Table 3.

Representative clinical evidence for electromagnetic neuromodulation modalities in pain.

Modality	Pain condition/Population	Methods	Target	Study type/Evidence level	Outcomes	Direction/Size	Safety	Author (first)/Year
SCS	Chronic pain (mixed conditions)	SCS vs conventional medical management (CMM); includes direct and indirect comparisons across SCS types	Spinal cord pathways (dorsal columns/dorsal horn circuits)	Systematic review + network meta-analysis of RCTs	Pain intensity; ≥50% pain responder rate; physical function; health-related quality of life (HRQoL)	All SCS types were superior to CMM for pain intensity at last follow-up (mean difference range −2.37 to −5.55 on 0–10 scale) and for ≥50% pain relief [Odds ratio (OR) range 9.75–63.4]; HRQoL also improved	Safety not quantitatively synthesized; adverse event (AE) reporting was limited/inconsistent across trials	Eldabe 2026 (223)
SCS	Intractable spine and limb pain	Implanted spinal stimulation vs medical therapy (MT); indirect comparison of newer tech vs conventional via common comparator	Spinal cord pathways (dorsal columns/dorsal horn circuits)	Systematic review of RCTs + meta-analysis	≥50% pain relief; pain intensity [visual analog scale (VAS)]	SCS vs MT: ↑ ≥50% pain relief (OR 13.01, 95% confidence interval (CI) 4.96–34.17); VAS ↓ ([weighted mean difference (WMD) 1.43, 95% CI 0.16–2.71). Newer tech (10 kHz/burst/DRG] vs conventional: ↑ pain relief (OR 2.07, 95% CI 1.35–3.19)	AEs mainly hardware/procedure-related; serious neurologic complications were rare, with one subdural hematoma-associated death reported	Lamer 2019 (92)
SCS	Intractable spine and/or limb pain	Implanted SCS vs MT; includes conventional vs high-frequency SCS comparisons (RCTs; min follow-up ≥3 months)	Spinal cord pathways (dorsal columns/dorsal horn circuits)	Systematic review of RCTs + meta-analysis	Opioid/pain medication use (opioid reduction/cessation; Medication Quantification Scale (MQS); morphine equivalent dosage (MED)	Opioid reduction ↑ (OR 8.60, 95% CI 1.93–38.30); MQS ↓ (WMD −1.97, 95% CI −3.67 to −0.27). HF vs conventional SCS: opioid reduction not statistically significant (NS) (OR 1.43, 95% CI 0.74–2.78); MED change NS	AEs variably reported; mostly hardware/implant site pain and lead migration; rare serious neurologic events reported	Pollard 2019 (224)
SCS	Low back pain	Implanted epidural spinal cord stimulation (electrodes + IPG); trials include placebo-controlled cross-over designs and parallel trials adding SCS to medical management	Spinal cord pathways (dorsal columns/dorsal horn circuits)	Cochrane review	Pain intensity, function, quality of life (QoL), AEs (incl. serious AEs (SAEs), need for revision surgery)	6 mo vs placebo: SCS probably does not improve pain/function/QoL (moderate certainty); ≥12 mo mean pain not evaluated. SCS + MT vs MT: pain/QoL uncertain (very low certainty); function/opioid use may slightly improve (low certainty)	Surgical/device AEs (infection, lead migration; occasional SAEs. Revision surgery reported (up to ∼31% at 24 mo in one trial); AE reporting inconsistent across studies	Traeger 2023 (118)
DRG	Complex regional pain syndrome (CRPS) I/II; causalgia (lower extremity focal/dermatomal pain)	Implanted DRG stimulation vs implanted conventional SCS	Dorsal root ganglion → dorsal horn circuits	Implanted DRG stimulation vs implanted conventional SCS (randomized comparative effectiveness trial)	Treatment success (composite incl. ≥ 50% VAS reduction + no stimulation-related neurologic deficit), VAS pain intensity; paresthesia specificity/postural variation	Treatment success higher with DRG vs SCS at 3 mo: 81.2% vs 55.7% (P < 0.001); maintained at 12 mo: 74.2% vs 53.0%; less postural variation in paresthesia and less extraneous stimulation in non-painful areas with DRG	Device-related and SAEs not different between groups; typical implant/device risks (e.g., infection, lead migration/lead issues, pocket pain) reported	Deer 2017 (106)
MCS	Refractory chronic neuropathic pain	Implanted epidural motor cortex stimulation with randomized active vs sham crossover comparison	M1	Single center RCT	Pain intensity [numerical rating scale (NRS)]; responder status	Pain reduction during active stimulation was variable; long-term responders were observed in a subset of patients (∼39% long-term responders), but overall efficacy was heterogeneous	Invasive neurosurgical/device-related risks, including infection, seizures, and hardware complications; no major unexpected safety signal in this small trial	Hamani 2021 (52)
VNS (implantable/invasive)	Chronic pain (mixed conditions)	Implanted cervical VNS	Cervical vagus nerve	Narrative review; invasive-VNS evidence subset	Pain intensity/pain relief	Limited and heterogeneous pain-specific clinical evidence; possible benefit reported in selected conditions, but no firm efficacy conclusions	Implant-related AEs possible	Shao 2023 (225)
VNS (implantable/invasive)	Treatment-resistant fibromyalgia	Implanted cervical VNS	Cervical vagus nerve	Prospective open-label Phase I/II proof-of-concept trial	Pain intensity; patient global improvement; physical function; widespread pain and tender-point criteria	Preliminary improvement was reported in a subset, with additional responses during follow-up; interpretation is limited by the small, uncontrolled design	Implantation-, device-, and stimulation-related AEs occurred; one device required revision and one was later explanted for lack of perceived benefit	Lange 2011 (130)
aVNS	Chronic pain (mixed conditions)	Auricular VNS, transcutaneous or percutaneous, vs sham/control	Auricular branch of the vagus nerve	Systematic review and meta-analysis	Pain intensity/analgesic use	Significant reduction in chronic pain intensity vs. sham/control; greater effect with percutaneous than transcutaneous aVNS	Very well tolerated; mainly mild local AEs	Duff 2024 (132)a
n-VNS	Migraine	Noninvasive VNS, including cervical (n-cVNS) and auricular (n-aVNS) approaches, vs sham/control	Cervical vagus nerve/auricular branch of the vagus nerve	Systematic review and meta-analysis of RCTs	Monthly migraine/headache days, pain-free rate within 2 h, ≥50% responder rate, headache intensity, AEs	n-cVNS: ↑ ≥ 50% responder rate (OR 1.64, 95% CI 1.10–2.47); no significant reduction in migraine days or headache days. n-aVNS: ↓ migraine days (mean difference (MD) −1.8, 95% CI −3.34 to −0.26) and ↓ headache intensity (standardized MD (SMD) −0.7, 95% CI −1.23 to −0.17)	n-cVNS was safe and well tolerated; no significant increase in AEs vs sham	Song 2023 (226)
n-VNS	Cluster headacheb	Noninvasive VNS, mainly cervical n-VNS, vs sham/control	Cervical vagus nerve	Systematic review	Attack frequency, pain intensity, responder rates, AEs	Moderate-to-high-quality evidence suggested benefit in reducing cluster headache frequency and intensity; benefit was clearer in episodic than chronic cluster headache, and heterogeneity prevented meta-analysis	Generally safe and well tolerated in noninvasive studies	Fernandez-Hernando 2023 (227)
tDCS	Chronic pain (mixed conditions)	tDCS vs sham	M1 (most studies)/other cortical targets	Cochrane systematic review of sham-controlled randomized and quasi-randomized trials	Pain intensity, disability, QoL, AEs	Short-term pain reduction vs sham: Standardized Mean Difference (SMD) −0.43 (95% CI −0.63 to −0.22), corresponding to ∼0.82 points on a 0–10 scale or ∼17% reduction; no clear disability benefit; long-term effect not supported. Evidence quality very low, with heterogeneity and possible small-study bias	Mostly minor/transient AEs (e.g., headache, scalp irritation, tingling, redness); no major lasting safety signal reported	O’Connell 2018 (153)
tDCS	Nonspecific chronic low back pain	tDCS (multiple sessions) vs sham	M1	Systematic review and meta-analysis of RCTs	Pain intensity, disability, AEs	No significant post-treatment benefit for pain (SMD 0.378, 95% CI −0.264 to 1.020; P = 0.249) or disability (SMD 0.143, 95% CI −0.214 to 0.499; P = 0.434); evidence does not support clinical use based on available data	No significant AEs reported	Alwardat 2020 (228)
tDCS	Painful diabetic polyneuropathy	Anodal tDCS over left M1 vs sham; 2 mA, 20 min, 5 consecutive days	M1	Randomized, sham-controlled trial	VAS pain intensity, Clinical Global Impression, pain threshold	Significant pain reduction vs sham, sustained at 2 and 4 weeks	No major safety signal reported	Kim 2013 (161)
rTMS	Chronic pain (mixed conditions)	rTMS vs sham	M1 (most studied). Some studies targeting prefrontal cortex	Cochrane systematic review of sham-controlled randomized and quasi-randomized trials	Pain intensity, disability, QoL, AEs	Small short-term effect overall: SMD −0.22 (95% CI −0.29 to −0.16), corresponding to ∼0.40 points on a 0–10 scale; high-frequency single-dose M1 stimulation showed a somewhat larger short-term effect but still did not clearly meet clinical importance thresholds; no clear disability benefit	Mostly minor/transient AEs; two seizures reported with active rTMS across included studies; AE reporting inconsistent	O'Connell 2018 (153)
rTMS	Neuropathic pain	rTMS vs sham	M1 (most studied)	Systematic review and meta-analysis of RCTs	Pain intensity	rTMS was better than sham in reducing pain; significant analgesic effects were associated with M1 stimulation and high-frequency rTMS (≥5 Hz); no evident difference between single- and multi-session protocols	No significant difference in AEs vs. control; reported adverse effects were generally mild/transient	Jiang 2022 (229)
rTMS	Fibromyalgia	10-Hz rTMS vs sham	M1 and DLPFC	Systematic review and meta-analysis of RCTs	Pain intensity, QoL, depression	10-Hz rTMS reduced pain and improved QoL; subgroup analysis found no significant difference between M1 and DLPFC for pain, although M1 showed a nonsignificant trend toward greater analgesia	No pooled AE analysis reported	Zhu 2023 (230)
Deep TMS	Central neuropathic pain	Deep rTMS (H-coil) vs sham; superficial rTMS comparator	M1	Randomized sham-controlled multicenter crossover trial	Pain intensity	Deep and superficial M1 rTMS both produced analgesic effects of similar magnitude; deep-rTMS effects appeared later but tended to last longer on some secondary outcomes.	Mostly mild-to-moderate AEs (mainly stimulation-site pain and headache); no treatment-related SAEs reported	Bouhassira 2024 (231)
TBS	Chronic pain (mixed conditions; mainly neuropathic pain)	Theta burst stimulation alone or as priming before HF-rTMS	Primarily M1	Systematic review	Pain intensity/pain relief	Some studies reported pain relief, particularly with iTBS of M1 alone or as priming before HF-rTMS, but evidence remains limited and insufficient for firm conclusions	Not systematically reported in review summary	Mussigmann 2025 (232)a
tACS	Chronic pain (mixed conditions)	tACS vs sham	Mixed targets (M1, bilateral frontal lobe, occipital lobe)	Systematic review of RCTs	Pain intensity	Insufficient evidence for a conclusive effect overall; findings were mixed across fibromyalgia, low back pain, and migraine	Safety not established through pooled analysis; evidence base remains limited	Chang 2023 (233)
CES	Chronic pain (mixed conditions)	CES vs sham/usual care	Cranial application	Systematic review of RCTs	Pain intensity, function, AEs	Evidence was insufficient/inconclusive for clinically important pain benefit across chronic pain conditions	No SAEs reported; minor symptoms occurred, but AEs reporting was limited/inconsistent	Shekelle 2018 (173)
CES	Fibromyalgia	CES vs sham vs usual care	Earlobe electrodes	RCT	Pain intensity, fatigue, sleep disturbance, functional status	CES was associated with greater decreases in pain and sleep disturbance, with improved functional status vs. sham and usual care over time; fatigue also trended to improvement	Very few AEs reported; no blood-pressure safety signal was observed	Taylor 2013 (178)
TENS	Chronic pain (mixed conditions)	TENS vs sham/placebo	Surface electrodes at the site of pain or over nearby peripheral nerve bundles	Systematic review and meta-analysis of RCTs	Pain intensity during or immediately after TENS	TENS was associated with lower pain intensity during or immediately after treatment vs. placebo; subgroup analyses did not identify pain duration (acute vs chronic) or diagnosis as effect modifiers	No SAEs directly attributable to TENS were reported; AEs were generally mild/infrequent and not significantly different from comparators	Johnson 2022 (234)a
TENS	Fibromyalgia	TENS vs sham/placebo	Surface electrodes over painful areas	Cochrane systematic review of RCTs	Pain intensity, fatigue, function, QoL, AEs	Very low-certainty evidence; unable to determine whether TENS improves pain vs. sham	No SAEs clearly attributable; AE reporting limited/inconsistent	Johnson 2017 (235)
TENS	Fibromyalgia	TENS vs sham, no intervention, or active comparators (e.g., exercise, physiotherapy modalities, pharmacologic therapy)	Surface electrodes over painful areas	Systematic review and meta-analysis of controlled clinical trials	Pain intensity, disability, QoL	TENS was associated with ↓ pain (SMD −0.61, 95% CI −1.00 to −0.16), ↓ disability (SMD −0.27, 95% CI −0.41 to −0.12), and ↑ physical QoL (SMD 0.26, 95% CI 0.08 to 0.44)	AEs not systematically reported; no pooled safety analysis	Garcia-Lopez 2024 (236)
PNS	Chronic pain (mixed conditions)	PNS vs comparator/sham/usual care	Peripheral nerves	Systematic review	Pain relief, functional improvement, AEs	For chronic pain, evidence was strongest for refractory peripheral nerve injury (level II), with lower-level evidence for some other indications	Safety generally acceptable, but AEs varied by device, target, and study	Helm 2021 (203)
PNS	Hemiplegic shoulder pain after stroke	Single-lead percutaneous PNS vs usual care	Peripheral nerve target in the shoulder/deltoid region	Single center RCT	Pain intensity, pain interference, QoL, AEs	Greater reduction in pain vs. usual care, maintained at least 12 weeks after treatment	No serious infections; lead dislodgement, retained fragments, pruritus, and transient implant-site pain were reported	Wilson 2014 (237)
PNFS	Localized chronic intractable back pain	Randomized crossover comparison of PNFS stimulation conditions; follow-up in implanted responders	Subcutaneous painful field of the back	RCT	Pain intensity, disability, patient satisfaction, AEs	PNFS improved pain during the randomized crossover phase, and responders who proceeded to permanent implantation maintained significant pain reduction through 52 weeks	Generally well tolerated; device/procedure-related AEs were reported	McRoberts 2013 (238)
SCS + PNS	Brachial plexopathy/focal upper-extremity neuropathic pain	Combined SCS with PNS	Cervical spinal cord + radial nerve (selected peripheral target)	Case report	Pain intensity/medication use/functional status	After loss of benefit from prior SCS alone, addition of radial-nerve PNS was associated with disappearance of background pain, residual breakthrough pain at NRS 2–3/10, reduced analgesic use, and maintained benefit at 6 months; evidence is limited to a single case	No procedural complication was reported in this case; AE characterization is limited by single-case evidence	Choi 2016 (239)
SCS + PNFS	Chronic low back and leg pain	Combined SCS with PNFS	Dorsal columns + painful subcutaneous field	Case series	Overall pain control/pain coverage	Most patients reported better overall pain control and improved coverage with the combination than with either modality alone; in some cases, PNFS was added after SCS alone did not adequately control low back pain	No trial-period complications were reported (including no infections or dislodged/pulled wires); broader implant-related risks remain relevant, but formal AE characterization was limited in this small case series	Bernstein 2008 (212)

For each modality–indication pair, we prioritized higher-level evidence syntheses (Cochrane reviews, systematic reviews, and meta-analyses). When such evidence was limited, inconsistent, or unavailable, we selected representative sham-controlled randomized trials or other illustrative prospective studies, emphasizing studies with clearly described stimulation parameters and quantitative pain outcomes. This table is intended to summarize representative evidence rather than provide an exhaustive systematic inventory.

^aThe cited study includes chronic, acute, and/or experimental pain populations; this table emphasizes the findings most relevant to chronic pain.

^bIncludes episodic and chronic cluster headache populations.