PR_198: Accuracy of Intermittent Fluoroscopy to Detect Intravascular Injection During Transforaminal Epidural Injections

Abstract

Objective: To determine how accurately intermittent fluoroscopy detects accidental intravascular injection during transforaminal epidurals. Design: Comparison of intermittent verses live fluoroscopy. Setting: University spine clinic. Participants: 5 physiatrist interventionists independently interpreted 100 images from cervical and lumbar transforaminal epidural injections. Intervention: For each epidural, a 0.5mL contrast was injected under live dynamic fluoroscopy, and 2 images were saved. The first image, or “contrast” image, was saved just before completing the 0.5mL injection. The second image, or postcontrast image, was taken 1 second after contrast injection. The “true” contrast pattern was determined under live fluoroscopy at the time of each injection and classified as: epidural only, vascular only, simultaneous epidural and vascular, or nonepidural. All saved images were randomly presented to 5 interventionists not present during the injections. They were told to assume 0.5mL of contrast injection and asked to identify each image with any or all of 3 potential contrast patterns: epidural, nonepidural, and/or vascular. Main Outcome Measures: Accuracy of intermittent fluoroscopy was determined by comparing patterns identified in the saved images with the “true” contrast patterns. Results: On average, vascular contrast patterns were correctly identified in 50% of contrast and 32% of postcontrast images. When epidural and vascular injection occurred simultaneously, the vascular component was identified in 46% of contrast and 26% of postcontrast images. Overall, interpretation of contrast and postcontrast images accurately described the true pattern observed under live dynamic fluoroscopy in 43% of the images. Conclusions: The reported incidence of vascular penetration during transforaminal epidural injections is between 8.1% and 21.3%. Intermittent fluoroscopy will miss from 50% to 74% of these accidental vascular injections, depending on when the image is obtained and if the vascular injection occurs simultaneous to epidural injection. Live dynamic fluoroscopy during contrast injection is recommended to avoid vascular injection.