Does Magnetic Resonance Imaging (MRI) Improve Interventional Outcomes for Lumbosacral Radiculopathy?

Overview

Epidural steroid injections (ESI) are the most frequently performed procedures in pain clinics. When performing ESI, there is no consensus about how to best select candidates for this intervention, and which level(s) to inject. Some experts advocate basing the injection level on MRI findings, whereas others recommend clinical symptoms. In order to determine whether MRI is necessary before referring patients with chronic LBP radiating into the leg(s) for pain management interventions, we will perform a randomized comparative study involving 131 patients with chronic low back and leg pain who are clinically candidates for epidural steroid injections into two groups. Group I will receive two ESI based solely on historical and physical exam (PE) findings. Group II will receive treatment only after MRI is reviewed. The investigators' hypothesis is that MRI will not have a significant effect on treatment outcome, and will have minimal impact on patient care.

One hundred and thirty-one patients referred to the Blaustein Pain Treatment Center with back and leg pain will be randomized to receive one of two treatments. Sixty-three patients will be allocated to group I and up to 68 to group II (see below and statistical analysis). All patients will be candidates for ESI based on history and physical exam. All 63 group I patients will receive two ESI based solely on history and PE (i.e. the treating physician will be blinded to MRI results). Group II will receive treatment based on MRI, history and PE. This probably but may not include ESI (i.e. it is conceivable that a normal MRI might result in conventional therapy instead of ESI, though chemical irritation of nerve roots often occurs without disc protrusion. We estimate this to occur in no more than 10% of cases (probably less); hence, we are planning to randomize up to 68 patients to group II. If an ESI is done, the patient will receive two injections, similar to patients in group I. Randomization will be done by a research assistant via a computerized randomization system in groups of 13. In order to ensure no serious pathology is missed in group I subjects, a separate doctor with no knowledge of treatment allocation will review the films. If there is serious pathology such as evidence of spinal metastases or infection, the patient will be excluded from the study. There will be no "control" group in this study. Both evaluating physicians and patients will be blinded until they exit the study.