Randomized trial of epidural injections for spinal stenosis published in the New England Journal of Medicine: further confusion without clarification.

Randomized controlled trials are considered the hallmark of evidence-based medicine. This conveys the idea that up-to-date evidence applied consistently in clinical practice, in combination with clinicians' individual expertise and patients own preference/expectations are enjoined to achieve the best possible outcome. Since its inception in 1990s, evidence-based medicine has evolved in conjunction with numerous changes in the healthcare environment. However, the benefits of evidence-based medicine have not materialized for spinal pain including surgical interventions. Consequently, the debate continues on the efficacy and medical necessity of multiple interventions provided in managing spinal pain. Friedly et al published a randomized controlled trial of epidural glucocorticoid injections for spinal stenosis in the July 2014 edition of the highly prestigious New England Journal of Medicine. This was accompanied by an editorial from Andersson. This manuscript provided significant sensationalism for the media and confusion for the spine community. This randomized trial of epidural glucocorticoid injections for spinal stenosis and accompanying editorial concluded that epidural injections of glucocorticoids plus lidocaine offered minimal or no short-term benefit as compared with epidural injections of lidocaine alone, with the editorial emphasizing proceeding directly to surgical intervention. In addition media statements by the authors also emphasized the idea that exercise or surgery might be better options for patients suffereing from narrowing of the spinal canal. The interventional pain management community believes that there are severe limitations to this study, manuscript, and accompanying editorial. The design, inclusion criteria, outcomes assessment, analysis of data and interpretation, and conclusions of this trial point to the fact that this highly sophisticated and much publicized randomized trial may not be appropriate and lead to misinformation. The design of the trial was inappropriate with failure to include existing randomized trials, with inclusion criteria that did not incorporate conservative management,or caudal epidural injections. Simultaneously, acute pain patients were included, multilevel stenosis and various other factors were not identified. The interventions included lumbar interlaminar and transforaminal epidural injections with highly variable volumes of medication being injected per patient. Outcomes assessment was not optimal with assessment of the patients at 3 and 6 weeks for a procedure which provides on average 3 weeks of relief and utilizing an instrument which is more appropriately utilized in acute and subacute low back pain. Analysis of the data was hampered by inadequate subgroup analysis leading to inappropriate interpretation. Based on the available data epidural local anesthetic with steroids was clearly superior at 3 weeks and potentially at 6 weeks. Further, both treatments were effective considering the baseline to 3 week and 6 week assessment, appropriate subgroup analysis seems to have yielded significant superiority for interlaminar epidural injections compared to transforaminal epidural injections with local anesthetic with or without steroids specifically with proportion of patients achieving greater than 50% improvement at 3 and 6 week levels. This critical assessment shows that this study suffers from a challenging design, was premised on the exclusion of available high-quality literature, and had inadequate duration of follow-up for an interventional technique with poor assessment criteria and reporting. Finally the analysis and interpretation of data has led to inaccurate and inappropriate conclusions which we do not believe is based on scientific evidence.