Force plate gait analysis at the walk and trot in dogs with low-grade hindlimb lameness

Abstract

To evaluate the accuracy of force plate gait analysis at the walk and trot in dogs with low-grade hindlimb lameness. Nineteen healthy dogs and 41 dogs with low-grade unilateral hindlimb lameness due to stifle or hip joint problems were walked and trotted over a force plate. Peak vertical forces (PVF) were recorded, and a symmetry index (SI) was calculated from the PVF of the hindlimbs. 'Cut-off' values were determined from the SI of the normal dogs. These cut-off values were used to discriminate lame dogs from normal ones. Sensitivity and specificity were evaluated for measurements at walk and trot, and the Cohen's Kappa coefficient (k) was used to determine the agreement between clinical lameness and force plate measurements, and between force plate results at walk and trot. Receiver Operating Characteristics (ROC) curve were plotted for both gaits to evaluate accuracy. The sensitivity of the measurements at walk was 0.63, and specifity was 0.95. The sensitivity of the measurements at trot was 0.90, and specificity was 1.0. Moderate agreement was found between force plate measurements at walk and trot, and between clinical gait assessment and force plate measurements at walk. Good agreement was found between clinical gait assessment and measurements at trot. ROC analyses revealed the trot (94.7% [91.7%; 97.7%]) to be the more accurate test than the walk (85.0% [80.1%; 89.9%]). The trotting gait was more sensitive and accurate than the walking gait for the differentiation of dogs with a low-grade hindlimb lameness from normal ones using force plate gait analysis.