Learning and reliability of colour Doppler ultrasound in giant cell arteritis

To evaluate a standardized training program and a reliability exercise in colour Doppler ultrasound (CDUS) for giant cell arteritis (GCA).

Two workshops were conducted in 2007 and 2008 to train rheumatologists in CDUS for GCA ultrasound diagnosis. Twenty-nine and forty-three participants without previous experience in GCA ultrasound were admitted in 2007 and 2008, respectively. First, some theoretical knowledge about GCA ultrasound signs was provided; second, a reader evaluation session of temporal artery video recording examinations of 27 and 30 patients were projected in the 2007 meeting and the 2008 workshop, respectively (50% were cases and 50% were controls). Twenty-four cases were common to both reader sessions. A mean of six videos were shown of each patient. Each video had to be assessed as normal or pathologic. Finally, hands-on scanning training was performed. To assess the efficacy of the workshop: 1) a structured satisfaction questionnaire was graded (1-5 Likert scale), and 2) the reliability, specificity, and percentage of correctly classified cases by each participant were calculated.

The kappa coefficient of inter-reader agreement for the 29 and 43 participants was excellent (Kappa: 0.846) in 2007 and (Kappa: 0.848) in 2008. The intra-reader kappa result was also excellent (Kappa: 0.950). The satisfaction, sensitivity, specificity, and percentage of correctly classified patients and controls were very high.

The proposed learning method seemed to be effective and well accepted by the target audience. The inter-reader reliability of GCA ultrasound was excellent. These encouraging results support the need for planned standardized training programs.