Jane Shipley

Incidence and avoidance of neurologic complications with paddle type spinal cord stimulation leads.

Introduction:  While reference is frequently made to the risk of spinal cord or nerve root injury with the surgical implantation of paddle type spinal cord stimulation (SCS) electrodes, data are lacking on the frequency, causes, and prevention of these complications.

Spinal cord stimulation versus re-operation in patients with failed back surgery syndrome: an international multicenter randomized controlled trial (EVIDENCE study).

Objective:  This paper presents the protocol of the EVIDENCE study, a multicenter multinational randomized controlled trial to assess the effectiveness and cost‐effectiveness of spinal cord stimulation (SCS) with rechargeable pulse generator versus re‐operation through 36‐month follow‐up in patients with failed back surgery syndrome.

Prevention of Percutaneous Spinal Cord Stimulation Electrode Migration: A 15‐Year Experience

Percutaneous spinal cord stimulation electrodes have a propensity to migrate longitudinally, which is a costly complication that often compromises therapeutic effect. After implementing simple changes to our percutaneous electrode anchoring technique, we no longer encounter this migration. The current retrospective study updates previously reported results.

A Review of Economic Factors Related to the Delivery of Health Care for Chronic Low Back Pain

We describe tools used to evaluate the economic impact of health care interventions, discuss the economic burden of chronic low back pain, and review evidence on the cost‐effectiveness of treating failed back surgery syndrome with spinal cord stimulation, intrathecal drug delivery, acupuncture, epidural injections, disc prosthesis, lumbar fusion, and noninvasive therapies. We also mention the lack of cost studies for emerging therapies, such as vibrotherapy and peripheral nerve field stimulation. Topics include types of cost studies; the economic perspectives taken by such studies; direct and indirect costs; measures of success; definitions of cost‐effectiveness, incremental cost‐effectiveness, incremental cost‐utility ratios, and quality‐adjusted life years; the concept of maximum willingness to pay; and the use of cost‐effectiveness models.

Questions about Turner et al. Spinal cord stimulation for failed back surgery syndrome: outcomes in a worker's compensation setting.

We question the design and implementation of this prospective study of spinal cord stimulation (SCS), pain clinic (PC) treatment, and usual care (UC) in Washington State Department of Labor and Industry (DLI) worker’s compensation (WC) claimants with failed back surgery syndrome. We also question the way results were analyzed and reported. For a similar effort and cost, this study could have been a well-designed randomized controlled trial (RCT). This would have answered the lead author’s previous call for RCTs involving SCS [10]. The fact that the DLI covered SCS only for study participants would have facilitated recruitment, and an RCT would have improved the equity of access to SCS among subjects. The authors’ assertion that RCTs are not always generalizable to specific populations is a compelling reason for conducting an RCT in this WC population. The choice of study design requires explanation. The three non-randomized comparison groups were sourced and filtered differently, which introduced selection bias. The study sponsors approved SCS for only 61 (54%) of the 112 physiciannominated patients, whereas potential PC and UC subjects were ‘‘identified from administrative data.” Percentages deemed ineligible were notably imbalanced (48% SCS, 28% PC, and 23% UC). Furthermore, the presence of a chronic co-morbid pain condition that could confound pain relief results was not a criterion for exclusion. The impact of these subject selection decisions, especially the rationale for allowing the study sponsor to handpick the SCS group, requires consideration. The composite primary outcome measure (at least 50% leg pain relief, less than daily opioid use, and improvement in function) is unreferenced and, to our knowledge, unprecedented. Elements of the measure might cancel each other out, for example, increased function might increase pain, and the benefit of opioids for neuropathic pain is controversial. This outcome measure requires justification. Follow-up was linked to the time of study entry (or SCS referral for ‘‘crossovers”) not to treatment start dates; thus, some SCS outcomes were assessed before SCS was administered (e.g., three patients had no screening trial at 6 months). These data require adjustment. The authors obscure their unacknowledged presentation of worst-case scenario data in Table 2 by reporting percentages without raw numbers. Figure 1 does not agree with numbers at 24 months in Table 4 or with numbers at 6 months in Table 6. An interim report of this study, issued on 4/30/07 [4], notes that 31 PC patients visited the pain clinic more than several times per week. This is 9 (17.6%) more than the 22 reported in Pain. Logically, any change in the number treated between the time of an interim report and a final publication should be an increase, not a decrease. These reporting discrepancies require correction. The investigators apparently did not require the clinicians to follow the accepted SCS screening and treatment protocols [8] that help identify potential SCS patients and protect patient safety. This lack might have contributed to the unusually low SCS trial success rate, especially in a WC population [6], the unusually high SCS complication rate [3], and the occurrence of a rare life-threatening complication in this small sample. The authors’ claim that they evaluated SCS ‘‘in actual practice” rings hollow in the absence of prior ‘‘actual practice” in DLI claimants. The level of experience of the unidentified implanting physicians [1,7] and how they were chosen require description. The authors write that ‘‘industry-sponsored studies” introduce sponsorship bias that ‘‘yield[s] more favorable results.” Sponsorship bias, however, can also produce unfavorable results when a payer/sponsor does not want to cover a therapy. Such ‘‘policybased evidence making” [2,5], is contrary to the premise behind a recent editorial in Pain advocating publication of ‘‘negative” trials [9]. Any steps the authors might have taken to counteract the bias of conducting a study for a payer that has consistently refused to cover SCS in actual practice require discussion. We would be the first to agree with the authors’ observation: ‘‘SCS outcomes may vary according to patient selection criteria, physician technical expertise, and SCS implant techniques and hardware.” This report, however, details only one of these variables – SCS patient selection criteria, which we believe were insufficient. Without information on additional, crucial variables of SCS treatment, how can policy makers judge the merits of the reported results and how can researchers test the validity of these results through study replication?

The Cost-effectiveness of Spinal Cord Stimulation

Publisher Summary For chronic pain syndromes, the least expensive therapy is the one that offers sufficient clinical benefit to reduce the patients consumption of health care resources by a sufficient degree for a sufficient amount of time to recapture the cost of the therapy. An additional bonus accrues if the pain therapy provides more than symptomatic relief and improves the underlying condition that is causing the pain. Like most medical devices, SCS incurs high upfront costs and thus it must substantially improve the health of patients and/or produce later savings to be cost-effective. This chapter shows how SCS can improve a patients state of health, which has a direct impact on the patients quality of life, and lead to savings associated with a reduced consumption of healthcare resources. Among the benefits that have been documented in patients with SCS therapies are improved quality of life/ability to engage in the activities of daily living, reduction in the symptoms of depression, improved neurologic function, and ability to return to work. The types of cost studies that have been conducted vary from simple cost descriptions to full economic evaluations, although no cost evaluation has been performed from a societal perspective. Which costs are identified, how they are measured and valued, and how the data are collected are all important factors of cost studies. SCS also poses special challenges for healthcare economists who are developing models and analytical techniques for conducting economic evaluations. Improving equipment can also result in cost savings if the improvement increases the number of successful patient outcomes. The timing of SCS treatment is also of the utmost importance in patients with critical lower-limb ischemia, since SCS can promote healing only if trophic ischemic lesions have not progressed to 3 cm 2 .

Clinical Study Designs for Neuromodulation

Abstract Neuromodulation has inherent advantages over many other invasive treatments, in that it is inherently nondestructive and reversible, and clinical evidence attests to its favorable risk/benefit ratio. Spinal cord stimulation (SCS), the most widespread application of neuromodulation, became established before the era of evidence-based medicine, which privileges the results of blinded, randomized controlled trials (RCTs) over other forms of clinical evidence. SCS elicits paresthesia when delivered at conventional frequencies, precluding blinded trials and thus “high-quality evidence” by contemporary standards. Contemporary clinical study design methods are presented here as they apply to neuromodulation in general and SCS in particular, so as to demonstrate advantages in practice. New SCS parameters that reduce or eliminate paresthesia offer new opportunities for blinded clinical trials.

Generating evidence on spinal cord stimulation for failed back surgery syndrome: not yet fully charged. Clin j pain. 2008;24:757-758.

To the Editor: In his editorial accompanying a systematic review, Chou claims the reviewers, ‘‘conclude that spinal cord stimulation is clinically effective for failed back surgery syndrome... (and) ... is cost-effective.’’ The reviewers, however, simply report that the studies that met their inclusion criteria came to those conclusions. By conflating the study and review conclusions, Chou tacitly acknowledges the general agreement that the results of a study meeting the selection criteria of a properly conducted systematic review can reasonably be considered valid. Thus, to discredit the conclusions of the studies that meet review criteria, Chou must either discredit the review or turn away from the review and look to his opinions, which are, of course, subject to error and bias—the very problems the systematic review process is designed to reduce. Chou does not find fault with the method used by Bala et al to assess the clinical studies in their review—a method designed for the purpose of ‘‘generating sound conclusions.’’ Thus, when Chou then finds that ‘‘some of their (the reviewers’) conclusions seem to overstate the case for spinal cord stimulation,’’ he is forced to try to justify this statement by making his own conclusions about the studies. In this attempt, he makes significant factual errors. In his second paragraph, Chou states that, ‘‘most (spinal cord stimulation) complications do not lead to serious morbidity or mortality.’’ The fact is that spinal cord stimulation rarely leads to serious morbidity and has never to our knowledge caused mortality. Chou claims that Bala et al noted that ‘‘both randomized trials (2.3) enrolled relatively small samples... and had some methodologic flaws.’’ Bala et al duly report the sample sizes in the randomized controlled trials (RCTs), but the reviewers do not call them ‘‘relatively small’’ because they recognize that the sample sizes (50 and 100 patients) were appropriately powered. Furthermore, what Chou considers ‘‘methodologic flaws,’’ the reviewers’ term ‘‘limitations.’’ Chou criticizes the Prospective Randomised Controlled Multicentre Trial of the Effectiveness of Spinal Cord Stimulation study for using conventional medical management, which the patients have presumably failed already, as a comparator treatment. Bala et al, however, understand that conventional treatment is a valid comparator in an RCT (9% of those randomized to the Prospective Randomised Controlled Multicentre Trial of the Effectiveness of Spinal Cord Stimulation comparator arm achieved the primary outcome at 6 months). Chou mistakenly claims that we failed to perform an intention-to-treat analysis in our cost study, yet he quotes details from precisely that analysis. He presumes that our costs increase with longer follow-up, yet the procedures occurred within 6 months of each other. Chou also criticizes our clinical study for ‘‘a lack of sensitivity analyses...,’’ yet sensitivity analysis is properly reserved for modeling studies (we appropriately reported prognostic factor analysis of the RCT data set in our clinical paper). Chou criticizes the preliminary Taylor and Taylor modeling study for using ‘‘questionable parameters’’ (which were the best data available at the time—and which led the Taylors to call for a head-to-head trial) and mistakenly claims that the Taylors failed ‘‘to perform sensitivity analyses on costs, utilities, or rates of pain relief,’’ when in fact the data Chou claims is lacking can be found in their tables 3 and 5. Remarkably, despite these problems with his second guessing of the systematic review, Chou concludes that ‘‘...spinal cord stimulation seems to be a reasonable option for pain relief in patients with persistent or recurrent radiculopathy after back surgery.’’ He modifies his endorsement by stating ‘‘It is quite conceivable that future studies might lead to different conclusions regarding clinical effectiveness...’’ It is, of course, equally conceivable that future studies will confirm our conclusions. Chou ends by calling for ‘‘studies comparing spinal cord stimulation to interdisciplinary rehabilitation or other promising alternatives.’’ Because patients must fail all noninvasive therapies before receiving spinal cord stimulation, these other therapies must necessarily have failed. Presumably then, were Chou to review the studies he is calling for, he would feel compelled to make the same criticism that he employed against the use of conventional medical management as a comparator in the Prospective Randomised Controlled Multicentre Trial of the Effectiveness of Spinal Cord Stimulation study. Editorialists would do well to follow the fact checking procedures employed by researchers who conduct systematic reviews.