A comparative study of full hindlimb flexion in horses: 5 versus 60 seconds

Abstract

The flexion test is routinely used in lameness and pre-purchase examinations. There is no accepted standard for duration of flexion or evidence that interpretation of results would differ with different durations of flexion. We hypothesized there would be no difference in interpretation of full hindlimb flexion for 5 or 60 seconds. Video recordings of lameness examinations of 34 client owned horses were performed that included: baseline lameness, upper hindlimb flexion for 60 seconds, and flexion of the same leg for 5 seconds. Videos were edited to blind reviewers to the hypothesis being tested. The baseline lameness video from each horse was paired with each flexion to make 2 pairs of videos for each case. Twenty video pairs were repeated to assess intra-observer repeatability. Fifteen experienced clinicians reviewed the videos and graded the response to flexion as either positive or negative. Potential associations between the duration of flexion and the likelihood of a positive flexion test were evaluated using generalized linear mixed models. A kappa value was calculated to assess the degree of intra-observer agreement on the repeated videos. Full hindlimb flexion of 60 seconds was more likely to be called positive than flexion of 5 seconds (p<0.0001), with the likelihood of the same interpretation 74% of the time. The first flexion performed was more likely to be called positive than subsequent flexions (p=0.029). Intra-assessor agreement averaged 75% with κ=0.49. Full hindlimb flexion of a horse for 5 seconds did not yield the same result as 60 seconds.