Response to: Reply to Sayeed et al

To the Editor: On behalf of the Spine Intervention Society (SIS), we would like to thank you for the opportunity to address the response from Dr Andersson et al1 to our previous letter to the editor2 addressing concerns with the American College of Occupational and Environmental Medicine’s (ACOEM) Guidelines: Invasive Treatments for Low Back Disorders.3 While the authors indicate that the ACOEM guidelines’ methodology “follows quality standards and incorporates evidence-based medicine principles of GRADE, Institute of Medicine, AGREE, and AMSTAR,” they fail to acknowledge a critical limitation of their methodology—randomized controlled trials do not tell the whole story and often do not necessarily represent the best available evidence. We agree that the best available evidence is often found in randomized controlled trials (RCTs); however, this is not always the case. Poor quality RCTs are found throughout the medical literature. And high quality, prospective, observational trials with greater methodological rigor than poor quality RCTs are also prolific. When assessing the evidence regarding any intervention, the best available evidence must be considered. Determining what constitutes the best available evidence involves consideration and weighing of the body of evidence with respect not only to study design, but to study quality and applicability—a hallmark of the evidence assessment methodologies that ostensibly serve as the basis for the ACOEM methodology. When addressing procedures where poor quality RCTs conflict with high quality observational studies, the authors state that, “Under these circumstances of somewhat conflicting evidence, one must accord the highest priority to the highest quality studies and study designs.” We principally agree with the statement but feel that the authors have accorded undue weight to study design, allowing poor quality, flawed RCTs disproportionate influence in formulating their recommendations. Study methodology is meaningless unless the procedures being assessed are performed on appropriately selected patients with appropriate indications using accurate and current technique, and that the data on outcomes are stratified by the specific condition and diagnosis being treated. The RCTs cited by the authors to support statements made against interventional procedures fail to present data stratified by diagnosis, different technical approaches, and/or to account for inadequate technique.4–9 An RCT with sound randomization, excellent blinding, and no losses to follow-up is of no value if the data are not stratified by confirmed diagnosis/ condition with appropriate technical approach or if the therapeutic procedure was not performed accurately. The evidence cited in our letter to the editor as high quality or “benchmark” studies on epidural steroid injections and facet procedures were those with rigorously applied selection criteria and meticulous technique.10–12 The Ghahreman ESI trial10 applied such technical rigor in epidural steroid injection procedural performance, as the procedures were performed in accordance with the very specific Practice Guidelines for Spinal Diagnostic and Treatment Procedures of the Spine Intervention Society.13 This study alone, of all the RCTs addressing epidural steroid injections, reasonably reflects current clinical practice. It is supported by a large body of prospectively acquired data in outcomes studies ignored by this guideline that also utilized appropriate patient selection criteria, meticulous technique, and reported categorical data to present the percentage of patients that obtain varying degrees of relief and/or functional improvement.14,15 While these other outcome studies do not shed light on efficacy, they certainly speak to the procedure’s effectiveness in the real world. The two benchmark studies of lumbar medial branch radiofrequency neurotomy used 80% to 100% relief thresholds following dual comparative local anesthetic blocks and a parallel treatment technique11,12—rigorous selection criteria to provide insight into the effectiveness of the procedure in patients confirmed to have facet-mediated pain. An impressive 55% to 60% of patients experienced at least 80% pain relief—a result that is surely meaningful to patients and should be an available option for patients who have already failed more conservative treatments and who wish to avoid the alternative treatments available to them (eg, opioids, surgery, life with unmanageable chronic pain). These findings are also supported by a large body of evidence ignored by this guideline, which relied on appropriate patient selection criteria, meticulous procedural technique, and reported categorical data to present the percentage of patients that obtain varying degrees of pain relief and/or functional improvement.16 We are pleased to hear that the authors are monitoring the quickly evolving evidence supporting high frequency 10 kHz spinal cord simulation (HF10 SCS). However, waiting for additional studies comparing HF10 SCS to a rehabilitation program or sham procedure is a bar set unrealistically high. By the time a patient is considered a candidate for implantation of a spinal cord stimulator, they have already failed to achieve adequate pain relief or functional improvement with rehabilitation, and most would not likely volunteer to enroll in a sham-controlled study. This raises an important challenge in pain research. It is difficult to recruit pain patients to participate in RCTs—especially explanatory RCTs. These patients do not want to risk being randomized to the placebo or sham arm. For this reason, it is far more practical to enroll patients in pragmatic RCTs, where patients are randomized to one or more treatments that are expected to be potentially effective. The evidence generated by both pragmatic RCTs and well designed, prospective, observational studies provides important real-world information on effectiveness, and can most certainly serve as the basis for treatment recommendations. On a related note, conservative management with non-opioid analgesics and rehabilitative treatment is recommended by most physicians and guidelines as the first-line treatment for back pain prior to any more invasive interventional spine procedures. As it pertains to the evidence, most studies assessing effectiveness of spine interventions require that patients fail to obtain adequate pain relief and functional improvement from conservative management prior to enrollment. We fully support a trial of conservative treatment for patients who can tolerate and respond well to them in the form of pain relief and functional improvement. Many patients will respond to conservative management negating the need for more invasive therapies. However, there will be patients who do not improve, their clinical progress plateaus, further physical/occupational therapy sessions will not be approved or are limited by their payors, or they cannot tolerate these treatments without aggravation of their existing conditions. Like any therapy, these patients should not be expected to continue prolonged ineffective treatments when complementary interventional procedures can provide effective pain relief and functional improvement in the form of expedited return to function and work, improved quality of life, decreased unnecessary healthcare resource utilization, and potential avoidance of surgery, reliance on opioids, or a lifetime of unmanageable chronic pain. Finally, we suggest that future guideline development panels include an adequate number of members with clinical expertise in the procedures being evaluated to provide guidance regarding technical aspects that impact the quality of the evidence assessment. Most of the authors of the current versions of the practice guidelines do not perform interventional spine procedures and would not be expected to understand the nuances of patient selection, indications, and specific technical variables of these procedures that necessarily impact outcomes. Without adequate practicing interventional spine physician representation to provide interventional expertise and perspective, there is inadvertent opportunity for introduction of bias. Conversely, the inclusion of several authors with significant conflicts of interest as employees of the health care industry, responsible for scrutinizing medical utilization and costs, may have also inadvertently introduced bias against supporting such procedures. Again, we hope that our comments further clarify for the ACOEM membership the utility of these procedures for our workers’ compensation patients. We appreciate the time and work of the panel members and reviewers that created this guideline, and hope that ACOEM will reconsider their weighting of study design, place additional emphasis on study quality, and ensure appropriate and adequate representation on future panels of experts with relevant clinical expertise on the procedures being evaluated. We encourage ACOEM to revisit the evidence and their recommendations to ensure patients have access to safe, high quality, evidence-based evaluation and treatment options for effective patient care when standard conservative care does not adequately alleviate symptoms and as an alternative to surgery, potential opioid use, or prolonged suffering. Yusef Sayeed, MD, MPH, MEng, CPH Department of Physical Medicine and Rehabilitation Department of Family Medicine Uniformed Services University of the Health Sciences Bethesda, MDRichard G. Chang, MD, MPH Department of Rehabilitation and Human Performance Mount Sinai Medical Center New York, NYAkhil Chhatre, MD Johns Hopkins University School of Medicine Department of Physical Medicine and Rehabilitation and Neurosurgery Baltimore, MDJayesh Vallabh, MD Department of Physical Medicine and Rehabilitation The Ohio State University College of Medicine Columbus, OHAmol Soin, MD Wright State University Dayton, OHWilliam D. Mauck, MD Division of Pain Medicine Department of Anesthesiology and Perioperative Medicine Mayo Clinic Rochester, MNBelinda S. Duszynski, BS Spine Intervention Society Hinsdale, ILD. Scott Kreiner, MD Health Policy Division of the Spine Intervention Society Barrow Brain and Spine Phoenix, AZ