Recent studies document a 629% increase in Medicare expenditures for epidural steroid injections in use for chronic low back pain, despite the fact that these increases have not been accompanied by population-level improvements in patient outcomes or disability rates [Deyo, J Am Board Fam Med 2009]. And this makes even more acute the issue of safety and efficacy, which I summarize briefly here from a recent commissioned internal review I undertook:
It is now well-established that epidural steroid injections are associated with a wide spectrum of adverse events [Bellini, Anaesthesiol Intensive Ther 2013], including elevated temperature, euphoria, depression, mood swings, local fat atrophy, depigmentation of the skin, pain flare, with insomnia (39%), facial erythema (20%), nausea (20%) and rash and pruritus (8%) having been observed following betamethasone injection [Dreyfuss. Fifth Annual Meeting International Spinal Injection Society 1997].
Furthermore, more series adverse events include hyperglycemia specifically after an epidural corticosteroid injection which may potentially contribute to the development of insulin resistance [Millefert, Rev Rheum Engl Ed 1995; Younes, Joint Bone Spine J 2007] in healthy non-diabetics, and also cause an increase in the blood glucose levels even in diabetics [Even, Spine J 2012]. On the neurological front, complications that have been reported following epidural corticosteroid injections include neurotoxicity and neurologic injury [Cohen, Reg Anesth Pain Med 2013], which the FDA is especially concerned about. Still others include steroid-induced myopathy (characterized by progressive proximal muscle weakness), which has been reported even following a single epidural dose of triamcinolone [Boonen, Br J Rheumatol 1995], and there is also evidence of paraplegia following thoracic and lumbar transforaminal epidural steroid injections [Shah, Pain Pract 2014].
It should also be noted that intraspinal bleeding is observed, a potentially devastating complication that can result in paraplegia or quadriplegia, often with back pain and headache as main or only the presenting complaints, and both epidural and subdural hematomas have been reported following epidural steroids, especially but not exclusively in patients on anticoagulant and/or antiplatelet therapy [Cohen, Reg Anesth Pain Med 2013]. And one recent review describes the growing number of cases associated with spinal cord and posterior circulation ischemia as a complication of cervical epidural steroid injection [Popescu, J Neuroimaging 2013].
As to efficacy, systematic review and meta-analysis finds that we lack plausible evidence of benefit of epidural steroid injections in the management of axial neck pain, axial low back pain, and lumbar spinal stenosis (using either lumbar or cervical interlaminar injections) [Abdi, Pain Physician 2005]. In confirmation and extension of these findings, the Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology (AAN) concluded that epidural steroid injection for radicular lumbosacral pain does not impact average impairment of function, nor the need for surgery, nor provide long-term pain relief beyond 3 months, so that their routine use for these indications is not recommended (Level B, Class I-III evidence), and with insufficient evidence to make any recommendation for the use of epidural steroid injections to treat radicular cervical pain (Level U) [Armon, Neurology 2007].
In addition, the Special Interest Group of the Canadian Pain Society has produced consensus-based guidelines for the pharmacological management of neuropathic pain, finding that the evidence is not strong or robust, and often limited or conflicting [Mailis, Pain Res Manag 2012], with the quality of the cumulative evidence judged only fair-to-poor, with strength of recommendations being only Grade B or Grade C, and with many specific recommendation achieving only Grade D status, and more importantly, they concluded that there is no evidence, to support that multiple injections will generate long-term relief.
Furthermore, epidural steroid efficacy is not superior for pain relief compared to placebo in sciatica, with more side effects in inducing a substantial enhancement of adverse events [Roncoroni, Rheumatology (Oxford) 2011; Cohen, Ann Intern Med 2012], and for low back pain, reviews have concluded that "despite over a century of use, the effectiveness of epidural GC injections remains controversial" [Balagué, Swiss Med Wkly 2012].
Finally, a recent multicenter, blinded, randomized controlled trial found that neither caudal epidural steroid injections nor caudal epidural saline injections are effective for chronic lumbar radiculopathy and are not recommended as an adjunct to recovery in patients whose symptoms have extended beyond 12 weeks [Iversen, BMJ 2011].
Where We Stand
Collectively, a review of the best methodologically assessed and critically appraised data to date finds (1) limited and inconsistent evidence of short-term benefits, (2) no robust evidence of long-term-benefits, and (3) inappreciable evidence of superiority to conservative management interventions and/or watchful waiting, and (4) a broad spectrum of adverse events and potential complications, supporting that (5) their use in pain therapies remains, and should remain, unapproved by the FDA and other regulatory counterparts elsewhere, and be subject to new and explicitly FDA-mandated warnings, and (6) that their clear overuse and abuse requires new close scrutiny and investigation, until such time, should if appear, that more robust and mature long-term data arrive to suggest any material reconsideration of these conclusions.