SummaryTransforaminal epidural injection is an effective method for treating spinal pain but can cause devastating complications that result from accidental vascular uptake of the injectate or a direct vascular injury. We prospectively evaluated the patient factors that might be associated with intravascular uptake during transforaminal epidural injections. A total of 2145 injections were performed on 1088 patients under contrast-enhanced real-time fluoroscopic guidance. The collected data included the patient's age, sex, body mass index, diagnosis, injection level, side of injection, history of spinal surgery at the targeted level, and the number of injections at the targeted site. The overall incidence of intravascular injection was 10.5% (224 ⁄ 2145). The highest incidence was at the cervical level (28 ⁄ 136; 20.6%), followed by the sacral level (111 ⁄ 673; 16.5%), the thoracic level (23 ⁄ 280; 8.2%) and the lumbar level (64 ⁄ 1056; 6.1%). The difference was significant for the cervical and sacral level compared with the lumbar and thoracic levels (p < 0.001). Intravascular injection was not associated with the other patient characteristics studied. Transforaminal epidural injection is an effective method for treating radicular pain [1] and is superior to other forms of therapeutic epidural injections [2,3]. This is a target-specific approach that facilitates better ventral epidural flow to the target site with a smaller volume of injectate. Although serious complications are rare [4,5], there are several reports of paraplegia or quadriplegia from spinal cord infarction [6,7], brainstem infarction [8,9], cortical blindness [10], coma or death [7,11]. The presumptive reasons for these serious complications are needle-induced direct mechanical injuries or vascular spasm in the vertebral or radicular arteries, as well as embolus formation of the particulate steroid [6,7,12,13]. In previous reports, the incidence of accidental intravascular needle placement during fluoroscopically-guided transforaminal epidural injection ranged from 9.9% to 32.8% and varied according to the spinal level [14][15][16][17]. It is recommended that contrast-enhanced real-time fluoroscopy be used to confirm correct needle placement and prevent intravascular injection during the procedure [15,18]. Knowledge of the associated factors may also help to reduce the incidence of intravascular needle placement but there is limited information on these. In particular, there are no reports on whether the patient's previous spinal surgery, body mass index (BMI) and repeated injection at the same site influence the risk of intravascular uptake. The incidence of intravascular injection during transforaminal injection in the thoracic region has not been reported. We therefore prospectively evaluated the incidence and factors associated with intravascular injection during transforaminal epidural injection.