Cochrane Review of SCS for Low Back Pain: A Review by the American Society of Pain and Neuroscience

Spinal cord stimulation (SCS) is a form of neuromodulation that has existed since 1967, with the first implants performed by Norman Shealy, MD, in six patients with a variety of chronic pain conditions.¹ This therapy is supported by high-quality evidence for the treatment of low back pain (LBP), spanning decades of research.^2-7

We were surprised and disappointed to read a recent Cochrane Review⁸ published with such negative findings that are inconsistent and misrepresentative of the aforementioned positive evidence. Because of this dissonance, the review deserves further inspection and clarification to identify the issues of this analysis by which we are concerned.

The Cochrane Library prides itself in three principles: trusted evidence, informed decisions and better health for patients.⁹ Unfortunately, this review appears not to have attained the publisher’s stated goals. This SCS for LBP review eliminates accepted evidence, validates a predetermined and potentially biased result that may decrease access, and could potentially increase healthcare costs for patients in Australia and around the world.

The authors of this editorial have four major issues with this SCS review for LBP:

• biased initiation for this review;
• aggregation of all types of SCS and various waveforms as equal and similar;
• characterization of LBP as a diagnosis when it is actually more representative of a symptom; and
• potential ethical issues associated with attempting to justify restricting access to a long-standing therapy for chronic LBP.

The Cochrane Library has been considered as an unbiased source of systematic reviews of different types of therapies for a given medical condition. The Methodological Expectations of Cochrane Intervention Reviews (MECIR) are described in the library’s handbook, defining conduct.¹⁰ Some of the mandatory requirements include predefining objectives of the review such as participants, interventions, comparators and outcomes (PICOs), along with considering important adverse effects of the intervention, “and ensur(ing) they are addressed.”¹¹ Further, the organization’s guidance states: “It is important that adverse effects are addressed in order to avoid one sided summaries of the evidence,” and “sometimes data on adverse events are best obtained from non-randomized studies or qualitative studies.”¹¹ These requirements have served to enhance the outstanding reputation of the Cochrane Library for many years.

The architecture of Cochrane Reviews also requires mandatory predefined criteria for participants. These include eligibility criteria for participants in the studies, as well as predefined and unambiguous criteria for interventions and comparators. These criteria include specification of the comparator for an inactive control (placebo, no treatment, standard care or with an active control intervention, i.e., different variant of the same intervention), with clear explanations of any restrictions utilized.¹¹ Further, the library mandates the choice of eligible study designs, with appropriate justification to include or restrict randomized and nonrandomized studies, and potential for bias.¹¹

The review by Traeger et al⁸ regarding which we have questions was initiated as a response to safety concerns from a prior Australian review by Jones et al in 2022 that noted a concerning number of adverse events and the reported unfavorable explant-to-implant ratio of SCS.¹² However, the explant-to-implant ratio reported in the Jones et al review was considerably higher and inconsistent with failure rates noted previously.^13,14 If the group was attempting to establish efficacy, they failed to designate any specific measures (e.g., PICOs criteria), neglected to identify a justified appropriate comparator (choice of placebo vs. SCS vs. standard of care), and excluded available literature, both published and unpublished.

Contributing to the questionable premise for the Traeger et al review is the fallacy that any single therapy in isolation can cure a chronic multifactorial pain condition, such as LBP. LBP as a specific diagnosis represents a challenge, as it is more characteristic of a symptom. “Low back pain,” in and of itself, is no longer a diagnosis per the International Classification of Diseases, Eleventh Revision.¹⁵ LBP encompasses many diagnoses, and in the cascade of a degenerative spine, homogenizing LBP oversimplifies the complexity of the differential diagnosis and its response to neuromodulation.¹⁶ Current pain management maintains that SCS is not a panacea for chronic pain. Ideally, it is an aspect of sound multimodal care, including restorative therapies, medication management as necessary and behavioral therapies.¹⁷

SCS, although often referenced as a generic term, varies widely in its approaches.^18-22 High-frequency stimulation at 10,000 Hz, Burst DR, closed-loop stimulation—to name a few—represent innovations in stimulation paradigms. It is clear from published works that these innovations offer strategic advancements in clinical outcomes and longevity, while also behaving as electrophysiologically distinct.²³

These innovative approaches often outpace the ability to accommodate investigations architected as randomized controlled trials (RCTs). Despite this research challenge, analyses of RCTs are present in the literature, including those of patient cohorts with LBP,^4-7 with one being a prospective, randomized, double-blind multicenter trial.^2,3 These studies compare one neuromodulation treatment strategy with another and not with conventional medical management or placebo, and they should not be ignored in the analysis efficacy, as doing so would create an unbalanced and incomplete evidence analysis. Further, the MECIR mandates that a comparator could include standard of care. Of note, Traeger et al chose to accommodate unpublished works that appear to have confirmed their biases against SCS, while others were ignored^24,25 or unfairly assessed.

While modern pain management is ideally multidisciplinary, none of the authors of the Cochrane Review in question had any experience with neuromodulation. Traeger and his group are primarily physiotherapists and did not include any other colleagues within the pain management discipline or SCS specifically, not including the regional resources available of the Neuromodulation Society of Australia and New Zealand, the International Neuromodulation Society, or the International Society of Pain and Neuroscience, to name a few.^26-28

We are concerned that some of the authors of this review may have been at least unconsciously biased, evidenced by the authors’ prior works,^29-33 which focused on alleged nonintervention and overutilization. From an ethical perspective, the funding of this review by the Australian government, which has been focusing aggressively on reducing healthcare utilization, is particularly problematic, potentially representing a conflict of interest. For example, the Cochrane Review’s first author was recently “awarded a $1.12 million for NUDG-ED: Trial of behavioural nudging interventions to reduce unnecessary care for LBP.”³⁴ The review’s senior author, Christopher Maher, PhD, also received Australian government funding. When funding agencies’ agendas emphasize reduction of healthcare utilization, those who are funded are likely to be either consciously or unconsciously motivated to produce reviews that are “consistent” with the goals of funders.

In addition, in his disclosures,⁸ Traeger acknowledged that he has served as a paid consultant “on models of physiotherapy care to a health service provider.” If he is a consultant being paid to further physiotherapy as a discipline, there exists the possibility that his involvement in this Cochrane Review would potentially be biased to underestimate the efficacy of a treatment modality that is in direct “competition” with physiotherapy. Further, Ian Harris (another of the review’s authors) published a book in 2016, titled “Surgery, the Ultimate Placebo: A Surgeon Cuts Through the Evidence.”³² We are not sufficiently naive not to believe that in some cases, unnecessary surgeries are performed. However, to suggest that surgeries are merely “placebos” uncovers potentially problematic agenda, which brings some question to this author’s participation in the Cochrane Review in question, and may have been inappropriate due to the possibility of formidable bias.

In addition to the potential conflicts of interest of the funder and authors of the Traeger et al review, we are most concerned regarding the potential for third-party healthcare payors to weaponize the authors’ invalid conclusions. The quality of chronic pain management in National Health Service (NHS) nations is not at its pinnacle at this juncture, and is sadly even more inadequate in the United States, where the profitability of extraneous stakeholders such as the health insurance and hospital industries have continued to utilize limited financial resources that should be going toward improving patient care.³⁵

For-profit insurers, for example, have a lengthy history of weaponizing any “data” that they can find that support their decisions to deny access to available high-quality care, regardless of the actual evidence bases of these treatments. In the insurance universe, the review by Traeger et al may serve as an excuse to deny coverage of SCS for LBP, despite insurance corporations’ full awareness of the strength of the extant body of empirical literature that supports its use among well-selected patients. The NHS, despite being one of the world’s top ranked healthcare systems, has been struggling financially over recent years. Given these financial concerns, bias associated with the review’s funding should be examined closely.

As ethical pain care professionals and pain scientists, we certainly understand the need for cost containment in healthcare. However, we also opine that research that drives pain care policy ought to be transparent and free of potential conflicts of interest and methodologically robust in order to produce findings that will help relieve patients’ pain and improve their quality of life. The American Society of Pain and Neuroscience has the aforementioned concerns regarding this Cochrane Review, as we believe that cost containment and profitability are not the ends of pain management research—which should be the relief of pain and suffering of an extremely vulnerable population.

As a final note, we urge the Cochrane Collaboration to reconsider the publication of further reviews of therapies that are not methodologically robust and might raise ethical concerns. As pain scientists, we have held the Cochrane Collaboration in high esteem, trusting the conclusions of its authors beyond other publishers of systematic reviews. Hopefully, the organization will right its ship, and return to its mission of promoting and publishing the excellent quality science that has brought it such a positive reputation.

References

1. Shealy CN, et al. J Neurosurg. 1970;32(5):560-564.
2. Mekhail N, et al. Cet Neuro. 2020;19(2):123-134.
3. Mekhail N, et al. JAMA Neurol. 2022;79(3):251-260.
4. Kapural L, et al. Anesthesiology. 2015;123(4):851-860.
5. Kapural L, et al. J Neurosurg Spine. 2022;11:1-12.
6. Deer T, et al. Neuromodulation. 2018;21(1):56-66.
7. Fishman MA, et al. Pain Pract. 2020;20(7):761-768.
8. Traeger AC, et al. Cochrane Database Syst Rev. 2023;3(3):CD014789.
9. Cochrane Library. About Cochrane Reviews. Accessed May 4, 2023. https://www.cochranelibrary.com/about/ about-cochrane-reviews
10. Cochrane Community. MECIR Manual. Accessed May 4, 2023. https://community.cochrane.org/mecir-manual
11. Higgins JPT, eds, et al. Cochrane Handbook for Systematic Reviews of Interventions. Accessed May 8, 2023. https://training.cochrane.org/handbook/ current
12. Jones CM, et al. J Patient Safety. 2022;18(5):507-511.
13. Pope JE, et al. Neuromodulation. 2017;20(6):543-552.
14. Van Buyten JP, et al. Neuromodulation. 2017;20(7):642-649.
15. World Health Organization. ICD-11 coding tool. Accessed May 2, 2023. https://icd.who.int/ct11/ icd11_mms/ en/ release
16. Deer TR, et al. J Pain Res. 2022;15:1325-1354.
17. Department of Health and Human Services. May 2019. Accessed May 4, 2023. https://www.hhs.gov/ash/ advisory-committees/ pain/ reports/ index.html
18. Eldabe S, et al. Pain Pract. 2023;23(3):232-233.
19. Abd-Elsayed A, et al. Pain Pract. Published online March 2, 2023. doi:10.1111/papr.13218
20. Arle JE, et al. Neuromodulation. 2016;19(4):385-397.
21. Arle JE, et al. Neuromodulation. 2020;23(5):582-593.
22. Heijmans L, et al. Postgrad Med. 2020;132(suppl 3):17-21.
23. Falowski S. Neuromodulation. 2019;22(2):219-228.
24. Deer TR. DISTINCT study. Paper presented at: the Annual Conference of the North American Neuromodulation Society; January 2023; Las Vegas, NV.
25. Patel NP, et al. J Neurosurg Spine. 2022;38(2):249-257.
26. Neuromodulation Society of Australia and New Zealand (homepage). Accessed May 4, 2023. https://www.nsanz.org.au
27. International Neuromodulation Society (homepage). Accessed May 4, 2023. https://www.neuromodulation.com
28. American Society of Pain and Neuroscience. International Society of Pain and Neuroscience. Become an ISPN member. Accessed May 4, 2023. https://aspnpain.com/international-society-of-pain-and-neuroscience/
29. Zadro JR, et al. J Orthop Sports Phys Ther. 2020;50(3):113-115.
30. Traeger AC, et al. J Orthop Sports Phys Ther. 2021;51(1):1-4.
31. Traeger AC, et al. J Physiother. 2017;63(2):63-65.
32. Harris IA. Surgery, the Ultimate Placebo: A Surgeon Cuts Through the Evidence. New South Publishing; 2016.
33. Buchbinder R, Harris IA. Hippocrasy: How Doctors Are Betraying Their Oath. New South Publishing; 2021.
34. University of Sydney. Sydney researchers secure over 12 million clinical trial funding. May 24, 2022. Accessed May 4, 2023. https://www.sydney.edu.au/news-opinion/ news/ 2022/ 05/ 24/ sydney-researchers-secure-over-12-million-clinical-trial-fundin.html
35. Schatman ME. Pain Med. 2011;12(7):1060-1062.