An assessment of intra- and interobserver agreement of reporting orthopaedic findings on presale radiographs of Thoroughbred yearlings

M. A. Jackson, A. L. Vizard, G. A. Anderson, J. S. Mattoon, R. B. Lavelle, B. T. Smithenson, N. V. Lester, A. F. Clarke and R. C. Whitton
September 2014
Equine Veterinary Journal

Reasons for performing study

Radiography is commonly used in clinical practice but agreement of reporting of radiographically detected orthopaedic findings in horses has rarely been studied.

To assess agreement within and between observers for reporting of orthopaedic findings on presale radiographs of Thoroughbred yearlings.
Study design

Retrospective analysis of archived radiographs.

Four veterinary radiology specialists each twice examined 167 sets of radiographs for orthopaedic findings in the fore feet, fore and hind fetlocks, carpi, tarsi and stifles. There were 27 findings analysed for agreement. Kappa statistic (κ), percentage of positive agreement (Ppos) and percentage of negative agreement are reported.

An excellent percentage of negative agreement was observed for all findings, with the exception of regular vascular channels of the proximal sesamoid bones. Ppos and κ results were variable. The presence of extra carpal bones, osseous cyst-like lesions of the ulnar carpal bone, sagittal ridge defects of the third metacarpus, fracture of the fore and hind proximal sesamoid bones, regular vascular channels in the hind proximal sesamoids, osteochondrosis lesions of the distal intermediate ridge and/or medial malleoli of the tibia, and osseous cyst-like lesions in the medial femoral condyle was consistently observed with an intra- and interagreement κ≥0.5 and Ppos≥50%. Lucency within the proximal sesamoids consistently had an observed intra- and interagreement κ0.4 and Ppos40%.

Observation of orthopaedic findings on yearling repository radiographs showed generally excellent agreement on the absence of findings, but variable agreement on the presence of findings. Agreement was good for larger and easy to categorise radiographic findings. More accurate definitions and training need to be developed to improve agreement within and between observers for orthopaedic findings with poor or fair to good agreement.