To the Editor:
We were excited to see the multicenter, randomized, comparative-effectiveness study of cervical epidural steroid injections comparing conservative treatment or combination treatment for cervical radicular pain by Cohen et al.1 published in Anesthesiology. However, we are concerned regarding the potential flaws with the study’s concept and design as well as the techniques used to examine the outcomes.
Pertinent information is missing in the introduction. The authors state that for cervical radiculopathy, two small controlled studies with fewer than 50 patients had evaluated epidural steroid injections2,3 with one demonstrating benefit.2 It appears that the authors have omitted a double-blind, randomized, active-controlled trial published in 2012 with a 2-yr follow-up that included 120 patients.4 The results of this trial are important because at the end of 2 yr, significant pain relief and functional status improvement of at least 50% were observed in 72% of the patients receiving local anesthetic only and 68% of those who received local anesthetic and steroids. Furthermore, the results were even more robust if patients were selected based on their response to the initial two procedures. This trial was practical in nature, including only patients with chronic pain who had failed conservative management, and the injection was repeated only if the pain had recurred. During a 2-yr period, patients received approximately five procedures.
As the authors have stated, in patients with radiculopathy, there is minimal benefit from physical therapy and pharmacotherapy as stand-alone treatments.
Therefore, addition of conservative management as a comparative treatment factor does not contribute to better understanding the role of epidural steroid injections. Furthermore, the authors have used epidural steroids without local anesthetic. As one would expect, there was no significant difference in these groups, and the combination therapy, which was reported as successful, demonstrated only minimal improvement.
The study design was somewhat surprising in that Cohen (Bicket et al.)5 had previously demonstrated that nonsteroidal solutions may have been superior to steroid solutions. Thus, it might have been more appropriate to use a local anesthetic with a steroid injection or design a group with local anesthetic alone similar to other studies. As Manchikanti et al.4,6–8 have elucidated, there was no significant difference between local anesthetic alone and local anesthetic with steroids. Furthermore, another systematic review also demonstrated a lack of superiority of steroids over local anesthetic in all conditions except lumbar disc herniation; whereas short- to mid-term, there was significant improvement with caudal and lumbar interlaminar epidural injections in managing lumbar disc herniation.9
Trials studying interventional treatments of chronic pain need to match a treatment with the incident disease. Methodologic limitations can and will lead to an inability to demonstrate the effect of the intervention.10,11 The study by Cohen et al. suffers from a failure to include local anesthetic in the treatment group and a focus on conservative therapy.
The authors declare no competing interests.