P305 Validating diagnostic GCA ultrasonography skills against an expert sonographer

Abstract

Abstract Background/Aims Ultrasonography has become the recommended first line investigation for suspected giant cell arteritis (GCA). However, it is still not yet widely available in the UK due to a paucity of skilled rheumatologists. Effective training programmes have been developed, but subsequent hands-on practice is needed to become truly proficient. There is no consensus on the logistics of gaining the ‘appropriate’ level of skill. This project demonstrates a potential mechanism for validation against an expert sonographer. Methods Over a 6-month period, 12 scans of patients referred with suspected GCA were sequentially performed by first a trainee (FC) and then an expert sonographer (CM) on the same day. FC had theoretical knowledge and supervised practical experience but had not scanned independently prior to this project. CM has performed >1000 scans. 6 vascular territories considered to be the core GCA US data set (common, frontal and parietal branches of the superficial temporal artery, and the 1st, 2nd and 3rd part of the axillary artery) were marked as either ‘normal’ or ‘halo’. CM was blinded to FC’s results. Inter-observer reliability was calculated for all vascular territories using Cohen’s kappa. Results A total of 144 vascular territories were scanned in 12 patients. Table 1 details the inter-observer variation, with a near perfect level of agreement demonstrated, kappa=0.90 (95% CI 0.83 - 0.98). Conclusion This is the first study that demonstrates a mechanism for validation of ultrasonography skills which will allow GCA US to become a feasible reality in clinical practice. This project demonstrates that after acquiring supervised training, near perfect levels of agreement can be arrived between a trainee and expert. Disclosure F.L. Coath: None. C. Mukhtyar: None.