Clinical Question

What are the benefits and harms of using systemic corticosteroids for the management of radicular and nonradicular low back pain or symptomatic spinal stenosis in adults?

Evidence-Based Answer

In patients with radicular low back pain, systemic corticosteroids increase the likelihood of improvement in function at short-term follow-up (absolute improvement = 19% better; 95% CI, 8% to 30% better). In adults with nonradicular low back pain, the use of systemic corticosteroids does not lead to a discernible effect on pain or function. Systemic corticosteroids are not beneficial in treating pain or function in those with spinal stenosis. A short course of systemic corticosteroids does not appear to cause harm.¹ (Strength of Recommendation: B, inconsistent or limited-quality patient-oriented evidence.)

Practice Pointers

Low back pain can be categorized by etiology as radicular (lumbar disc or nerve root abnormalities), nonradicular mechanical (a combination of muscle, ligament, tendon, or bony abnormalities not resulting from spinal stenosis or disc or nerve root abnormalities), symptomatic due to spinal stenosis (narrowing of the spinal canal from bony and/or soft tissue structures), or back pain secondary to rheumatologic, inflammatory, metabolic, or malignant conditions. Management of low back pain may be specific to etiology and guided by duration of symptoms: acute pain (less than four weeks), subacute pain (four to 12 weeks), or chronic pain (more than 12 weeks). The authors sought to determine the benefits and harms of systemic corticosteroids two to 12 weeks after administration for low back pain that was radicular, nonradicular mechanical, or symptomatic due to spinal stenosis.

The Cochrane review included 13 randomized controlled trials with a total of 1,047 participants, and sample sizes ranged from 29 to 269.¹ Nine trials evaluated participants with radicular low back pain. Of the nine trials, three evaluated participants with acute symptoms, two evaluated those with mixed acute and nonacute symptoms, one evaluated those with nonacute symptoms, and in three trials the duration of symptoms was unclear. Two trials studied participants with acute nonradicular low back pain. Two trials studied participants with spinal stenosis: one trial included patients with chronic symptoms whereas the other did not report the duration of symptoms. Radiologic confirmation of conditions was required in only two of the nine trials for radicular low back pain; radiologic confirmation was also required in both of the trials for spinal stenosis. Patient demographics were not consistently provided; however, in studies that reported this information, the median age of participants was 40 years in the radicular and nonradicular low back pain trials and 58 years in the spinal stenosis trials. Corticosteroids were given as a one-time treatment or over the course of several days and were administered orally, intravenously, or intramuscularly; studies that evaluated epidural injections were excluded. Corticosteroid type, dosage, and length of treatment varied between studies. Total doses of prednisone equivalents ranged from 50 to 1,050 mg. Multiple scales were used to determine treatment effect.

There was no evidence that systemic corticosteroids improved pain or function in the short-or long-term in patients with nonradicular low back pain or spinal stenosis. Adverse effects were inconsistently reported but, in studies that included this information, no serious adverse effects were noted. Outcomes and adverse effects were not made more or less likely by the type of corticosteroid, route of administration, or dosing. Administration of a single dose vs. multi-day course of corticosteroids also did not make a difference in outcomes or adverse effects.

Updated guidelines from the U.S. Department of Veterans Affairs and U.S. Department of Defense (VA/DoD) and the North American Spine Society recommend against systemic corticosteroid administration for the treatment of low back pain because of poor evidence and possible adverse effects.^2,3 The VA/DoD guidelines apply to both radicular and nonradicular low back pain, whereas the North American Spine Society guidelines apply to low back pain without a neurologic defect or with radicular pain that does not travel beyond the knee. Family physicians may consider prescribing corticosteroids for improving function in patients with radicular low back pain but should discuss the relatively small benefit.