Assessing the Agreement Between Radiologic and Clinical Measurements of Lumbar and Cervical Epidural Depths in Patients Undergoing Prone Interlaminar Epidural Steroid Injection.

BACKGROUND: Fluoroscopy-guided epidural steroid injection (ESI) commonly is performed to treat radicular pain yet can lead to adverse events if the needle is not advanced with precision. Accurate preoperative assessment of the distance from the skin to the epidural space holds the potential for reducing the risks of adverse effects from ESI. It was hypothesized that the distance from the skin to the epidural space as measured on preoperative magnetic resonance imaging (MRI) would agree with the distance traveled by a Tuohy needle to reach the epidural space during midline, interlaminar ESI. This study compared the final needle depth measurement at the point of loss of resistance (LOR) from cervical or lumbar ESI to the distance from the skin to the anterior and posterior borders of the epidural space on the associated cervical and lumbar preoperative MRI.
METHODS: This retrospective chart review analyzed the procedure notes, MRI, and demographic data of patients who received a prone, interlaminar ESI at an outpatient chronic pain clinic between June 1, 2013, and June 1, 2015. The following data were collected: body mass index (BMI), age, sex, intervertebral level of the ESI, and LOR depth. We then measured the distance from the skin surface to the anterior border of the ligamentum flavum (ligamentum flavum depth [LFD]) and dura (dura depth [DD]) on MRI. A total of 335 patients were categorized into the following patient subgroups: age ≥65 years, age <65 years, BMI ≥30 kg/m (obese), BMI <30 kg/m (nonobese), male, and female. Secondary analyses were then performed to compare the agreement between LOR depth and DD with that between LOR depth and LFD within each patient subgroup. Intraclass correlation coefficient (ICC) and Bland-Altman plot were used to assess the agreement between DD or LFD and LOR depth.
RESULTS: Data from 335 ESIs were analyzed, including 147 cervical ESIs and 188 lumbar ESIs. Estimated ICC values for the agreement between LOR depth and LFD for all lumbar and cervical measurements were 0.88 (95% confidence interval [CI], 0.85-0.91) and 0.72 (95% CI, 0.64-0.79), respectively. Estimated ICC values for the agreement between LOR depth and DD for all lumbar and cervical measurements were 0.86 (95% CI, 0.82-0.89) and 0.69 (95% CI, 0.60-0.77), respectively.
CONCLUSIONS: This study assessed the agreement between MRI-derived measurements of epidural depth and those determined clinically. MRI-derived measurements from the skin to the anterior border of the ligamentum flavum, which represents the most posterior aspect of the epidural space, revealed stronger agreement with LOR depths than did measurements to the dura or the most anterior aspect of the epidural space. These results require further analysis and refinement before supporting clinical application.