability was 0.60 for gray-scale, and intra-reader reliability was 0:60 for gray-scale and 0.62 for PD.Most inflammatory hand pain patients without clinical evidence of an inflammatory arthritis had US synovitis (55%); and US (p