Spinal cord stimulation for low back pain (Cochrane review)

Key points

1. 01Moderate-certainty evidence from the only available placebo-controlled trial at six months shows SCS probably does not improve back pain (4 points better on a 0-100 scale, 95% CI 8.2 better to 0.2 worse), function, or quality of life
2. 02No placebo-controlled trial has measured outcomes at 12 months or beyond, so long-term efficacy is completely unknown
3. 03When SCS was added to medical management (unblinded parallel trials), effects on pain and function were uncertain due to very low certainty evidence and high risk of bias
4. 04Surgical revision due to adverse events was required in 4-31% of SCS recipients across studies, depending on follow-up duration
5. 05Ten of 13 included studies had financial ties to SCS manufacturers, raising concern about publication and reporting bias

How it was conducted

Design

Cochrane systematic review and meta-analysis of randomised controlled trials and cross-over trials

Included studies

13 studies (10 placebo-controlled cross-over trials; 3 parallel-group trials comparing SCS plus medical management versus medical management alone)

Participants

699 adults with chronic low back pain (with or without leg pain); 55% female; mean age 47-59 years; mean symptom duration 5-12 years

Comparators

SCS versus placebo (stimulator switched off or inactive); SCS plus medical management versus medical management alone

Primary outcome

Mean low back pain intensity at long-term follow-up (>=12 months)

Search date

10 June 2022 (CENTRAL, MEDLINE, Embase, CINAHL)

What they found

• No placebo-controlled trials assessed SCS at long-term follow-up (>=12 months); this primary outcome was not estimable
• At 6 months (medium-term), SCS probably does not reduce back pain versus placebo: MD 4.00 points better (95% CI 8.20 better to 0.19 worse) on a 0-100 VAS; 1 study, 50 participants; moderate-certainty evidence
• At 6 months, SCS probably does not improve function versus placebo: MD 1.3 points better (95% CI 3.9 better to 1.3 worse) on 0-100 RMDQ/ODI; 1 study, 50 participants; moderate-certainty evidence
• At 6 months, SCS probably does not improve health-related quality of life versus placebo: MD 0.04 points better (95% CI 0.16 better to 0.08 worse) on EQ-5D 0-1 scale; 1 study, 50 participants; moderate-certainty evidence
• At immediate-term (<1 month), it is uncertain whether SCS reduces back pain versus placebo: MD 13.8 points better (95% CI 20.6 better to 7.0 better); I2=80%; 8 studies, 139 participants; very low-certainty evidence; effect disappeared in trials at low risk of detection bias (MD 3.00 points, 95% CI 9.3 better to 3.2 worse; 2 studies, 62 participants)
• Adding SCS to medical management may slightly improve function at medium-term follow-up: MD 16.2 points better (95% CI 19.4 better to 13.0 better); I2=95%; 3 studies, 430 participants; low-certainty evidence (reduced to 7.7 points, 95% CI 11.8 to 3.6, excluding the high-frequency SCS outlier)
• Adding SCS to medical management may slightly reduce opioid use: 15% fewer participants using opioids (95% CI 27% lower to 0% lower); I2=0%; 2 studies, 290 participants; low-certainty evidence
• Adverse events: 9 of 50 participants (18%) experienced adverse events over 12 months in the sole long-duration placebo-controlled trial; 4 of 50 (8%) required surgical revision within 12 months
• Surgical revision rates across trials of new SCS implants ranged from 4.1% at 8 weeks (1 study, 24 participants) to 30.9% at 24 months (1 study, 42 participants)
• At 24 months in one parallel trial, 13 of 42 SCS recipients (31%) required revision surgery; 45% experienced at least one adverse event

Limitations

• No placebo-controlled evidence exists beyond 6 months, making long-term efficacy unknowable from this review
• Most included studies had high or unclear risk of bias, particularly from inadequate blinding, failure to account for carryover effects in cross-over designs, attrition bias, and selective reporting
• The three unblinded parallel-group trials comparing SCS plus medical management versus medical management alone had uncontrolled and poorly reported co-interventions, making attribution of any benefit to SCS uncertain
• 10 of 13 studies had financial ties to SCS manufacturers, and 92% were judged at high risk of other potential biases, suggesting the evidence base may overstate benefits

Why it matters

For patients

Patients considering SCS for chronic low back pain should know there is currently no evidence it provides lasting benefit over a sham procedure, and that roughly 1 in 10 to 1 in 3 patients require repeat surgery for device-related complications within 2 years.

For clinicians

Clinicians should not offer SCS for low back pain outside a clinical trial given the absence of placebo-controlled long-term data, the high surgical complication and revision rates, and the very high proportion of studies with industry ties and methodological limitations.

For readers

This Cochrane Review is the most rigorous synthesis available on SCS for low back pain; its conclusion that the evidence does not support routine use aligns with the need for well-designed, long-term, blinded trials before this expensive, invasive procedure can be recommended.