15.2 Basic EMG Interpretation (02.G)
Key Takeaways
- Nerve conduction studies in Guillain-Barré syndrome may be normal or limited to F-wave and H-reflex abnormalities during the first days; do not delay intravenous immunoglobulin or plasma exchange for a prettier EMG.
- Demyelinating GBS (AIDP) shows slow velocities, prolonged distal latencies, conduction block, temporal dispersion, and late F waves; AMAN shows low CMAP amplitudes with relatively preserved speed.
- Low-rate (2–3 Hz) repetitive nerve stimulation with a compound muscle action potential decrement of 10% or more supports a postsynaptic neuromuscular-junction disorder such as myasthenia gravis.
- ICU-acquired weakness is usually critical-illness polyneuropathy, myopathy, or overlap: preserved sensory responses with low motor amplitudes point toward myopathy.
- Fibrillation potentials lag axonal injury by one to three weeks, so a day-2 needle exam cannot exclude new denervation.
EMG the intensivist actually has to read
Quick Answer: Nerve conduction studies (NCS) and electromyography (EMG) tell you whether weakness is demyelinating neuropathy, axonal neuropathy, myopathy, or neuromuscular junction (NMJ) failure. In Guillain-Barré syndrome (GBS), studies in the first days may be normal or show only late-response abnormalities — treat on clinical grounds. Acute inflammatory demyelinating polyradiculoneuropathy (AIDP) slows and blocks conduction; acute motor axonal neuropathy (AMAN) drops compound muscle action potential (CMAP) amplitude. Myasthenia gravis (MG) shows a ≥10% decrement on 2–3 Hz repetitive nerve stimulation (RNS). ICU-acquired weakness is usually critical-illness polyneuropathy (CIP), critical-illness myopathy (CIM), or both.
Independent OpenExamPrep teaching in this section covers basic EMG interpretation listed under Diagnostic studies in the ABPN Content Specifications. You are not sitting a neuromuscular medicine examination. You need enough pattern recognition to know when the study is too early, when it supports GBS subtype language, when it points to NMJ disease, and when “ICU weakness” is nerve, muscle, or both. Detailed GBS and myasthenic-crisis treatment lives later in this guide; here the tracing is the topic.
What the two tests measure
NCS stimulate a nerve and record a CMAP from muscle or a sensory nerve action potential (SNAP) from a sensory nerve. You care about distal latency, conduction velocity, amplitude, conduction block (abnormal drop in CMAP amplitude or area across a segment), and temporal dispersion (abnormal stretching of CMAP duration). F waves and H reflexes test proximal roots and plexus that a wrist-to-elbow shock never sees.
Needle EMG records muscle membrane at rest and motor-unit action potentials (MUAPs) during effort. Spontaneous fibrillations and positive sharp waves mean denervation or muscle-membrane irritability, but they take 1–3 weeks to appear after an axonal insult. Reduced recruitment of large, long MUAPs is neuropathic. Early recruitment of small, short, polyphasic MUAPs is myopathic. A patient who is too weak to activate muscle cannot give you a recruitment pattern — another reason early ICU studies are limited.
| Pattern | NCS headline | Needle EMG headline | ICU prototype |
|---|---|---|---|
| Demyelinating neuropathy | Slow velocity, long distal latency, block, dispersion, long or absent F waves; SNAP may show sural sparing in AIDP | Reduced recruitment; spontaneous activity if secondary axonal damage | AIDP |
| Axonal neuropathy | Low CMAP ± low SNAP; velocity relatively preserved | Fibrillations after a delay; large MUAPs later | AMAN, CIP, AMSAN |
| Myopathy | Low CMAP, preserved SNAP, velocity preserved | Small short MUAPs, early recruitment; irritability in necrotizing myopathy | CIM, toxic myopathy |
| Postsynaptic NMJ (MG) | Baseline CMAP often normal; low-rate decrement | Unstable MUAPs, jitter if single-fiber is done | Myasthenic crisis |
| Presynaptic NMJ (LEMS, some botulism) | Low resting CMAP; facilitation after brief exercise or high-rate stimulation | Similar instability | Lambert-Eaton, botulism |
GBS: demyelinating versus AMAN, and the first-week trap
AIDP, the common North American subtype, is a demyelinating polyradiculoneuropathy. Look for prolonged distal motor latencies, conduction velocity below about 70% of the lower limit in demyelinating range, conduction block, temporal dispersion, and F waves that are absent, impersistent, or markedly delayed. Sensory studies may show the sural-sparing pattern (abnormal median or ulnar SNAP with a relatively preserved sural), which favors AIDP over a length-dependent axonal polyneuropathy.
AMAN is an axonal motor neuropathy, often after Campylobacter, more often in other regions of the world but not rare in a U.S. ICU. Distal CMAP amplitudes fall; velocities and distal latencies stay relatively respectable. SNAPs are typically preserved (if sensory axons die too, the name becomes AMSAN). Early AMAN can show reversible conduction failure at nodes that mimics demyelinating block without true myelin stripping. That is why a single day-5 study labeled “AIDP” sometimes becomes AMAN on a week-3 repeat, and why serial NCS exist in the neuromuscular literature. For the intensivist, the actionable point is: axonal physiology predicts a harder recovery, but acute treatment of typical GBS is still IVIG or plasma exchange regardless of the first-pass label.
When EMG is too early: Lambert's old observation still holds in modern series — a fraction of GBS patients have no conduction abnormality in the first days. The earliest electrodiagnostic clues, when present, are often absent H reflexes and abnormal F waves, because roots are inflamed before the forearm segment slows. Distal CMAP amplitude drop can appear early in AMAN. Classic demyelinating block and dispersion often declare after the first week. Albumino-cytologic dissociation on CSF is also frequently missing on day 1–3 (next section). None of that delays intubation decisions or IVIG 0.4 g/kg/day for 5 days or plasma exchange when the clinical diagnosis is GBS and the patient is deteriorating.
Worked example: a 34-year-old with areflexic ascending weakness for 36 hours, FVC 18 mL/kg, NCS “normal motor velocities, F waves not obtained in the legs.” That study is compatible with very early GBS, not a reason to cancel immunotherapy and wait for a prettier EMG on day 10. Repeat electrodiagnosis later if the diagnosis is uncertain or if you need subtype language for counseling.
Phrenic NCS and diaphragm EMG exist for ventilator weaning questions; they are technically fussy and not required to start treatment. Independent practice at /practice/abim-neurocritical-care should punish waiting on the EMG, not failing to recite every Uncini criterion.
Myopathy versus neuropathy versus NMJ
Sort ICU weakness with four questions: are reflexes and sensation involved, are SNAPs preserved, is there a decrement on RNS, and was the patient already weak before this admission?
Neuropathy (GBS, CIP, vasculitis): distal more than proximal in length-dependent forms, hypo- or areflexia, sensory symptoms if sensory axons are hit. NCS: low SNAP if sensory fibers die; neuropathic recruitment.
Myopathy (CIM, statin/immune necrotizing myopathy, rhabdomyolysis recovering): often proximal, reflexes relatively preserved until weakness is profound, sensation spared. NCS: SNAPs present, CMAPs may be low if muscle cannot generate voltage. CK may be high in necrotizing disease and normal or modest in classic myosin-loss CIM.
NMJ: fatigable, extraocular and bulbar clues in MG; autonomic features and hyporeflexia in Lambert-Eaton myasthenic syndrome (LEMS). Pupils and descending paralysis raise botulism. NCS at rest may look almost normal in MG until you repeat-stimulate.
Repetitive stimulation in myasthenia — the one maneuver to know
Low-rate RNS (typically 2–3 Hz, train of 5–10 stimuli) in a weak, preferably proximal or facial muscle: compare the first CMAP to the fourth or fifth. A decrement ≥10% that repairs after 30–60 seconds of rest or after brief exercise (post-activation facilitation, then post-activation exhaustion a few minutes later) supports a postsynaptic disorder such as MG. Warm the limb; cold hides the decrement. Acetylcholinesterase inhibitors can mask it — hold pyridostigmine if the question is diagnostic and the airway is safe.
High-rate RNS (or 10 seconds of maximal exercise) looks for increment. A large increment (classically >60–100%) of a low baseline CMAP supports presynaptic disease (LEMS; variable in botulism). High-rate trains hurt; exercise facilitation is kinder in an awake patient. Do not perform painful 50 Hz trains on an unsedated ICU patient “because the textbook said so” if exercise will do.
Single-fiber EMG and jitter are more sensitive for MG and are not a 02:00 ICU procedure. A negative RNS in a strong distal hand muscle does not exclude ocular MG. In crisis, you treat on clinical and pulmonary mechanics first; RNS confirms the junction and helps separate MG from CIM when both are plausible after days of ventilation.
ICU-acquired weakness: CIP, CIM, overlap
After sepsis, multiorgan failure, and days of ventilation, ICU-acquired weakness is common. Electrodiagnosis splits it coarsely:
- CIP: axonal sensorimotor polyneuropathy. SNAPs and CMAPs both reduced. Needle: neuropathic recruitment, fibrillations after the first week or two.
- CIM: myosin thick-filament loss, often after glucocorticoids plus neuromuscular blockade stacked on sepsis. SNAPs preserved, CMAPs low, velocities preserved. Direct muscle stimulation (rarely available at 03:00) shows an inexcitable muscle in CIM versus an inexcitable nerve in CIP.
- Critical-illness neuromyopathy: the usual real-world overlap. Report both, treat the sepsis, stop unnecessary paralysis, and rehabilitate. Electrodiagnosis here is for sorting (is this still undiagnosed GBS? unmasked MG? spinal cord compression you missed?) more than for a unique drug.
GBS versus CIP: GBS is the admission weakness, often with a preceding infection, albuminocytologic dissociation, and demyelinating features in AIDP. CIP is acquired in the unit after days of critical illness and is axonal. Do not call day-2 septic weakness “CIP” before you have thought about GBS, spinal epidural abscess, and hypophosphatemia.
Limits of a basic intensivist read
Skin temperature, edema, ICU electrical noise, central lines, and an uncooperative or encephalopathic patient wreck amplitudes. A single low CMAP in a cold, edematous leg is not AMAN. Needle EMG in anticoagulated patients carries a hematoma risk — often skip deep paraspinals. You do not need to memorize every demyelinating cutoff percentage. You do need to know too early GBS studies can be normal, AIDP versus AMAN headlines, 10% decrement at 2–3 Hz, and SNAP-present points to muscle in the weak ICU patient.
Exam traps
Delaying IVIG because day-2 NCS is “not yet demyelinating.” Labeling AMAN as AIDP forever after one study with reversible conduction failure. Diagnosing MG from a 4% decrement in a cold abductor digiti minimi. Calling preserved SNAPs “normal, therefore not ICU weakness” while CMAPs are nearly absent (that is CIM until proven otherwise). Expecting fibrillations 12 hours after intubation for GBS.
A patient develops areflexic ascending paralysis over 40 hours. Forced vital capacity is 19 mL/kg. Nerve conduction studies on the same day show normal distal latencies and velocities; F waves are absent in the legs. Which action is most appropriate?
Which nerve-conduction pattern best supports acute motor axonal neuropathy rather than AIDP?
In a patient with fatigable ptosis and respiratory failure, which repetitive-stimulation result supports myasthenia gravis as a postsynaptic disorder?
After 12 days of septic shock and mechanical ventilation, an awake patient is flaccid. SNAPs are normal; CMAPs are markedly low; conduction velocities are preserved. Needle EMG shows small, short motor-unit potentials. Which localization is most consistent?