Inter-tester reliability of an ultrasound protocol for Knee-Apophysitis

Intro
Apophysitis of the knee (Sinding-Larsen-Johansen and Morbus Osgood-Schlatter) are prevalent in physically active adolescents. The sonographic composite Flaviis severity score have shown prognostic value and other sonographic measures have been proposed to characterize the morphology of these conditions. We therefore investigated inter-tester reliability of the Flaviis score and other ultrasound measures specific for knee apophysitis.

Methods
Two ultrasound operators received standardized training, and each scanned 27 knees diagnosed with apophysitis (n=17, 13.4±1.6 years, Tegner score 9) in a random order, following a standardized protocol including the Flaviis score and 16 different measures. Thirty-two activity-matched asymptomatic knees (n=18, 12.5±1.9 years, Tegner score 9) were included to ensure valid blinding of potential diagnosis, symptoms, examination findings, and each other’s assessments. Measures were deemed reliable if kappa exceeded 0.40 (moderate agreement).

Results
The following measures were reliable: Flaviis severity score (κ=0.60) (composed of distal tendon thickening [κ=0.76], tibial tubercle cartilage thickening [κ=0.49], free ossicles [κ=0.77], hypoechogenic nucleus [κ=0.76], thickening of the infrapatellar bursa [κ=0.23), maturation stage (1-4 score) (κ=0.61), apex patella cartilage thickening (κ=0.41), proximal tendon thickening (κ=0.50), and color-doppler (apex patella [κ=0.50], distal tendon [κ=0.48], tibial tubercle [κ=0.50]). Measures of fragmentation (1-4 scale) (κ=0.20-0.24; 41-56% agreement; p=0.28-0.5), color-doppler of distal tendon and infrapatellar bursa (1-4 scale) (κ=0.27-0.32; 48-93% agreement; p<0.001), and thickening of the infrapatellar bursa were not reliable.

Conclusion
The Flaviis score and most of its components (4/5) and other knee apophysitis measures (6/11) showed moderate to substantial inter-tester re