Development of a Mathematical Model to Predict Skin to Lumbosacral Epidural Distance in Dogs

Abstract

Epidural anaesthesia is frequently the pillar for pain relief in dogs undergoing surgery of the hindlimb or abdomen, but alarmingly, failure rates have been reported as high as 32% even for experienced veterinary anaesthetists. In human anaesthesia, development of equations to predict the skin to epidural distance has shown promise in improving epidural success alongside the use of conventional epidural techniques. It is therefore feasible that the knowledge of a predictive skin to lumbosacral epidural distance may positively influence failure rates in dogs. This thesis aims to identify correlation relationships of canine external body variables with the skin to lumbosacral epidural distance. Predictive equations were developed based on canine computed tomography images, and dog cadavers were then used to further validate the mathematical equations. Finally, a randomised prospective study was conducted, and a non-inferiority analysis was performed to determine the effectiveness of the mathematical equation in a practical clinical setting.