Gait scoring in dogs with thoracolumbar spinal cord injuries when walking on a treadmill

BACKGROUND An inexpensive method of generating continuous data on hind limb function in dogs with spinal cord injury is needed to facilitate multicentre clinical trials. This study aimed to define normal fore limb, hind limb coordination in dogs walking on a treadmill and then to determine whether reliable data could be generated on the frequency of hind limb stepping and the frequency of coordinated stepping in dogs with a wide range of severities of thoracolumbar spinal cord injury. RESULTS Sixty-nine neurologically normal dogs of different body sizes including seven lame dogs were videotaped walking on the treadmill without prior training and all used the lateral gait of right fore, left hind, left fore, right hind (RF-LH-LF-RH). Severely paraparetic dogs were able to walk on the treadmill for a minimum of 75 seconds, scoring of which generated data representative of function in animals with extremely variable gaits. Fifty consecutive stepping cycles were scored by three observers in 18 dogs with a wide range of disability due to acute thoracolumbar spinal cord injury using a stepping score (hind limb steps/fore limb steps ×100), and a coordination score (coordinated hind limb steps/total hind limb steps ×100). Dogs were also scored using a previously validated ordinal open field score (OFS). Inter- and intraobserver agreement was high as assessed with Cronbach's alpha test for internal reliability. The stepping and coordination scores were significantly correlated to each other and to the OFS. CONCLUSIONS Dogs with naturally occurring spinal cord injury can walk on a treadmill without prior training and their hind limb function can be scored reliably using a stepping score and coordination score. The only requirements for data acquisition are a treadmill and appropriately positioned video camera and so the system can be used in multicentre clinical trials to generate continuous data on neurologic recovery in dogs.