The scan showed what the nerve looks like. Nobody asked what it is doing.

EMG AND NERVE CONDUCTION STUDIES

An MRI shows what a nerve looks like. Electrodiagnostic testing shows what it is doing. In the arm those are different questions, and the second one is the one that decides treatment.

Why imaging cannot answer this

A cervical MRI in anybody past forty shows foraminal narrowing and disc protrusion at several levels. Those findings are near-universal and mostly silent. The scan cannot tell you whether the narrowing it found is producing the symptoms you have, and it cannot see the wrist or the elbow at all.

Electrodiagnostic testing measures function. It asks whether a nerve is conducting normally, where along its course conduction fails, and whether the muscles that nerve supplies show the electrical signature of denervation. That is a physiological answer to a physiological question.

The two halves, and what each is for

Nerve conduction studies stimulate a nerve at one point and record downstream, measuring speed and amplitude. They are excellent at localizing a peripheral entrapment — the median nerve at the wrist, the ulnar nerve at the elbow — and are the reference standard for confirming carpal tunnel syndrome and grading its severity.

Needle electromyography records electrical activity from inside the muscle. This is the half that matters for a nerve root, because a root lesion produces changes in the muscles that root supplies. Finding those changes in a myotomal distribution — and not in muscles supplied by other roots — localizes the level.

The two are done together for a reason. Testing only conduction misses radiculopathy; testing only the needle misses the entrapment sitting alongside it.

The limitation to understand before you have it

A normal study does not exclude cervical radiculopathy. Needle EMG detects motor axon loss. A root that is irritated and painful but not yet losing axons — which describes a great many radiculopathies, particularly early or predominantly sensory ones — can produce an entirely normal study.

This gets people into trouble. A normal EMG is reported, the patient is told their neck is fine, and a real radiculopathy goes untreated. The correct reading is narrower: the study did not find evidence of axon loss, which is not the same as finding evidence of nothing.

The reverse also holds. Electrodiagnosis is genuinely useful in cervical radiculopathy precisely because it excludes the mimics — and a well-conducted study that finds a median neuropathy at the wrist instead has changed your treatment entirely.

What it feels like

Two parts. The conduction half involves brief electrical pulses at the skin — startling rather than painful, like a static shock. The needle half uses a fine electrode into several muscles, held briefly while you contract. Uncomfortable, over quickly, and no sedation.

It takes forty minutes to an hour. There is no preparation beyond avoiding lotion on the arms, and you drive yourself home.

When it changes the plan, and when it does not

• Worth doing: hand symptoms where the distribution is unclear, suspected double pathology, weakness without a clear cause, symptoms that do not match the imaging, and before any decompression where the level is in doubt.
• Usually not needed: a textbook C6 or C7 story with matching examination and matching imaging, improving on its own.
• Timing matters: denervation changes take two to three weeks to appear on needle EMG. A study done in the first two weeks after symptom onset can be falsely normal for that reason alone.

What we do with the result

A confirmed entrapment goes to the treatment for that entrapment, which is frequently not a neck treatment at all. A confirmed radiculopathy with a named level is treated as radiculopathy. Both together is double crush, and the order of treatment then becomes the question.

A normal study in somebody with a convincing story does not close the case here. It narrows it, and the workup continues on examination rather than stopping because a test was unremarkable.

Who should not have it, or should wait

People on anticoagulants can generally still have nerve conduction studies; needle EMG in deeply anticoagulated patients is a judgment call and sometimes deferred. A pacemaker or implanted stimulator is not a contraindication but the operator needs to know. Significant lymphoedema in a limb changes where the needle goes.

And timing again, because it is the most common avoidable error: wait three weeks from symptom onset before the needle study, or accept that a normal result may mean nothing more than that the denervation has not developed yet.

Reading your own report

Two words carry most of the meaning. Latency is how long the signal takes — prolonged latency across the wrist is the carpal tunnel finding. Amplitude is how much signal arrives; reduced amplitude means axons have been lost rather than merely slowed, and it is the more serious finding.

On the needle side, fibrillations and positive sharp waves are the electrical signature of denervated muscle. Finding them in muscles supplied by one root, and not in neighboring myotomes, is what localizes a level. A report describing “chronic neurogenic changes” without saying which muscles is not much use, and it is reasonable to ask.

What it cannot do

It cannot assess a purely sensory root problem well. It cannot see the C5 root reliably because the muscles are awkward to sample. It cannot tell you whether a disc or a bone spur is doing the compressing. And it cannot grade pain.

Those limits are why it is one input rather than the answer, and why it belongs alongside an examination rather than instead of one.

What people ask about nerve testing

Does a normal EMG mean nothing is wrong with my neck?

No, and this is the most important thing on this page. Needle EMG detects axon loss. An irritated, painful root that has not yet lost axons produces a normal study, which is common in early and sensory-predominant radiculopathy. What a compressed root actually causes.

Is it painful?

Uncomfortable rather than painful. Brief shocks for the conduction half and a fine needle into several muscles for the other. Most people describe it as easier than they expected and are glad it is over. Double crush, where two sites compress one nerve. The published tolerability data is better than the reputation.

Can it tell the difference between my neck and my wrist?

That is precisely what it is best at, and it is the reason to do it when the hand symptoms are ambiguous. It can also find both at once. When the hands are the problem.

Should I have it before an MRI?

Often they answer different questions and both get done. If the ambiguity is where along the arm the problem sits, electrodiagnosis is the more informative test and imaging alone will not settle it. How the two are separated.

Related reading

• Cervical radiculopathy
• Carpal tunnel, or the neck?
• Double crush syndrome
• When your hands are the problem
• Cubital tunnel, or the neck?
• Why your MRI is normal

For clinicians referring a patient for electrodiagnostic testing, what the study answers and what comes back is set out on information for referring physicians.

The study confirms a level that the examination has already proposed. How the finger map generates that proposal.

The study is tolerated better than it is anticipated. What the two halves of it actually feel like.