Neck Pain Months After a Car Accident: Where It Often Comes From

What this video covers

• What cadaveric and biomechanical models show happens to the cervical facet joint capsules during a rear-end collision
• Why most whiplash resolves on its own, and the point at which that stops being the right answer
• Why a normal X-ray or MRI does not rule out a real, identifiable pain source
• How diagnostic medial branch blocks confirm the joint before anything is treated, and why a negative block means no procedure
• What the sham-controlled evidence actually shows for cervical radiofrequency neurotomy, how long relief lasts, and where the evidence is still weak
• Why exercise and early movement are supported for whiplash, while ketogenic nutrition, pro-resolving mediators, and red or near-infrared light therapy remain unproven for this condition
• MEDICAL DISCLAIMER: This content is for educational purposes only and is not medical advice. It does not substitute for professional diagnosis or treatment. Always consult a licensed healthcare provider regarding your condition. Viewing this video does not establish a doctor-patient relationship.

You were stopped in traffic when the car behind you did not stop. They X-rayed your neck, told you nothing was broken, and sent you home. That was four months ago. You still cannot check your blind spot without turning your whole torso. Your MRI was unremarkable. And somewhere between the third visit and the fourth, the language in your chart changed — from injury to complaint, from patient to claimant.

This article explains what the biomechanical models show happens to the cervical spine in the first fraction of a second after a rear impact, why that can produce pain that outlasts every normal image, what a normal MRI does and does not rule out, and how a specific pain source is tested rather than guessed at. It is not an argument for a procedure. Most whiplash resolves — six to twelve weeks of analgesia, early movement, graded exercise, and return to normal activity. Inside that window no procedure is indicated, and a clinic offering you one is selling you something.

What happens to your neck in the first 150 milliseconds

Between fifty and seventy-five milliseconds after impact — before your neck muscles can generate useful protective force — the seat drives your torso forward while your head stays behind. The cervical spine briefly assumes an S-shape: the lower segments extend while the upper segments are still flexed. Only past about a hundred and fifty milliseconds does the neck swing into simple extension, the motion most people picture when they hear the word whiplash.

In those models it is the early S-shape, not the later swing, that loads the paired zygapophysial joints — the facet joints. These are the small paired joints at the back of each spinal level, roughly fingernail-sized, that let your neck rotate and extend. Each one is wrapped in a capsule.

Here the evidence has to be labeled honestly. Cadaveric and computational models show capsular strain in that S-curve window rising above the capsule’s physiologic range — enough for what the biomechanics literature calls sub-failure injury, at collision speeds people survive without breaking a bone. Sub-failure is the precise word: loaded past the point of damage, without tearing through. That mechanism is modeled, not measured in you. The dorsal horn sensitization often invoked alongside it is rodent work. Laboratory inference should not be dressed up as bedside proof, and it is not being dressed up here.

Why the symptom pattern points at the joint

Whether that modeled capsular loading is what generates any particular person’s pain has not been demonstrated in living patients, and no scan settles it. What can be described is the pattern that makes the facet joint worth testing: pain that sits deep and axial in the neck rather than shooting down the arm, worse with extension and rotation — which is why the blind-spot check is the movement that gives you away — and tender to press over the articular pillars, the bony column you can feel a couple of finger-widths to either side of the midline at the back of your neck.

That pattern is a reason to test, not proof of a mechanism. And none of it has a shadow on film. A joint capsule generating pain traffic is not a structural finding. It is a physiological one, and physiological questions need physiological tests.

What has been measured, in living patients, is how often this joint turns out to be the source. In 1996, in Spine, Lord and colleagues studied sixty-eight consecutive patients referred with neck pain persisting beyond three months after whiplash, using comparative and placebo-controlled anesthetic blocks. Sixty percent had cervical facet joint pain (95% confidence interval 46 to 73 percent). That figure deserves its boundaries stated plainly: it is the yield in a referred population that had already failed to recover, one in which headache-dominant patients were screened first at the C2-3 level. It is not the rate among everyone in a collision, and it is not a pure axial-neck-pain number.

What the standard approach misses

Standard early care for whiplash is largely right. Analgesia, early movement, and graded exercise are what most people need, and most people get better on them. The failure is not in the first six weeks. It is at month four, when the assembly line offers two exits and neither one contains a diagnosis.

Behind the first door: the images are normal, therefore the injury healed, therefore what remains is stress, litigation, or catastrophizing. Behind the second: serial injections chosen by body region rather than by diagnosis — a trigger point here, a steroid there, because the neck is where the needle goes.

Both doors share the identical defect. Nobody made a diagnosis.

A normal MRI does not tell you your neck is fine. It tells you there is no tumor, no cord compression, no disc on a nerve root, no gross ligamentous disruption. It is also not the study that excludes a non-displaced fracture — that is a CT. If nobody ordered one, that should be said out loud.

One more thing you are owed in plain language: no corticosteroid preparation is FDA-labeled for epidural or intraspinal use anywhere in the spine. That use is off-label whoever performs it, and you should hear that sentence before a needle goes in, not after. (Corticosteroid is not used at all in radiofrequency neurotomy, described below.) This is educational information only. Do not start, stop, or change any medication without consulting your physician.

What this means for evaluation

First, close the differential

Two things come off the list before anything else, because they are what hurt people.

• Cervical artery dissection. Sudden severe neck or back-of-head pain after a collision, with dizziness, visual change, drooping, or slurred speech, is a dissection until a CT or MR angiogram says otherwise. That is an emergency department, not a pain clinic.
• Concussion. Ongoing headache, fogginess, light sensitivity, or trouble holding a thought is a brain injury, managed on its own track. No needle in your neck addresses it.

Then the rest of the list: radiating arm pain with weakness or reflex loss is radiculopathy. Gait change, hand clumsiness, and hyperreflexia suggest myelopathy, which is urgent. Occult fracture and craniocervical instability must be excluded first.

When what remains is axial neck pain — worse with extension and rotation, tender over the articular pillars, no neurologic deficit, past three months — the facet joint becomes a testable hypothesis. That is a different thing from a diagnosis, and the difference is the whole point.

Then test it, with blocks that can fail

From C3-4 downward, each cervical facet joint draws its nerve supply from two medial branches, one above and one below, wrapping the waist of the articular pillar. Under live fluoroscopy those nerves are reachable with a fraction of a milliliter of local anesthetic. If that joint is the generator, your pain stops for the pharmacologic life of the drug and returns when it wears off.

C2-3 is the exception. It is supplied by the third occipital nerve, which crosses the dorsal face of the joint itself rather than a pillar waist, so it is blocked and lesioned differently and carries its own risk of temporary unsteadiness.

One positive block is not enough — single blocks carry an unacceptable false-positive rate. The standard is dual comparative blocks: a short-acting anesthetic on one visit, a long-acting one on another, with your relief tracking the agent. Longer relief with the longer-acting drug is a concordant result.

And the sentence that costs a practice money: a negative block means no procedure. If the joint does not answer, we do not burn it. If you want the procedural detail — what the visits look like, what is and is not done at each step — it is laid out on the car accident injuries treatment page.

Only then, neurotomy — and only with the numbers on the table

When the blocks are concordant and positive, radiofrequency medial branch neurotomy becomes an option. Under fluoroscopic guidance the electrode is laid parallel to each medial branch — parallel, because the lesion forms along the shaft of the active tip, not across it — and the tissue is heated to approximately eighty degrees Celsius. That denatures the nerve carrying the joint’s pain traffic. It is symptomatic denervation, not repair. Anyone calling it healing is not describing this procedure.

The evidence, and its limits:

• Effect size (established, small trial). In the New England Journal of Medicine, also in 1996, Lord randomized twenty-four patients with axial post-collision facet pain to real neurotomy or a true sham — identical procedure, current never switched on, patient and assessor blinded. Median time for pain to return to half of baseline: 263 days treated versus 8 days sham (P=0.04); seven of twelve versus one of twelve were pain-free at twenty-seven weeks. Twenty-four patients is a small trial, and it should be read as one.
• Durability and grade (established, with a stated conflict). In 2016, in Pain Medicine, Engel and the Standards Division of the International Spine Intervention Society analyzed the published data: the majority pain-free at six months, more than a third at one year, number needed to treat of two for complete relief at six months. Those are pooled figures from a systematic review of what has been published — not the result of a second sham-controlled trial. The authors graded effectiveness high quality, and only conditionally: the grade holds where the procedure meets their own society’s technical standard. They graded the evidence on risks low quality. Both grades are reported here because they published both, and it is worth knowing that the body issuing the standard is the body that graded compliance with it.
• Scope (contested where it is stretched). The controlled evidence above was generated in axial neck pain, not headache. It should not be stretched there.

A number needed to treat of two means roughly half of correctly selected patients do not get complete relief. Individual results vary.

The material risks, stated flatly: numbness in the treated territory; post-neurotomy neuritis that can burn for weeks; transient unsteadiness after third occipital neurotomy; deep paraspinal weakness; new or increased pain; bleeding; infection; and rarely nerve injury. The evidence on those risks is the part the systematic review graded low quality, which means the honest answer about how often they happen is that it is not precisely known.

It is also not permanent. Peripheral nerves regenerate. Benefit typically ends between six and eighteen months, and repeat treatment is expected. If nobody told you that, you were not consented — you were recruited. Movement and deep neck flexor retraining are what hold the result.

About the adjuncts you may have been offered

You may have been offered a ketogenic protocol, an omega-3 pro-resolving mediator, or red-light therapy for this neck. For whiplash specifically, what stands behind all three is cell-culture and rodent work. No controlled trial has shown that any of them shortens a whiplash-associated disorder or improves a facet-mediated neck. They are adjuncts under study — investigational for this condition, not treatment for what happened in that collision.

What a real course can look like

The following is a composite — a picture assembled from many patients with this condition, not one person’s chart. It is an illustration, not data.

Five months after a rear-end collision. Axial right-sided neck pain, worse with extension and rotation, no arm symptoms, normal exam, unremarkable MRI. Comparative blocks followed: a short-acting agent gave a little over an hour of near-complete relief; weeks later, a long-acting agent gave most of a day. Longer relief with the longer-acting drug. Concordant. Only then was neurotomy discussed.

The hard parts are the ones nobody advertises. Authorization took weeks. The first fortnight afterward was worse, not better — burning, patchy numbness, neuritis, and a message wondering whether the nerve had been damaged for good. Relief settled in during weeks three and four. Around ten months the ache returned as the nerves regrew, exactly as predicted.

That is a good outcome. It is not everyone’s outcome. Individual results vary.

Frequently asked questions

Key takeaways

• In the first 50-75 milliseconds of a rear impact the cervical spine forms a transient S-curve that loads the facet joint capsules — a mechanism shown in cadaveric and computational models, not measured in living patients.
• In patients referred with neck pain lasting beyond three months after whiplash, 60% (95% CI 46-73%) had cervical facet joint pain on comparative and placebo-controlled blocks; that is a referral-population figure, not a rate for everyone in a collision.
• A normal MRI rules out several serious structural problems but cannot confirm or exclude a facet joint as the pain generator; only a diagnostic block can, and a negative block means no procedure.
• The sham-controlled evidence cited here comes from a twenty-four-patient trial conducted in axial neck pain, not headache, and should not be stretched to headache; the systematic review that grades effectiveness high quality grades the risk evidence low quality.
• Roughly half of correctly selected patients do not get complete relief, benefit typically ends in six to eighteen months as the nerves regrow, and repeat treatment is expected.
• If your neck pain has outlasted three months since your collision, the next step is not another injection — it is a diagnosis.

Medically reviewed by Gurpreet Singh Padda, MD, MBA, MHP — Board Certified in Anesthesiology, Pain Medicine, Interventional Pain Management, Addiction Medicine, and Obesity Medicine. Last reviewed July 2026.

This article is educational and is not a substitute for evaluation, diagnosis, or treatment by a physician. Individual results vary. Do not start, stop, or change any medication without consulting your physician. To be evaluated, call (314) 481-5000 or text (314) 886-5902 — bring your imaging and your timeline.

References

1. Lord SM, Barnsley L, Wallis BJ, McDonald GJ, Bogduk N. Percutaneous radio-frequency neurotomy for chronic cervical zygapophyseal-joint pain. N Engl J Med. 1996;335:1721-6. PMID 8929263. PubMed
2. Lord SM, Barnsley L, Wallis BJ, Bogduk N. Chronic cervical zygapophysial joint pain after whiplash. A placebo-controlled prevalence study. Spine (Phila Pa 1976). 1996;21:1737-44. PMID 8855458. PubMed
3. Engel A, Rappard G, King W, Kennedy DJ; Standards Division of the International Spine Intervention Society. The Effectiveness and Risks of Fluoroscopically-Guided Cervical Medial Branch Thermal Radiofrequency Neurotomy: A Systematic Review with Comprehensive Analysis of the Published Data. Pain Med. 2016;17:658-69. PMID 26359589. PubMed

Dr. Gurpreet Singh Padda, MD, MBA, MHP