Quality of Evidence Base

Because this Key Question does not concern the causal relationship between decompression therapy and treatment outcome, an assessment of study quality is not relevant.

Findings

We present the inclusion and (for three studies) exclusion criteria reported in the eight studies that address this Key Question in Table 6. The only inclusion/exclusion criterion consistently listed across all of the studies was that enrolled patients must have suffered from chronic low back pain related to radiographically confirmed disc degeneration or herniation. Common exclusion criteria described in the two studies reporting them were tumor, infection spinal instability and surgical implants. Other reported inclusion/exclusion criteria were unique to individual studies. Of particular relevance to the Medicare population is the fact that the two studies which included patients over the age of 65 specifically excluded patients with “severe osteoporosis.”(1,10)

Table 6

Inclusion and Exclusion Criteria in studies of Vertebral Axial Decompression Therapy.

Subsection Summary

The studies that addressed this question required that enrolled patients have radiographically confirmed disc degeneration or herniation or facet arthritis. Common exclusion criteria reported in the three studies reporting them were tumor, infection spinal instability and surgical implants.

Quality of Evidence Base

Because this Key Question does not concern the causal relationship between vertebral axial decompression or decompression therapy and treatment outcome, an assessment of study quality is not relevant.

Findings

The most frequently reported outcome measures evaluated by these studies of the efficacy or effectiveness of decompression therapy were pain relief (as pre-, post- or change in pain scores) or percentage improvement in pain score (six studies), functional outcomes (three studies) and physiological outcome measures (three studies). For the studies reporting physiological measures, only one also reported a patient-oriented outcome (percent improvement in pain).(9) The physiological measures were current perception threshold (for detecting sensory changes),(7) dermatomal somatosensory evoked potentials (for detecting sensory changes)(9) and intradiscal pressure (to explore a possible mechanism for the clinical effects of VAX-D therapy).(8) The functional outcome measures included self-report of ADLs (0 – 5 scale),(1) a disability rating specific to the individual’s most affected activities (0 – 4 scale),(3) and limitation of ambulation (0 – 3 scale).(5) The study which collected patient-reported ADL information did not report the results, but incorporated the data into an overall assessment (remission, partial remission or no response), which also incorporated pain relief and return to work.

Table 13 of Appendix D lists the outcomes in the decompression studies along with those commonly assessed in studies of other non-surgical modalities for chronic low back pain due to herniated disc or degenerative disc disease.(11,21,22,44–46) A number of outcomes have been evaluated and reported by studies of these other non-surgical modalities that have not been reported by studies of decompression therapy. These include absenteeism, return to work, overall health, analgesic consumption, low back pain-related disability rates, recovery time, gait analysis, and quality of life.

Subsection Summary

The outcomes in these studies of spinal decompression therapy also reported in studies of other non-surgical treatment options for low back pain are pain scores, pain relief and functional status. Data pertaining to a number of outcomes that are commonly reported by studies of other non-surgical treatment modalities (absenteeism, return to work, overall health, analgesic consumption, low back pain-related disability rates, recovery time, gait analysis, quality of life), are not yet available in the spinal decompression therapy literature.

Quality of Evidence Base

The results of our assessment of the quality of the studies that comprise the evidence base for Key Question 3 are presented in Table 7. Details of the quality assessment are presented in Table 16 of Appendix D.

Table 7

Results of Assessment of Study Quality.

The study by Ramos(1) was an open (unblinded) comparative trial in which no attempt was made to ensure the patient groups were comparable at baseline. ECRI’s evaluation of the study found it to be highly susceptible to bias. Consequently, we do not consider this study further for Key Question 3.

The study by Sherry et al.(3) was an open trial in which patients were randomized by sequential order, generally considered an inappropriate method of randomization. Our assessment of the quality of this study found it to be of low quality.

The study by Shealy and Borgmeyer(6) of the DRS® system was a blinded randomized trial. Despite this, our assessment of the quality of this study found it to be of low quality. A number of factors led to this categorization. First, the method of randomization was not described in the article describing the study. This precludes one from determining whether randomization was stochastic and whether concealment of allocation to treatment groups occurred. Second, although patients were blinded to their treatment assignments, the study did not report on the success of blinding or whether unblinded investigators were involved in ascertaining patients’ ratings of their response to treatment. Third, it appears that patients’ ratings depended on their recollection of their symptoms prior to entering the study. Lastly, the authors of the study have significant financial interests in the company that manufactures the DRS® system; Dr. Shealy is the inventor of the DRS®system(47) and director of the Shealy Institute, a pain management facility which utilizes DRS® in its treatments, and Ms. Borgmeyer is a Research Coordinator at the same institute(6)

Characteristics of Enrolled Patients

Important characteristics of the patients enrolled in the two studies that comprise the evidence base for Key Question 3 are summarized in Table 8. We present further information on the characteristics of the patients enrolled in the two included studies in Table 12 in Appendix D.

Table 8

Characteristics of Enrolled Patients (Key Question 3).

In both included studies, enrolled patients suffered from chronic low back pain that was unresponsive to conservative treatment. As indicated below, patients in the Sherry et al. study had (on average) been symptomatic much longer than those in the Shealy and Borgmeyer study. Disc problems were confirmed by imaging studies in both studies. Patients with facet arthritis in the study by Shealy and Borgmeyer underwent MRI to rule out other pathology.(6) The enrolled patients are therefore representative of the type of patient likely to be treated by decompression therapy in the clinic. Neither study included patients over 65 years of age.

Findings

The findings of the two low quality studies that form the evidence base for Key Question 3 are presented in Table 17 and Table 18 of Appendix D. Both studies found that the two types of decompression therapy were effective in reducing pain.

Sherry et al. compared a typical VAX-D decompression protocol in 22 patients to treatment with TENS in 22 patients. The TENS protocol used has been criticized as being suboptimal, and not a typical TENS protocol in that patients received TENS therapy for thirty minutes once daily, five days a week, rather than continuously or as needed for pain. The protocol described for each type of therapy was to provide 24 treatments (five days a week for four weeks, then once a week for four weeks). However, the VAX-D treated patients received a mean of 24.1 treatments, with a range of 18 to 36, while the TENS group received a mean of 18.0, range 10 to 24 treatments. Whether the overall duration of treatment was longer in the VAX-D group is not noted.(4)

Post-treatment VAS scores in the study by Sherry et al(3) indicate that the “evaluable” patients treated with decompression therapy had greater pain relief than patients in the control group. The difference was both statistically and clinically significant. Of note, three of four randomized patients who were not considered “evaluable” were treated with VAX-D. Two of these patients were noted after treatment to have had baseline VAS scores <2.0, and another withdrew because treatment was no longer required. One patient in the TENS group did not wish to continue treatment. In addition to reporting changes in VAS scores, Sherry et al. also reported the percentage of patients achieving at least 50% pain relief. No patients in the control group achieved this level of pain relief, as compared to 68% of patients in the decompression group who did achieve this level of pain relief. This difference is both statistically and clinically significant.

Sherry et al. also asked the participants to rate their disability on four activities most affected by their low back pain. Mean disability scores pre- and post-treatment were reported without measures of variance or tests of statistical significance. On a scale of 1 – 4, with 1 indicating complete disability and 4 representing no limitation of activity, the mean scores for the VAX-D group were 2.2 (pre-) and 2.9 (post-treatment); mean scores for the TENS group remained 2.2.(3)

Shealy and Borgmeyer(6) compared a typical DRS decompression protocol in addition to TENS to a standard traction therapy protocol in addition to TENS. They did not report changes in VAS scores, but did report the percentage of patients reporting “poor,” “good” or “excellent” improvement in symptoms. The article does not provide specific details on severity or nature of baseline symptoms or how patients’ assessments of improvement were ascertained. Patients treated with DRS decompression therapy were more likely to have excellent improvement in symptoms than patients treated with traction. However, when patients with excellent and good improvement after treatment were combined, the difference between treatment groups was not statistically significant.

Key Question 3 a. Do patients with chronic low back pain (due to herniated disc or degenerative disc disease) who are treated with decompression therapy have more, less, or the same level of pain relief than patients who are treated with other therapies?

Sherry et al. stated that after completion of VAX-D treatment, 13 of 19 (68.4%) of evaluable patients reported a successful treatment (defined as a 50% reduction in pain and any improvement in disability). Patients in the control group received TENS therapy, with none of 21 (0%) evaluable patients reporting a successful treatment. Shealy and Borgmeyer stated that 18 of 22 (81.8%) patients in the DRS therapy/TENS group reported an excellent/good treatment outcome (defined as ≥50% improvement). For patients in the traction/TENS control group this level of improvement was achieved in nine of 17 (52.9%). However, no evidence-based conclusion can be drawn from the limited, low quality evidence.

Key Question 3 b. Do patients treated with or decompression therapy for chronic low back pain due to herniated disc or degenerative disc disease utilize more, less, or the same number of adjunctive therapies (e.g., medications, bracing) than patients treated with other therapies?

The included studies did not provide data that would allow us to address this sub-question.

Key Question 3 c. Do patients treated with decompression therapy for chronic low back pain due to a herniated disc or degenerative disc disease return to work more quickly than patients treated with other therapies?

The included studies did not provide data that would allow us to address this sub-question. The study by Sherry et al. incorporated return to work into their definition of “remission,” but did not report return to work separately.(3)

Key Question 3 d. What is the duration of pain relief achieved, if any?

Sherry et al. (low quality study) reported that of 13 patients who had been “successfully treated” (defined as >50% improvement in pain and any improvement in disability ratings), two were lost to follow up, 1 had “suffered a significant other injury,” and of the 10 still available at 6 months, 7 still met criteria for a successful outcome.(3)

Key Question 3 e. If the therapy is effective, what are the patient characteristics/indications of those for whom it appears to work? Is the therapy effective for the Medicare population (over 65 years of age)?

As noted in Table 8, patients in the included studies were younger than the typical Medicare population. The limited quality and quantity of the evidence for efficacy of decompression therapy precludes the formulation of evidence-based conclusions regarding the efficacy of decompression therapy as a therapy for chronic low back pain for the Medicare population over the age of 65.

Key Question 3 f. If it works, which, if any, particular decompression protocol provides the most pain relief?

The two included studies did not compare different decompression protocols.

Quality

The quality and generalizability of the information were not formally evaluated because we included uncontrolled trials in the evidence base for this question. Uncontrolled studies cannot be used to determine causality or to estimate frequencies of adverse events; they can only be used to generate a list of adverse events possibly attributable to the device.

Findings

Adverse events reported in the included articles are presented in Table 9. The technology assessment performed for the Australian government included unpublished information on adverse events submitted by the manufacturer of the VAX-D system.

Table 9

Adverse Events and Harms Associated with Decompression Therapy.

Adverse events have been reported to occur in association with vertebral axial decompression therapy (one case report of an enlargement of an existing disc protrusion and reports of treatment-related pain). According to the MSAC report presented to the Australian government in 2001,(4) information supplied by the manufacturer indicated that approximately 10% of individuals who undergo vertebral axial decompression therapy are unable to tolerate “the positioning of the table or the distractive pressures” and discontinue treatment. However, none of the published studies utilized in our report, including the large case series of Gose et al.(5) (which enrolled 778 patients) reported that any patients were unable to tolerate treatment, although one percent of the patients reported an increase in pain. Of note, this case series was limited to patients who had received at least 10 treatments, suggesting that those who did not tolerate the therapy were screened out.

Key Question 4 a. Would the characteristics of the Medicare population (osteoporosis, etc.) increase the likelihood of adverse events compared to the trial populations?