Deliberate epidural air injection in dogs: a radiographic study.

To investigate whether and to what extent air bubbles are present in the epidural space after injections of air, the authors designed an experiment using a chronic dog model. The epidural space was entered at the sixth or seventh lumbar interspace in six mongrel dogs (weight, 20 to 27 kg). Loss-of-resistance (LOR) to saline was tested to identify the epidural space. Three milliliters of air were injected via a Tuohy needle, followed by 5 ml of contrast material (Isovue 300, Squibb Diagnostics). Anterior-posterior (AP) and lateral radiographs were taken on day 1. On days 2, 4, and 8, the procedure was repeated, omitting the air injection. Radiographs were read by a neuroradiologist. Air bubbles were radiographically demonstrated in all six dogs on day 1, and in three of six dogs on day 2. No air was seen on later films. The air bubbles seen were small and few in number but often located near intervertebral spaces. The authors' findings suggest that injection of large volumes of epidural air should be avoided, particularly when N2O is used as part of the anesthetic or when performing diagnostic peridurography since N2O may expand the volume of existing bubbles and interfere with filling of the epidural space by local anesthetics or contrast material.