Shoulder scan, shoulder injection, shoulder therapy. The shoulder was never the problem.

SHOULDER PAIN THAT STARTS IN THE NECK

You have had a shoulder scan, a shoulder injection and shoulder physical therapy. The pain is unchanged, because the shoulder was never the problem.

Why this happens so reliably

The C5 nerve root refers into the outer shoulder and upper arm convincingly enough that the patient points at their deltoid and everyone reasonably starts there. Meanwhile a subacromial or AC joint problem refers into territory people describe as their neck. The traffic runs both ways and it runs constantly.

A systematic review exists specifically on overlapping, masquerading and causative cervical spine and shoulder pathology, which is a fair indication of how often the two get confused by people who are paying attention.

The questions that separate them

• How far does it go? Shoulder pain stops around the deltoid insertion. Neck-referred pain frequently continues past the elbow, and anything reaching the hand is not the shoulder.
• What provokes it? Overhead reaching and lying on it are shoulder. Neck extension and rotation are cervical.
• Is there numbness? Shoulders do not produce numbness. A dermatomal strip of altered sensation is a root.
• Does moving the shoulder passively hurt? If somebody else moves your arm through range and it hurts, the shoulder is involved. If it does not, the pain is being referred there.
• Is there true weakness? Painful weakness is common in both. Weakness without pain points at a nerve or a cuff tear.

Frozen shoulder, which is neither

Adhesive capsulitis restricts passive external rotation, and that single finding separates it from almost everything else. It is frequently mislabelled as impingement and occasionally as a neck problem, and it is treated completely differently — the natural history is long, the treatment is directed at the capsule, and rotator cuff strengthening does not touch it.

It is worth naming here because it is a common answer in the group of people who have had shoulder treatment that did nothing, and because the examination finding that identifies it takes about fifteen seconds.

When both are true

Frequently. A degenerate cervical spine and a degenerate shoulder coexist in the same population, and a person can have a C5 radiculopathy generating the arm pain and subacromial pathology generating the overhead pain. Treating one and declaring failure when half the symptoms remain is the predictable outcome.

Working out the proportion matters. If most of the pain is cervical, shoulder surgery will disappoint. If most is subacromial, a cervical injection will disappoint. The examination is what apportions it, and it needs to cover both regions in the same appointment.

What we do about it

Examine both, every time, in the same visit. Where it is genuinely ambiguous, a diagnostic injection resolves it — a well-placed subacromial or AC joint block that abolishes the pain for a few hours has answered the question, and one that does nothing has answered it equally well.

And we do not route anybody to a shoulder surgeon on the strength of an MRI report showing a cuff tear, because asymptomatic cuff tears are extremely common with age and the tear on the film is not automatically the source of the pain.

The order we work in

Distribution first, because it costs nothing and narrows the list faster than any test. Then active and passive range, which separates a joint problem from a referred one. Then resisted testing to localize within the shoulder. Then the cervical provocation tests. Then, and only then, imaging read against all of it.

That sequence takes one appointment. The alternative — a shoulder service and a neck service seen six weeks apart, each doing half the examination — takes months and frequently still leaves the question open.

The scapula, which nobody mentions

Pain around the shoulder blade is one of the most reliable referral patterns from the lower cervical spine, and it is routinely investigated as a shoulder or a muscular problem. Interscapular pain with neck extension provocation is cervical until proven otherwise.

It also runs the other way: poor scapular control loads the AC joint and the subacromial space, so a shoulder blade that does not rotate properly produces shoulder symptoms without any structure being damaged.

What the injections tell us

A diagnostic block is the tie-breaker when the examination cannot separate the two, and its value lies as much in a negative as a positive. A well-placed subacromial injection that changes nothing has removed the subacromial space from the list, which is a real result and a cheap one.

The requirement is that it be well placed. Blind injections into small structures miss often enough that a negative result from one carries much less weight than people assume.

What a shoulder examination adds that a scan does not

Active range against passive range, which is the single most informative comparison in shoulder medicine and is not obtainable from an image. Resisted testing in specific planes. The painful arc. And direct palpation of the AC joint, which localizes better than anything else in the region.

None of that requires equipment. All of it is more useful for deciding what to treat than a report listing degenerative findings that most shoulders of that age also have.

If you have already had shoulder surgery

Persistent pain after a technically successful shoulder operation is a recognized situation, and a cervical contribution that was present all along is one of the explanations for it. The examination is the same either way, and having had the operation does not remove the question.

It does change the emphasis: post-operative stiffness, altered mechanics and deconditioning all become part of the picture, and separating those from a referred component is the work.

What people ask about shoulder and neck pain

My shoulder MRI showed a rotator cuff tear. Is that my pain?

Not necessarily. Cuff tears are found in large numbers of shoulders that do not hurt, and the frequency rises steeply with age. It earns the diagnosis when it matches the examination. Rotator cuff, or the neck?.

The shoulder injection did nothing. What does that mean?

If it was accurately placed, it is a useful negative result and it points the search back toward the neck. If it was placed blind, it means less than it appears to. Neck or shoulder?. It has already narrowed the field, which is worth more than it sounds.

Can a neck problem cause shoulder weakness?

Yes, and distinguishing that from a cuff tear is one of the main jobs of the examination. The pattern of which movements are weak is what separates them. How facet pain is confirmed.

Why did nobody check both?

Because the two are usually managed by different services, and each examines the region it owns. Doing both in one appointment is the whole point of this page. Neck or shoulder?.

Related reading

• Rotator cuff, or the neck?
• Subacromial bursa
• AC joint pain
• Neck or shoulder?
• Cervical radiculopathy
• Why your MRI does not explain it

A shoulder injection that changed nothing has already narrowed the field. That result is a finding, not a failure.