Most cervical problems get better without an operation, and most of the work is done by time, movement and sensible pain control rather than by anything dramatic. This page goes through each rung of the ladder, what the trial evidence behind it actually shows, and how long it is reasonable to give it before asking the question again.
The short version
More than 85%
Of acute cervical radiculopathy improves without specific treatment, most of it inside eight to twelve weeks.
6 weeks
Where physiotherapy and a collar helped most in trial data. By six months the treated and untreated groups matched.
Stay active
Gentle activity generally beats rest. Inactivity tends to make a stiff neck stiffer.
Last rung
Surgery is considered when the rungs below have not worked, or when the spinal cord is at risk.
THE LADDER
For almost every cervical condition, treatment climbs in the same order. Each rung is worth a fair go before the next one, with one exception: where the spinal cord is involved, the ladder is skipped.
FIRST
Doing the ordinary things you can tolerate, avoiding the positions that flare it, and letting the natural history do its work. This rung treats more people than all the others combined.
ALONGSIDE
Targeted exercise, a short course of appropriate pain relief and decent sleep positioning. The aim is to make the weeks tolerable while the problem settles, not to force it.
SOMETIMES
Where one nerve or joint is clearly the culprit and the pain is not settling, a carefully placed injection can calm it and can help confirm the source.
ONLY IF NEEDED
Considered for weakness that is worsening, pain that will not settle after a fair trial of everything above, or compression of the spinal cord.
Where you sit on this ladder depends on which condition you have. The conditions pages start from how each one feels, and do I need neck surgery covers what has to be true before an operation belongs in the conversation.
THE FIRST RUNG
It is an unsatisfying thing to be told when your arm is keeping you awake, but it is the single most important fact about cervical nerve pain.
8 in 10+
More than eight in ten people with acute cervical radiculopathy improve without any specific treatment, most of them within about eight to twelve weeks, and 83 per cent have regained satisfactory function by three years. Where a disc fragment is pressing on the nerve, the body often reabsorbs part of it over months, so follow-up scans frequently look better than the first one. Everything else on this page is aimed at making that period bearable and keeping you moving through it.
Staying gently active matters more than any particular exercise. Keeping the neck and shoulder girdle moving, continuing to work where you can, and avoiding the specific positions that flare the pain gives a better result than resting and waiting it out.
PUBLISHED RESEARCH
Non-surgical treatment for the neck is better studied than most people assume, and the findings are more modest than the marketing around them.
Faster, not different
Collar or physiotherapy against waiting, 205 patients
In a randomised trial of 205 people with cervical radiculopathy of less than a month, a semi-hard collar with rest and a course of physiotherapy with home exercises each produced a clinically meaningful extra reduction in arm pain at six weeks compared with a wait-and-see approach. By six months the median arm pain score was zero in all three groups, including the group that did nothing.
BMJ, 2009
Moderate quality
Adding traction to physiotherapy, 5 trials, 449 patients
A meta-analysis of five randomised trials found that adding mechanical traction to physiotherapy gave a significant reduction in pain in the short and middle term, on moderate-quality evidence, with smaller and mostly non-significant effects on function and disability. Useful as a supplement for some people, rather than a treatment in its own right.
Physical Therapy, 2018
How to read this
The pattern in both studies is the same: these treatments change the shape of the first six weeks rather than the eventual outcome. That is genuinely worth having when the pain is severe, and it is also a reason not to spend a year and a great deal of money chasing the perfect regimen.
It is also why a lack of improvement after a fair trial is informative. If nothing on the lower rungs has shifted it after a few months, that is a reason to look again at the imaging and the examination rather than to keep repeating the same thing.
MEDICATION
The purpose is to let you sleep, keep moving and get through the weeks. It is not to fix the disc, and no tablet does that. Prescribing is a matter for your GP or specialist.
01
Regular paracetamol taken properly is more useful than people expect, mostly because it lets the other measures work. It is the least interesting rung and often the most used.
02
A short course can take the edge off an acutely inflamed nerve root. They are not suitable for everyone, particularly with stomach, kidney or cardiovascular history, which is why they are worth discussing rather than assuming.
03
Nerve pain often responds poorly to ordinary painkillers, and agents aimed specifically at neuropathic pain are sometimes used. They take time to work and have their own side effects, so they are a considered choice rather than a first move.
04
Nerve pain is often worst at night, and a few nights of sleep changes how the whole problem feels. Strong opioids have a narrow role in the acute phase and a poor one beyond it.
INJECTIONS
A well-placed injection has two jobs: to settle a specific inflamed structure, and to tell you whether that structure is the one causing your symptoms. Both are useful. Neither is a cure.
01
Around the nerve root that is compressed, or into a specific facet joint. It is placed under imaging guidance, because the target is small and the structures nearby are not forgiving.
02
Results vary widely. In pooled reporting, around 24 per cent of patients had complete resolution of symptoms, about 40 per cent achieved at least 75 per cent pain relief, and roughly 32 per cent got no relief at all.
03
Often not long. The evidence supports pain relief over a period of weeks rather than months, which is why an injection is usually a way of buying time for the natural history rather than an end point.
04
Sometimes the diagnostic value is the point. If numbing one specific nerve root abolishes your arm pain for a few hours, that is strong evidence about which level is responsible, whatever the scan shows.
Injections are not risk free. Minor side effects such as a sore neck or a headache are reasonably common, and serious complications, while rare, are the reason the approach and the imaging guidance are chosen carefully. Whether an injection is appropriate for you is a decision for the doctor performing it.
In a randomised trial of 205 people, a semi-hard collar and a course of physiotherapy each reduced pain faster than waiting did. By six months all three groups, including the one that did nothing, had a median arm pain score of zero. That is genuinely worth having when the pain is severe, and it is also a reason not to spend a year chasing the perfect regimen.
LESS USEFUL THAN EXPECTED
None of these is dangerous in itself. They are on this page because they are common, they feel productive, and they generally are not.
A few days of taking it easy in a severe flare is sensible. Weeks of it is not. Necks stiffen quickly, and deconditioning makes the return to normal activity harder than the original problem.
A collar has a defined short-term role in the acute phase of nerve pain, which is what the trial evidence tested. Wearing one for months is a different thing, and it tends to weaken the muscles that were helping.
Imaging changes slowly and symptoms change quickly. A repeat scan is useful when something clinical has changed, and mostly unhelpful when it has not, because the findings that worried you the first time will still be there.
Cycling through practitioners in the first few weeks makes it very hard to tell what helped. Giving one sensible plan a fair run tells you far more, and the natural history is on your side while you do.
WHEN TO STOP WAITING
Pain that is severe but gradually improving, numbness that is unchanged, and a plan you have only been on for a few weeks. Improvement in cervical nerve pain is usually measured in weeks to a few months, and a bad week inside that period is not a failure of the plan.
About cervical radiculopathy ›Weakness that is measurably worsening, pain that has not shifted at all after a fair trial, or any of the cord signs listed below. The last of those is not a wait-and-see situation at all, and it is assessed on a different timescale from nerve pain.
About cervical myelopathy ›Seek urgent care
Sudden or rapidly worsening weakness, new clumsiness or unsteadiness affecting both hands, a decline in walking, or any change in bladder or bowel control can signal pressure on the spinal cord. Do not keep waiting: call 000 or go to your nearest emergency department. GPs with an urgent concern can phone the rooms directly on (07) 3870 3708.
COMMON QUESTIONS
There is no single number, but the natural history of acute cervical nerve pain is measured in weeks to a few months. A plan that has produced no change at all after a fair trial is worth reviewing, and worsening weakness changes that timeline immediately.
Good physiotherapy should not. In the randomised trial evidence it reduced pain faster than doing nothing. Treatment that consistently and sharply increases your arm symptoms is worth reporting back rather than pushing through.
A semi-hard collar has a short-term role in acute nerve pain and performed as well as physiotherapy in the trial data. It is a matter of weeks rather than months, because long-term use tends to weaken the muscles you will need later.
It can be, where one nerve or joint is clearly the source and the pain is not settling. Be clear about what you are buying: often several weeks of relief and useful diagnostic information, rather than a permanent fix.
Not exactly, but displaced disc material is frequently broken down and reabsorbed by the body over months, and follow-up scans often show a herniation has shrunk or disappeared. Symptoms usually improve well before the scan changes.
Usually yes, and doing so is generally better for the outcome than stopping. Modifying what you do, how long you sit and how your screen is set up is more useful than taking the time off entirely.
NEXT
The condition most of this page is about: what it is, which level does what, and how it usually behaves over time.
Read about radiculopathy ›What has to be true before surgery belongs in the conversation, and the signs that change the timing.
Work through it ›Age-related change in the neck, why it shows on almost everyone’s imaging, and when it actually matters.
Read about spondylosis ›This page is general information about non-surgical treatment of the cervical spine. It is not a substitute for an individual assessment, and it is not a recommendation to take or avoid any particular medication or procedure, which are decisions for the doctor treating you. Where figures are given, they come from the published research listed below and describe groups of people rather than individuals. More about Dr Tollesson.
References
Cervical collar or physiotherapy versus wait and see policy for recent onset cervical radiculopathy: randomised trial. BMJ, 2009, for the six-week and six-month comparison quoted above.
Cervical Radiculopathy: Effectiveness of Adding Traction to Physical Therapy, a systematic review and meta-analysis of randomised controlled trials. Physical Therapy, 2018, for the traction findings.
Cervical Radiculopathy, StatPearls, National Library of Medicine, for the natural history of acute cervical radiculopathy.
Cervical Epidural Injection, StatPearls, National Library of Medicine, for the reported rates of pain relief and the duration of benefit.
A consultation will tell you what is causing your symptoms, whether anything needs treating, and which of the rungs above is worth your time. A GP or specialist referral is needed for a Medicare rebate.