Rates of lumbosacral transforaminal injections interpreted as intravascular: fluoroscopy alone or with digital subtraction.

We investigated the rate of injections interpreted as intravascular during imaging of lumbosacral transforaminal epidural injections, using fluoroscopy alone or with digital subtraction. We evaluated 732 injections performed on 348 patients: 8.1% (59/732) and 10.5% (77/732) of injections were interpreted as intravascular during fluoroscopy and digital subtraction, respectively, p = 0.13. The odds ratio (95% CI) for interpreting injections as intravascular increased for both fluoroscopy and digital subtraction fluoroscopy, with: each year of age, 1.04 (1.01-1.07) and 1.03 (1.00-1.06), p = 0.011 and 0.024, respectively; sacral compared with lumbar injections, 10 (5-19) and 8 (5-15), p < 0.001 for both. The odds ratio for intravascular injection increased with three other variables during digital subtraction fluoroscopy: spinal stenosis, 5.1 (1.5-17.1), p = 0.009; failed back surgery syndrome, 4.3 (1.2-15.8), p = 0.025; compression fracture, 8.0 (1.6-39.4), p = 0.011.