• intervertebral disc disease;
  • radiography;
  • dog

To assess the diagnostic accuracy of survey radiography for canine thoracolumbar intervertebral disc protrusion, survey radiographs (lateral and ventrodorsal) of 64 dogs with surgically-confirmed thoracolumbar intervertebral disc protrusion, 51 dogs with negative myelograms and 29 dogs with various spinal conditions other than disc protrusion were reviewed by three independent observers who were unaware of any clinical information. There were marked differences in observer performance for diagnosis of intervertebral disc protrusion, although there were no significant differences in intra-observer diagnostic accuracy for small vs. large dogs. Accuracy of observers for determining sites of intervertebral disc protrusion using survey radiography was in the range 51–61%. AH observers had low accuracy for identification of second sites of intervertebral disc protrusion. The most useful radiographic sign, narrowed intervertebral space, had only moderate sensitivity (range 64–69%) and moderate predictive value (range 63–71%) for intervertebral disc protrusion. Vacuum phenomenon was an infrequent but accurate sign of intervertebral disc protrusion. Recognition of multiple radio-graphic signs of intervertebral disc protrusion at one site was associated with increased accuracy of diagnosis. No observer was accurate enough to justify attempting targeted surgical treatment of intervertebral disc protrusion without myelography.