目的：开发首个类风湿性关节炎（RA）肌腱损伤的超声评分系统，并评估其观察者内和观察者间的可靠性。
方法：我们在35名具有肌肉骨骼超声经验的国际风湿病学家中，对RA中超声定义的肌腱损伤和超声评分系统进行了Delphi研究。包括12名RA患者，并由12名风湿病学家-超声医师进行了两次评估。在B模式下对5个腕部伸肌区室（桡侧腕肌短骨和长肌；四肢伸肌；指公伸肌；最小伸肌；尺侧腕肌）和一个踝肌腱（胫骨后肌）进行了盲目的超声检查，每个患者的独立和双侧。通过κ系数计算观察者内和观察者间的可靠性。
结果：在B模式下对肌腱损伤评分采用三级半定量评分系统。肌腱损伤评分的平均观察者内部可靠性优异（κ值0.91）。平均观察者间可靠性评估显示出良好的κ值(κ值0.75)。最可靠的是指伸肌的最小值，尺侧腕伸肌,和胫骨后腱。还开发了腱鞘炎和肌腱损伤的超声参考图像图谱。
结论：超声是评估RA肌腱损伤的可重复工具。这项研究有力地支持了一种新的可靠的肌腱损伤超声评分系统。
OBJECTIVE: To develop the first ultrasound scoring system of tendon damage in rheumatoid arthritis (RA) and assess its intraobserver and interobserver reliability.
METHODS: We conducted a Delphi study on ultrasound-defined tendon damage and ultrasound scoring system of tendon damage in RA among 35 international rheumatologists with experience in musculoskeletal ultrasound. Twelve patients with RA were included and assessed twice by 12 rheumatologists-sonographers. Ultrasound examination for tendon damage in B mode of five wrist extensor compartments (extensor carpi radialis brevis and longus; extensor pollicis longus; extensor digitorum communis; extensor digiti minimi; extensor carpi ulnaris) and one ankle tendon (tibialis posterior) was performed blindly, independently and bilaterally in each patient. Intraobserver and interobserver reliability were calculated by κ coefficients.
RESULTS: A three-grade semiquantitative scoring system was agreed for scoring tendon damage in B mode. The mean intraobserver reliability for tendon damage scoring was excellent (κ value 0.91). The mean interobserver reliability assessment showed good κ values (κ value 0.75). The most reliable were the extensor digiti minimi, the extensor carpi ulnaris, and the tibialis posterior tendons. An ultrasound reference image atlas of tenosynovitis and tendon damage was also developed.
CONCLUSIONS: Ultrasound is a reproducible tool for evaluating tendon damage in RA. This study strongly supports a new reliable ultrasound scoring system for tendon damage.
