Spinal Cord Stimulation

1.What Problem Is SCS Actually Solving?

The most important preoperative question is not “Has the patient failed conservative therapy?” It is more specific: is the dominant pain generator neuropathic, anatomically plausible, refractory, and measurable? SCS is best supported when the patient’s pain is burning, electric, allodynic, shooting, distal, radicular, or otherwise neuropathic. It is less reliable when the dominant complaint is axial mechanical pain from instability, severe stenosis, deformity, fracture, inflammatory arthropathy, or another nociceptive driver that has not been addressed.

Modern SCS works through several stimulation paradigms. Traditional tonic SCS creates paresthesia over the painful territory. High-frequency 10 kHz stimulation is paresthesia-free. Burst and differential targeted multiplexed programs attempt to modulate pain networks without relying entirely on paresthesia overlap. Closed-loop systems record evoked compound action potentials and adjust output to maintain a target dose of neural activation. These distinctions matter, but they do not rescue poor selection. The phenotype still comes first.

2.The Published Consensus Spine

The public guidance most useful for trainees comes from two overlapping families of work. First, the Neurostimulation Appropriateness Consensus Committee, or NACC, has published practical recommendations on infection prevention, surgical technique, neurologic-injury mitigation, cervical neurostimulation, and long-term optimization or salvage. Second, multisociety consensus guidance on patient selection and trial stimulation emphasizes psychological screening, anticoagulation and infection risk, trial interpretation, and expectation-setting.

Those documents do not replace local protocols, device labeling, or judgment. They do, however, give trainees a reliable safety culture: optimize modifiable infection risk, respect neuraxial anticoagulation rules, avoid deep sedation when patient feedback is needed, document device MRI conditionality, and make the trial answer a predefined clinical question rather than a vibe check.

3.Vocabulary That Changes Management

4.Painful Diabetic Neuropathy: The Strong New Indication

Painful diabetic neuropathy is the modern SCS indication that most clearly changed the teaching script. The SENZA-PDN randomized trial compared conventional medical management alone with conventional medical management plus high-frequency 10 kHz SCS in patients with refractory lower-limb painful diabetic neuropathy. The SCS group had large and durable improvements in pain and quality of life, with follow-up reported through 24 months and longer-term survey data extending the durability story.

That does not mean every patient with diabetes and foot pain should be sent for an implant. It means that after neuropathic medications, diabetes care, foot care, and vascular/infection risks are addressed, SCS should be part of the discussion for persistent disabling PDN. The patient needs a foot exam, a realistic discussion of wound and infection risk, and a clear understanding that the primary goal is pain and function, not “curing neuropathy.”

5.Nonsurgical Refractory Back Pain: Not a Shortcut Around Surgery

PSPS Type 1 and nonsurgical refractory back pain describe patients with chronic refractory back pain, with or without leg pain, who have not had prior spine surgery and are not appropriate candidates for corrective spine surgery. This category matters because randomized and prospective data support SCS in selected patients, including high-frequency and burst paradigms. It also matters because it can be abused if we use it before doing the harder diagnostic work.

The trainee’s job is to make sure “nonsurgical” really means nonsurgical. The patient should have appropriate imaging, surgical assessment when indicated, rehabilitation, behavioral health review, and nonoperative/interventional care. Progressive neurologic deficit, untreated severe stenosis, instability, deformity, fracture, infection, malignancy, or inflammatory disease should not be bypassed by an SCS referral.

6.Length-Dependent Peripheral Neuropathy: Same Shape, Weaker Evidence

Length-dependent peripheral neuropathy is clinically seductive because it can look like PDN: distal burning, painful numbness, allodynia, and stocking-glove symptoms. But evidence does not transfer automatically. A patient with chemotherapy-induced neuropathy, idiopathic small-fiber neuropathy, B12 deficiency, paraproteinemia, alcohol-related neuropathy, hereditary neuropathy, or autoimmune neuropathy is not the same evidence object as a SENZA-PDN patient.

That does not make SCS unreasonable. It makes the conversation more individualized. Before a trial, confirm the diagnosis and look for treatable causes: diabetes or prediabetes, B12 deficiency, thyroid disease, renal disease, paraproteinemia, alcohol or toxin exposure, medication toxicity, autoimmune disease, and hereditary patterns. If the neuropathy remains severe, refractory, and function-limiting, a trial can be considered, but the patient should know that the evidence base is less mature.

7.SCI and MS Central Pain: Proceed With Humility

Central neuropathic pain is different. In spinal cord injury, the pain may be at-level, below-level, radicular, musculoskeletal, spasticity-related, pressure-injury-related, or mixed. In multiple sclerosis, pain can come from central lesions, spasticity, trigeminal neuralgia, musculoskeletal compensation, or treatment effects. SCS may help selected patients, but the outcomes are less predictable than in peripheral neuropathic pain.

The most defensible posture is careful selection and honest uncertainty. Optimize spasticity, sleep, mood, rehabilitation, pressure injury, bowel/bladder triggers, and medication burden first. Ask neurology and rehabilitation colleagues to help define the pain phenotype. If a trial is offered, define a specific target symptom, such as lower-extremity burning pain, not global weakness, fatigue, gait decline, or “quality of life” in the abstract.

8.Before the Trial, Define Success

The old shorthand is “at least 50% pain relief.” That remains useful, but it is too narrow. For paresthesia-based therapy, the trial asks whether stimulation covers the painful territory and whether the induced sensation is tolerable. For paresthesia-free therapy, the patient may never feel stimulation, so you need other anchors: pain diary, walking tolerance, sleep, allodynia, medication use, patient global impression, and whether the patient understands what living with the device will require.

9.Safety Details Trainees Should Not Miss

Infection prevention starts before the prep stick. Optimize diabetes, smoking, nutrition, skin disease, immune suppression, and active infection risk. Anticoagulant and antiplatelet management should be treated with the seriousness of a neuraxial procedure. Device documentation matters: future MRI conditionality, model, lead location, restrictions, and implant date should be easy to find.

For percutaneous trial placement, avoid deep sedation or general anesthesia when patient feedback is needed, unless there is a specific technical or patient factor that justifies it. After implant, trainees should know the common failure modes: lead migration, inadequate coverage, uncomfortable stimulation, infection, wound breakdown, loss of efficacy, battery or pocket pain, need for reprogramming, revision, or explant.

10.Optimizing the Host: Antibiotics, Glucose, and Coagulation

The implant is only as safe as the host you implant into, and most of the modifiable risk is addressed before the patient reaches the table. The recurring theme of the NACC infection-prevention guidance is to optimize the host first: bring the HbA1c toward target (commonly cited goals are below roughly 7–8%), insist on smoking cessation (ideally several weeks preoperatively, where most of the wound-healing benefit accrues), address obesity and malnutrition, and clear remote sources of bacteremia such as active dental disease and urinary infection. High systemic opioid doses are associated with poorer device outcomes, so reducing the opioid burden before implant is part of selection, not an afterthought. Hair is clipped, not shaved, and chlorhexidine skin antisepsis is standard.

Antibiotic prophylaxis. A weight-based first-generation cephalosporin (cefazolin) given within 60 minutes of incision covers the great majority of relevant skin flora; NACC explicitly does not endorse routine vancomycin, which is reserved for documented MRSA risk or beta-lactam allergy (begun within 120 minutes), with clindamycin as the other alternative. The published evidence does not support routine prolonged postoperative oral antibiotics after an appropriate preoperative dose, although practice varies. Intra-pocket vancomycin powder and antimicrobial envelopes are extrapolated from cardiac-device data and remain judgment calls rather than requirements. A consolidated reference on prophylaxis dosing, intraoperative adjuncts, MRSA decolonization, and empiric therapy for established infection — across SCS, DBS, and epilepsy surgery — is available as a separate quick-reference page.

Anticoagulation and antiplatelet management. Because epidural lead placement carries the same catastrophic risk as any neuraxial procedure — epidural hematoma — antithrombotic drugs are held for drug-specific intervals and restarted on a drug-specific schedule, following the NACC bleeding-and-coagulation guideline and the parallel ASRA recommendations. The hold times are agent-specific and change as new drugs appear, so the safe habit is to look each drug up for each patient rather than rely on memory; modifiers such as advanced age can lengthen the recommended interval. Antithrombotics are generally held through the trial as well, which creates a practical tension in patients who cannot safely stop them for long: options include shortening the trial or performing a staged trial (placing and anchoring the permanent lead, tunneling an externalized extension for the trial, and creating the pocket) so the neuraxis is entered only once. A consolidated, drug-by-drug ASRA/NACC hold-and-restart table is available as a separate quick-reference page.

11.Surgical Technique in Brief

SCS leads come in two forms. Percutaneous cylindrical leads are placed through a Tuohy needle into the dorsal epidural space and are used for trials and for many permanent implants; paddle (surgical) leads are placed through a small laminotomy, offer a broader, more stable contact array, and are favored when percutaneous steering fails, when leads have migrated, or when the epidural space is scarred. Paddle placement is done asleep with neuromonitoring; percutaneous placement keeps the patient awake or lightly sedated because the patient is the best intraoperative monitor.

Percutaneous placement. The patient is positioned prone with the lumbar lordosis flattened, and fluoroscopy is set to a true AP with the midline marked. The Tuohy needle enters one to two vertebral levels below the target interspace so that it reaches the epidural space at a shallow angle (well under 45 degrees), which keeps the trajectory dorsal and protects against a ventral or intrathecal puncture. Epidural entry is confirmed by loss of resistance, and the lead is steered cranially with a curved stylet under intermittent live fluoroscopy. Dorsal, midline position is confirmed on the lateral view (the lead lies posterior in the canal) and on AP (contacts over the spinous-process line). Target levels follow the pain territory — roughly T8–T10 for low-back and leg pain, more rostral for upper-limb or cervical pain (leads driven to around C2–C4 from a low cervical/high thoracic entry). For bilateral pain, two leads straddle the midline.

Anchoring, pocket, and closure. For a permanent system, the lead is anchored to the thoracodorsal (supraspinous) fascia with a dedicated anchor as close to the entry point as possible, taking care not to advance or withdraw the lead while removing the needle; a strain-relief loop is left at the fascial entry and again beneath the generator, because lead migration and fracture are the commonest hardware failures. The pulse generator sits in a subcutaneous pocket — commonly in the upper buttock/flank below the iliac crest, away from the beltline — developed in a clean tissue plane and sized to the device to avoid both dehiscence and seroma. Tunneling runs from the lead toward the pocket, and the wound is closed in layers. Counsel the patient on the early warning signs that matter: fever or feeling unwell (infection) and new leg numbness with severe back pain (epidural hematoma — a surgical emergency).

12.Patient Selection and Psychological Evaluation

Selection is multidisciplinary, and the psychological evaluation is part of it — not a bureaucratic hurdle. The screening looks for factors that independently predict poorer device outcomes: untreated or poorly controlled depression and anxiety, high pain catastrophizing, somatization, active substance-use disorder, unrealistic expectations, secondary-gain or unresolved compensation/litigation issues, and the cognitive capacity and social support needed to operate and live with an implanted device. The point is not to disqualify patients but to identify and optimize what is modifiable before committing them to hardware.

Two practical companions to the psychological screen complete the selection picture. First, reduce the systemic opioid burden where possible: very high daily opioid doses are associated with lower device success, so weaning is part of preparation rather than something to defer until after implant. Second, set explicit, measurable goals with the patient — a target reduction in pain, better sleep, more walking, less medication — so that the trial answers a defined question and the patient understands that the aim is meaningful improvement, not cure. The trial itself is the strongest selection tool: it converts a prediction into an observation in that individual patient.

13.The MRI-Device Dilemma

Patients with implanted stimulators frequently need MRI for unrelated reasons, and the recurring question — can they be scanned? — turns on the difference between three labels. MR Safe means no electromagnetic hazard; MR Unsafe means not safe or not tested; MR Conditional means safe only when a specific set of conditions is met. “The patient has an MRI-conditional system” is therefore the beginning of the analysis, not the end.

Of the three magnetic fields in an MRI, the one that matters most for a stimulator is the radiofrequency field: a lead behaves like an antenna, concentrating RF energy and producing focal heating at the lead tip that can injure neural tissue, which is why scans are constrained by SAR (specific absorption rate) limits. The static (Tesla) field is usually the least problematic because these systems are largely non-ferromagnetic, and the gradient field is a secondary concern. Conditionality therefore depends on the specifics: 1.5T versus 3T, the device placed in a dedicated MRI mode or turned off, SAR and scan-time limits, and whether the labeling permits full-body scanning or only head-and-extremity coils.

The single most important principle is that a system is only as conditional as its weakest component. A fractured lead, an abandoned or capped-off retained lead, or a non-conditional extension or adapter spliced onto an otherwise conditional system can render the whole construct MR-unsafe — and a broken or disconnected lead is itself a focal-heating hazard. The major manufacturers (Boston Scientific, Medtronic, Abbott, Nevro, Saluda) all offer systems with differing MRI labeling that is updated over time, so the only safe practice is to verify the exact implanted combination against the current device manual rather than relying on the brand.

14.A Field Guide to Wound and Hardware Complications

The generator pocket is the largest dissection in the operation and the commonest site of trouble, so most complications are recognized and managed there. Distinguish the fluid collections: a seroma is a diffuse serous collection managed conservatively with a binder and observation, drained only if it is large (roughly > 25 mL), rapidly accumulating, or threatening the wound — and a drain is pulled once output falls below about 25 mL/day on two consecutive days; a hematoma is a confined blood collection that, if expanding or under tension, needs evacuation and control of the bleeding source (a nidus for infection). The catastrophic collection is the epidural hematoma — new or worsening leg numbness with severe back pain is a surgical emergency, not an overnight observation.

Infection declares itself as a spectrum: superficial cellulitis (erythema, non-tender) may respond to antibiotics, but a tense, tender, purulent pocket infection or abscess generally requires surgical drainage and, once infection reaches the device or the deep space, explantation of the system — partial salvage rarely succeeds, and reimplantation is deferred for months. Wound dehiscence and skin-flap ischemia come from too few closure layers, an oversized pocket, closure under tension, or dissection in the wrong plane; they are commoner in diabetics and smokers and are managed by debridement and layered re-closure. Lead migration and fracture are the dominant hardware failures: secure fascial anchoring and strain-relief loops prevent migration, and a sudden change in impedance points to fracture or disconnection. Prevention is mostly technical — atraumatic dissection, disciplined hemostasis, a correctly sized pocket, layered closure, and strain relief.

15.The Modern Mental Model