Rater agreement for assessment of equine back mobility at walk and trot compared to quantitative gait analysis.

Abstract

BackgroundLameness assessment in horses is still predominantly performed using subjective methods. Visual assessment is known to have moderate to good intra-rater agreement but relatively poor inter-rater agreement. Little is known about inter- and intra-rater agreement on the evaluation of back motion, for which no objective measurement technique in a clinical setting is available thus far.ObjectivesTo describe inter- and intra-rater agreement of visual evaluation of equine back mobility.Study designRater reliability study using a fully crossed design in which all horses are rated by all observers. This data is compared with objective gait analysis.MethodsSeventy equine professionals (veterinarians and physiotherapists) and veterinary students evaluated videos of 12 healthy horses at walk and trot on a hard, straight line. Nine parameters related to back mobility were scored: general mobility, thoracic, lumbar, lumbosacral flexion and extension and left and right thoracolumbar latero-flexion. All parameters were compared with simultaneously measured quantitative motion parameters. After 1 month, six randomly chosen horses were re-evaluated by 57 observers.ResultsFor each parameter inter- and intra-rater agreements were calculated using intra-class correlation coefficients. For all parameters, inter-rater agreement was very poor (Main limitationsHorses were scored from videos and by lack of any existing (semi-) quantitative system, a custom-made system had to be used.ConclusionsThe poor inter- and intra-rater agreements of visual scoring of mobility of the equine back and the disagreement between subjective and objective gait analysis data, demonstrate the need for the development and introduction of objective, quantitative and repeatable techniques to assess equine back motion.