The nerves grow back. That is the deal, and nobody explains it beforehand.

CERVICAL RADIOFREQUENCY NEUROTOMY

This is the one interventional treatment for neck pain with a genuinely durable answer — and it only works if the right nerves were identified first, which is where most of these fail.

What it actually does

Each cervical facet joint is supplied by two medial branch nerves, small sensory nerves whose only job is to report from that joint. Radiofrequency neurotomy heats a short segment of those nerves to interrupt them, so the joint stops reporting pain.

It is not a treatment for the joint. The arthritis, if there is any, is unchanged. What changes is that the joint can no longer send the signal, which is why the benefit lasts as long as the nerve takes to regrow — typically somewhere between six and eighteen months, sometimes considerably longer.

Why the block comes first, without exception

The medial branches are the target only if the facet joints are the source. Nothing on an MRI establishes that. What establishes it is a diagnostic medial branch block — anesthetising the specific nerves and seeing whether the pain goes.

Single blocks carry a false-positive rate high enough to matter, which is why the guideline position favors confirming before ablating. Treating on a single equivocal block is how a procedure with good evidence acquires a reputation for not working.

What the evidence supports

Consensus practice guidelines from the multisociety working group on cervical facet joint interventions support radiofrequency neurotomy in appropriately selected patients — the phrase is load-bearing, and selection means confirmed blocks rather than an imaging finding.

More recent guidance has been more cautious about interventional spine procedures generally, and it is worth saying so rather than quoting only the supportive documents. What the cautious readings share is a criticism of selection, not of the technique. Done on confirmed facet pain it performs; done on unconfirmed axial neck pain it does not, and both findings sit in the same literature.

What the day involves

• Prone or seated, depending on the levels. Local anesthetic at the skin.
• Needle placement under fluoroscopic guidance onto the articular pillar at each target level. This is not a landmark procedure and is not done without imaging.
• Sensory and motor stimulation testing before any heat is applied — confirming the electrode is on the target nerve and not near a motor root.
• Lesioning for 60 to 90 seconds per site. Several sites per level, because the nerve is not a point.
• Home the same day. No sedation is required and none is used routinely.

What recovery honestly looks like

The first week is frequently worse. Neuritic soreness and a deep ache in the treated area are expected, not a complication, and people who were not warned about it conclude the procedure failed. Benefit typically declares itself between two and four weeks as that settles.

When the nerves regenerate the pain returns, and the procedure can be repeated. That is the honest shape of it: this is durable relief, not a cure, and anybody describing it as permanent is overselling.

Who it does not suit

• Anyone whose blocks were negative or equivocal — the target has not been confirmed.
• Radicular arm pain, which is a nerve root problem and a different diagnosis.
• Progressive neurological deficit or cord signs, which need a surgical opinion.
• Segmental instability, which no ablation addresses.

When the pain comes back, which it will

The nerves regenerate. That is not a complication, it is the mechanism, and the honest framing is that this buys a defined period of relief rather than a cure. Most people come back somewhere between six and eighteen months.

Repeat procedures generally work about as well as the first, which is the useful thing to know. What changes the answer is whether anything else has changed — a new radicular symptom, a new level, a deteriorating gait — because a repeat should never be booked without asking whether the original diagnosis still holds.

If it does not work at all

Three questions, in order. Were the blocks genuinely positive, or was one equivocal and treated as positive because everyone wanted to proceed? Were the correct levels targeted, given that referral patterns overlap? And is there a second pain generator — a disc, a root, a shoulder — that was always going to limit the result?

Repeating an identical procedure that did not work, without changing any of those inputs, is not a plan. It is the most common thing that happens next, and it is why people arrive here having had three.

Conventional and cooled radiofrequency

Conventional radiofrequency creates a small lesion around the active tip. Cooled radiofrequency circulates fluid through the electrode so a larger volume can be lesioned without charring, which gives a bigger target zone where anatomy is uncertain.

For cervical medial branches, where the target is small and the anatomy well described, conventional technique with careful placement and stimulation testing is what is used here. Larger is not automatically better next to a motor root.

What people ask about cervical ablation

How long does cervical radiofrequency ablation last?

Typically six to eighteen months, occasionally longer, determined by how quickly the medial branch nerves regenerate. It is repeatable when the pain returns. What a diagnostic block proves.

What is the success rate?

In properly selected patients — meaning confirmed diagnostic blocks — the majority get meaningful relief. In unselected patients the numbers fall sharply, which is the single biggest driver of the variation you will read about. Ablation or fusion?.

How painful is the procedure?

Uncomfortable rather than severe, with local anesthetic at each site. The week afterward is usually the harder part, and it is temporary. How facet pain is confirmed. Test stimulation first, then the lesion, and an ache for a few days afterward.

Will it weaken my neck?

The medial branches are sensory to the joint and also supply the small multifidus muscles at that segment. Some multifidus atrophy occurs. It has not been shown to produce meaningful functional weakness, and it is a real trade-off rather than a nothing. The block that settles it.

Related reading

• Diagnostic medial branch block
• Cervical facet syndrome
• Ablation compared with fusion
• Cervicogenic headache
• What happens at the first visit
• Neck pain doctor in St. Louis