BackgroundThe responsiveness and the intertester reliability of the Finnish Canine Stifle Index (FCSI) were tested, and a cut-off between compromised and severely compromised performance level was set.MethodsThree groups of dogs were used, 29 with any stifle dysfunction (STIF), 17 with other musculoskeletal disease except stifle (OTHER) and 11 controls (CTRL). All dogs were tested with the FCSI by the same physiotherapist at three occasions, at baseline, at six weeks and 10 weeks, and once also by another physiotherapist.ResultsDogs in the STIF group demonstrated significantly higher (P<0.001) FCSI scores than in OTHER or CTRL groups at baseline. Only the STIF group showed a significant (P<0.001) change in FCSI score at all time points, indicating responsiveness to change. There were no significant differences between the evaluators (P=0.736), showing good intertester reliability, supported by moderate to good (0.78) intraclass correlation coefficient (ICC). The evaluator performing the FCSI did not have a significant effect when comparing the groups of dogs (P=0.214). The 95 per cent confidence intervals of the ICC per group were 0.79 (0.60, 0.91) for STIF, 0.83 (0.53, 0.96) for OTHER 0.78 (0.64, 0.88) for all dogs. A cut-off differentiating a severely compromised from a compromised performance was set at 120, having sensitivity of 83 per cent and specificity of 89 per cent.ConclusionThe FCSI is a recommendable measure of dogs’ stifle functionality.
USING CONFIDENCE INTERVALS TO OBTAIN A FAMILY OF ESTIMATORS OF THE INTRACLASS CORRELATION COEFFICIENT (OR HERITABILITY)
