Accuracy of intermittent fluoroscopy to detect intravascular injection during transforaminal epidural injections

Prospective validity study.

To determine how accurately intermittent fluoroscopy detects inadvertent intravascular injection during transforaminal epidurals.

Serious morbidity caused by transforaminal epidural injections is frequently related to inadvertent vascular injection of corticosteroids. Several methods have been proposed to reduce the risk of vascular injection, but none have demonstrated efficacy. Because of the fleeting appearance of vascular contrast patterns, live fluoroscopy is recommended during contrast injection. Despite this, many practitioners continue to use intermittent fluoroscopy.

During 50 epidural injections dynamic contrast flow was observed under live fluoroscopy, and the "dynamic true" image was determined. Two intermittent fluoroscopy images were saved from each injection, the first just before completing the contrast injection ("static C" image), and another 1 second after the contrast injection ceased ("static PC" image). Five physicians with experience performing these injections independently interpreted the 100 randomly ordered static images. Accuracy of intermittent fluoroscopy was determined by comparing the interpretation of the 100 static images with the dynamic true patterns observed under live fluoroscopy.

Overall, interpretation of the static images missed 57% of the vascular injections. Timing of the static images influenced accuracy with the static C images missing 50% of vascular injections, and the static PC images missing 68% of vascular injections (P = 0.075). Accuracy was significantly worse when vascular injections occurred simultaneous to the expected epidural injection (P = 0.041), and in lumbar images (P = 0.012).

Based on these findings, we recommend use of live fluoroscopy to observe dynamic contrast flow during transforaminal epidural steroid injections.