This study was a retrospective comparative review of chart data. Patient privacy and data confidentiality were maintained throughout the study. After the initial data collection, all patient identifiers were deleted from the dataset.

The study participants were patients who received a cervical FJ or TFE steroid injection at the outpatient Pain Centers of Ewha Womans University Seoul Hospital and Mokdong Hospital between 2019 and 2021. On the day of the procedure, before the injection, the patients were asked to fill out self-assessment questionnaires with their baseline information. To collect data and determine compliance with the inclusion criteria, the electronic clinical records and survey responses were retrospectively reviewed. We selected patients over the age of 18 who were diagnosed with FS based on their clinical profiles, medical examination results, and magnetic resonance imaging (MRI) results. Included patients also experienced cervical radicular pain caused by FS and underwent FJ or TFE injection to treat their symptoms.

The eligibility criteria were as follows: 1) cervical FS confirmed by MRI and concordant radicular pain along a specific nerve root in the upper arm; 2) the need for single-level FJ or TFE steroid injection; 3) cervical radicular pain for more than one month; and 4) a current pain score of 4 or more on a Numeric Rating Scale (NRS) of 0 (no pain) to 10 (worst pain imaginable) despite sufficient conservative treatment for at least one month, including physical therapy or oral medications such as pregabalin, gabapentin, tramadol, acetaminophen, or NSAIDs. The exclusion criteria were as follows: 1) the presence of psychiatric disorders; 2) evidence of vertebral fracture, tumor, pregnancy, or infection of the cervical spine; 3) coagulopathy; 4) allergy to iodinated contrast media; 5) laboratory results suggestive of inflammatory disease or rheumatoid disorders; 6) severe degenerative change in which the needle cannot approach the target area; and 7) cervical-referred pain characterized by axial neck pain, pain with pressure on the spinal column at the cervical facet joint level, and absence of neurologic symptoms.1415

We performed an FJ steroid injection aseptically using fluoroscopic guidance with a 22 ga, 2.5-inch spinal needle. Subjects were placed in the prone position, and the C-arm was rotated 90° to allow for visualization of the lateral FJ. The appropriate entry site was the lateral border of the cervical FJ, which was marked on the skin. The skin and subcutaneous tissue were anesthetized with 1% lidocaine. Using intermittent fluoroscopy, we advanced the needle via a lateral approach toward the FJ. The needle was not placed too far medially to avoid the spinal cord in anteroposterior (AP) fluoroscopy. Once the needle reached the FJ, we injected 0.5–1.0 mL of contrast media and acquired fluoroscopy images to exclude the intravascular position of the needle and confirm the adequate distribution of the contrast media (Fig. 1A and B). Then, we injected 1 mL of a solution mixed with 1% lidocaine and 2.5 mg dexamethasone. The subjects were observed for 30 minutes post-procedure.

The TFE steroid injection was aseptically performed using fluoroscopic guidance with a 22 ga, 2.5-inch Tuohy needle. The participants were in a supine position on a fluoroscopy table, and the C-arm was rotated between 45° and 55° obliquely to the ipsilateral side for visualization of the neural foramen with maximum transverse width. The appropriate entry site was posteriorly in the foramen at the division between the caudal and middle thirds. This was marked on the skin. The skin and subcutaneous tissue were anesthetized with 1% lidocaine. The Tuohy needle was advanced to the superior articular process via the tunnel vision technique. Then the needle was advanced into the neural foramen, touching its posterior border at the halfway point between the articular pillars’ medial and lateral borders in AP fluoroscopy. The needle should not be advanced beyond this point. The needle position was confirmed by real-time fluoroscopy with 0.5–1.0 mL of a contrast media (Fig. 1C and D). Intravascular injection was confirmed by diffusion of the contrast media through the vascular channel. Then, we injected 2 mL of a solution mixed with 0.5% lidocaine and 2.5 mg dexamethasone. The subjects were observed for 30 minutes post-procedure.

During the follow-up period, if the patient satisfaction score was equal to or above “good,” but the improvement of NRS score was < 50%, the injections were repeated at least two weeks apart. No additional injections were administered if the NRS score was 3 or less. If there was no pain relief or the pain worsened, or if the patient satisfaction rating was equal to or below “fair,” other treatments such as increasing the dose of analgesics, invasive procedures, or surgery were considered. “Excellent” indicated that the patient was ‘satisfied with the outcome, and the treatment was as effective as expected’; “good” indicated ‘not as effective as expected but willing to take the treatment again when the pain recurs’; and “fair” indicated ‘some effect but not enough to choose the treatment again when the pain recurs’; and “bad” indicated ‘same or worse effect as the previous treatment.’16

A standardized chart abstraction form was used to extract data on demographics, treatments, pain severity, and functional assessments. Follow-up interviews were conducted during hospital visits one, three, and six months after the initial injection by nursing personnel not involved in the injection procedure.

The primary outcomes included NRS scores for pain and functional disability. The degree of functional disability was measured using the Korean version of the Neck Disability Index (NDI; range, 0–50), which is the most widely used questionnaire assessing cervical spine abnormalities.1718 It consists of 10 questions, including seven questions related to functional activity, two questions about symptoms, and one question about concentration. The final score is obtained by summing the scores of all the questions. A higher NDI score indicates increased functional disability associated with cervical abnormality.17 Additionally, a subgroup analysis was performed to divide patients into moderate (NRS scores 4–6) and severe (NRS scores > 7) pain groups based on baseline pain scores.19

The secondary outcomes included the proportion of responders and the Medication Quantification Scale (MQS) score. A “responder” was defined as a patient experiencing a ≥ 50% improvement in their NRS score and a ≥ 40% or more improvement in their NDI score.1720 Patients who did not show this degree of improvement and those who required more analgesics, other invasive procedures, or surgery during the follow-up period were categorized as non-responders. We used the MQS to quantify medication use.21 The scores were calculated for each pain-related medication based on weights assigned by pharmacologic class and dosage level and were summed to yield the total MQS score.

The variables correlating with effectiveness one month after the initial injection were also evaluated. Independent variables, including the injection method, age, sex, number of injections, FS severity, MQS, pain duration, and the presence of CDH were documented in the medical charts. Age was classified into five groups: < 39, 40–49, 50–59, 60–69, and > 70 years. FS severity was classified into two grades according to routinely obtained cervical axial MRI findings based on Kim et al.22 (Fig. 2). Grade 1 included patients with non-severe FS, in which the narrowest width of the neural foramen was ≥ 50% of the width of the extraforaminal nerve root at the anterior margin of the superior articular process. Grade 2 included patients with severe FS, in which the narrowest width of the neural foramen was < 50% of the extraforaminal nerve root. MQSs were divided into two groups (≤ 3 and > 3).23

Finally, we reviewed the charts for immediate side effects, such as vasovagal reactions, facial flushing, or brief, severe neck pain within minutes of injection, and any complications that occurred after the procedure (e.g., motor weakness, sensory change, infection, or hematoma) at the one-month follow-up visit.

The Shapiro-Wilk test was used to check the normal distribution of continuous variables. Continuous variables are represented as mean ± SD or medians (interquartile ranges), and categorical variables are displayed as numbers (percent). The Student’s t-test and Pearson’s χ² test were used to compare the characteristics of the FJ and TFE groups regarding age, sex, number of injections, side of symptoms, target level, FS severity, MQS, pain duration, and the presence of CDH. At each time point (before the first injection and one, three, and six months after the initial injection), the NRS and NDI scores and MQS were compared by repeated-measures analysis of variance (ANOVA) with Bonferroni’s correction for post hoc comparisons. Univariate and multivariate logistic regression analyses with Pearson’s χ² test were performed to assess whether the injection method, age, sex, number of injections, FS severity, MQS, pain duration, and the presence of CDH were independent predictors of treatment success. All data were statistically analyzed using Statistical Package for the Social Sciences (SPSS) version 23.0 (IBM Corp, Armonk, NY, USA). P < 0.05 was considered statistically significant.

The Institutional Review Board centers of Ewha Womans University Seoul Hospital (SEUMC 2021-11-034) and Ewha Womans University Mokdong Hospital (EUMC 2021-12-035) approved the study. Since this study did not involve direct contact with the study population, the approval included a waiver of informed consent.