Management of carpal tunnel syndrome

About carpal tunnel syndrome

The distal part of the median nerve passes through the carpal tunnel, an anatomical compartment bounded by the transverse carpal ligament (flexor retinaculum) and carpal bones. Carpal tunnel syndrome results from compression of the nerve in the tunnel at the point it passes under the transverse carpal ligament at the wrist.1^,3 The clinical features are mainly explained by the functions of the branch of the nerve that passes through the carpal tunnel. These include carrying sensory impulses from the skin of the palmar aspects of the thumb, index, middle and radial half of the ring finger, and the terminal parts of their dorsal aspects (see Fig 1).3 The nerve also supplies the following muscles: abductor pollicis brevis, opponens pollicis, part of the flexor pollicis brevis and the two lateral lumbrical muscles.3^,4

Clinical picture

Carpal tunnel syndrome usually presents with numbness, tingling, and burning sensations, and a dull ache in the hand and fingers. These symptoms are usually restricted to the thumb, index, middle and ring fingers, but may affect the little finger and/or the palm as well.3 Symptoms usually occur at night, often waking the patient from sleep, but can be relieved within a few minutes by shaking the hand.3 Pain sometimes radiates up the forearm as far as the elbow, and even as high as the shoulder or root of the neck.3 Other, less common, symptoms include weakness or clumsiness of the hand, and dry skin, swelling or colour changes in the hand.2 Symptoms may recur during the day when the hands are used for carrying things, and for activities that involve holding them up, such as reading a newspaper, driving or using a keyboard.3

In one follow-up study involving 196 untreated patients (a total of 274 affected hands), about 30% had symptomatic improvement 10–15 months after diagnosis, while 45% had stayed the same and 21% had got worse.5 However, these figures may over-represent improvement, as they are derived from a selected sample of patients (i.e. one that excluded those in whom immediate surgery was indicated). It is also unclear from published literature whether the severity and character of symptoms are related to the structural and physiological condition of the median nerve, or whether the duration of symptoms correlates with the amount of nerve injury or predicts treatment outcomes.6^,7 In many people, intermittent symptoms persist for years without signs of median nerve damage, while others have progressive, and eventually permanent, neurological features due to nerve damage. In advanced cases, there is development of weakness of the thenar muscles, which particularly affects abduction of the thumb and is associated with atrophy of the lateral aspect of the thenar eminence. Sensory loss may appear over the tips of the fingers innervated by the median nerve.4 Rarely, median nerve deficit develops in the absence of symptoms, such that motor and sensory signs may be discovered incidentally, particularly in older people.3^,4

Epidemiology and risk factors

At any one time, at least 1 in 50 people have carpal tunnel syndrome.1 It is particularly common in women,1 with one study in the UK indicating an incidence of 139.4 cases per 100,000 women per year and 67.2 cases per 100,000 men per year.8 One UK study in women found genetic predisposition to be the single strongest factor in predicting the development of the syndrome.9 Other predisposing factors include diabetes mellitus, pregnancy, obesity, myxoedema, acromegaly, and infiltration of the flexor retinaculum in primary and hereditary amyloidosis.3 Carpal tunnel syndrome may develop as a consequence of wrist joint involvement in rheumatoid arthritis or osteoarthritis, or deformity related to an old fracture.3 Whether overuse of the hands is a cause of the syndrome is not clear, although most patients report that symptoms are aggravated by heavy use of the hands.2 Carpal tunnel syndrome is an independent risk factor for diabetes,10 so it is worth screening for diabetes in patients presenting with the syndrome.2

Diagnosis

Clinical assessment

Around 95% of patients with carpal tunnel syndrome report burning, numbness and tingling in the hand and a similar proportion have such symptoms at night.3 About 55–65% of patients first present with bilateral carpal tunnel syndrome.2 Percussion over the carpal tunnel or pressing over the median nerve may elicit paraesthesiae in a median nerve distribution (Tinel's sign); or symptoms may be provoked by sustained flexion or hyperextension of the wrist (Phalen's and reverse Phalen's signs, respectively). However, neither of these tests is very specific or sensitive (estimated sensitivity: Tinel's sign 28–73%, Phalen's test 46–80%; estimated specificity 44–95% and 51–91%, respectively).6 Differential diagnosis for carpal tunnel syndrome includes cervical radiculopathy; cervical cord disease (e.g. multiple sclerosis); thoracic outlet syndrome; ulnar neuropathy; Raynaud's phenomenon; osteoarthritis of the metacarpophalangeal joint of the thumb; and tendonitis. Also, carpal tunnel syndrome may occur as part of a generalised peripheral neuropathy.

Investigations

The diagnostic investigation most commonly used for suspected carpal tunnel syndrome is a nerve conduction study, which typically shows delayed median nerve sensory action potential, and then delayed and smaller median nerve compound muscle action potential, as nerve conduction becomes slower and more dispersed and, in severe cases, as more nerve fibres undergo axonal degeneration. However, nerve conduction studies are only appropriate if there is diagnostic doubt following clinical assessment or before surgery.2^,6 Nerve conduction studies can also help to ascertain whether the carpal tunnel syndrome heralds a more generalised neuropathy, and as a baseline for working out what has gone wrong after an unsuccessful operation for the condition. However, some patients meeting classic clinical criteria for carpal tunnel syndrome have normal nerve conduction studies, while others with only possible symptoms have delayed median sensory action potential.3^,6^,11

Research has found magnetic resonance imaging of the median nerve in the carpal tunnel to be no more reliable than nerve conduction studies in the diagnosis of carpal tunnel syndrome, and it is more expensive.2^,6 There is some evidence that the specificity and sensitivity of high resolution ultrasonography is comparable with that of nerve conduction studies, but there is comparatively less experience of its use in the diagnosis of carpal tunnel syndrome, so it is probably best viewed as a complementary investigation.2^,12^,13 We know of no adequate studies of the use of computer tomography in the diagnosis of this syndrome. None of these imaging techniques is in routine use in UK practice for the diagnosis of carpal tunnel syndrome.

Treatment

The standard treatment options for carpal tunnel syndrome are splinting, local corticosteroid injections and surgery. Other suggested treatments include diuretics, NSAIDs, vitamin B₆, acupuncture, manipulation, mobilisation, exercises, magnet therapy, therapeutic ultrasound and yoga.2^,14

Splinting

The simplest treatment for carpal tunnel syndrome is provision of a splint that will hold the wrist in a neutral position without applying direct compression over the carpal tunnel;14 the aim is to provide mechanical respite for the median nerve. Splints are available on the NHS, but local provision arrangements vary (e.g. via physiotherapists or orthotics departments). The basic cost of each splint to the NHS is around £3.70. They are also available to buy via the internet from around £15 each.

Clinical evidence

In one controlled trial, 83 patients with carpal tunnel syndrome were randomised to nocturnal splints or no splinting for 4 weeks.14^,15 The primary outcome measure was the score on the Boston Carpal Tunnel Questionnaire. This is a patient-orientated validated measurement that assesses symptoms with an 11-item scale plus functional status with an 8-item scale; the overall score is calculated as the mean of the responses of the individual items. There was a greater overall improvement after 4 weeks in the splint group than in the non-splint group (overall score fell from 4.64 to 3.02 in the splint group vs. from 4.79 to 4.64 in the “no splint” group, p<0.001; with the fall in the splint group being considered “clinically important”).16 A total of 38 out of 40 participants wore the splint for most or all of the nights during the study. Splinting has been shown to still be effective at 12 weeks in reducing symptoms and improving functional status,17 but there is a lack of evidence on longer-term use. It is unclear whether it is more effective to wear a splint at night only or full-time.14 However, some patients find the splints too cumbersome to wear during the day.2

Local corticosteroid injection

If splinting does not work within a few weeks or the symptoms become unbearable at an earlier stage, then one option is to give a local corticosteroid injection (e.g. betamethasone, dexamethasone)18 into or close to the carpal tunnel.2^,3^,19 The procedure needs to be undertaken using a sterile technique. The best location for injection is not clear, but care must be taken to avoid injecting the nerve itself, which is painful and might cause nerve damage.

Clinical evidence

One systematic review pooled data from two randomised controlled trials involving a total of 141 participants with carpal tunnel syndrome; it found that clinical improvement at 1 month or less was more likely following local corticosteroid injection than following placebo injection (relative risk [RR] 2.58, 95% CI 1.72 to 3.87).20 The review also found that symptoms improved more with local corticosteroid injections than with either systemic (intramuscular) injection at 1 month or oral corticosteroids at 12 weeks. However, it also found that local corticosteroid injections did not produce greater improvement in symptoms after 2 or 8 weeks than did oral anti-inflammatory medication plus splinting, and that two injections did not provide more clinical improvement than one.

The optimal corticosteroid, dose and site of injection are not known. One randomised controlled trial involving 123 patients with suspected carpal tunnel syndrome (and blinded to treatment allocation) found no significant difference in improvement in symptoms between injections of hydrocortisone 25 mg, hydrocortisone 100 mg, or triamcinolone 20 mg after 6 weeks or at 6 months of follow-up.21 (The primary outcome measure was a subjective change in symptoms on a five-point scale from “much worse” to “much better.”) However, 66% of patients treated with hydrocortisone 25 mg and 63% of those given hydrocortisone 100 mg reported their symptoms to be “better” or “much better” compared with 5% in the control group (no injection; p<0.05 and <0.02, respectively), at 6 weeks. In another randomised controlled study (blinding status not given) involving 42 patients with symptoms of carpal tunnel syndrome, there was no significant difference in the proportions of patients with a “favourable response” (assessed by patients' reports of symptoms) at any time up to 12 weeks following injection of methylprednisolone 35 mg proximal to the wrist skin crease or methylprednisolone 12 mg 2–3 cm distal to the skin crease.19 The authors stated that their rationale for using a smaller dose of corticosteroid for the distal injection was that the space available at this area is more limited than proximal to the transverse carpal ligament.

The likelihood of harm from local corticosteroid injection appears very small; there have been at least eight case reports of nerve or tendon injury, but no serious complications in published trials in which over 3,000 injections were given.2

Surgery

Patients with carpal tunnel syndrome who have severe or persistent median nerve deficit, or symptoms that do not respond to splinting or an injection, are typically referred for consideration of carpal tunnel decompression surgery. Such treatment involves division of the transverse carpal ligament (either via traditional open surgery or endoscopically) to increase the space in the carpal tunnel.2^,22 It is usually performed under local anaesthesia as day case surgery.2^,22

Clinical evidence

In published series of carpal tunnel release operations, the rates of success (defined as no further treatment needed after surgery) have varied from 27% to 100%.7 An NHS survey of 4,000 patients found that, about 2 years after surgery, 75% considered that the operation had been an “unqualified success” and 8% thought they were worse off.7 Similarly, data pooled from 209 surgical series involving a total of 32,936 procedures worldwide also found an average success rate of 75%.7

A systematic review including four randomised controlled trials involving a total of 317 participants with carpal tunnel syndrome compared surgery with non-surgical treatment (e.g. splinting [two trials]; local corticosteroid injection [two trials]).23 Three of the trials (295 participants in total) found a clinical improvement after 3 months (the primary outcome measure of the review) was more likely with surgery (relative risk [RR] 1.23, 95% CI 1.04 to 1.46). However, complications were more likely with surgery (two trials, 226 participants in total; RR 1.38, 95% CI 1.08 to 1.76). The reviewers concluded that surgical treatment for carpal tunnel syndrome was significantly better than splinting for relieving symptoms, but that further research was needed to discover whether it was better than local corticosteroid injection. Another review, which assessed variations of surgical techniques in carpal tunnel syndrome, included 33 studies comparing endoscopic and open approaches: there was no significant difference in relief of symptoms in the short or long term.22 However, in three studies (with a total of 294 participants), endoscopic carpal tunnel release resulted in earlier return to work or activities of daily living than standard open surgery (weighted mean difference -6 days, 95% CI -9 to -3 days).

A combination of symptoms and signs, and positive nerve conduction test results seems to increase the likelihood of a favourable outcome from surgery.6 In a postal questionnaire follow-up of 998 patients, around 75% who had had moderate pre-operative neurophysiological abnormalities were cured or much better after surgery (i.e. a success).24 By contrast, the outcome was not so favourable in patients with normal pre-operative nerve conduction studies (51% success) or those with extremely severe nerve conduction abnormalities suggesting axonal degeneration (47% success).

Surgery may be complicated by painful scarring, infection or complex regional pain syndrome.2 Failure of the treatment has been associated with advanced disease, incorrect diagnosis or incomplete section of the transverse carpal ligament.1^,2 A literature review estimated the incidence of structural damage to nerves, arteries or tendons was 0.49% with open carpal tunnel release based on 22,327 casess and 0.19% with endoscopic carpal tunnel release based on 5,669 cases (p<0.005).25

Other treatments

One systematic review including 21 published randomised controlled trials involving a total of 884 people with carpal tunnel syndrome who had not previously undergone surgical release found that most other interventions (such as NSAIDs, diuretics, vitamin B₆, magnet therapy, acupuncture, exercise, and chiropractic manipulation of the wrist) were no better than placebo in terms of clinical outcome.14 Possible benefit from therapeutic ultrasound was reported in one trial but not in two others, and from carpal bone mobilisation in one trial and from yoga in another trial.14

When to refer

The diagnosis and initial management of patients with carpal tunnel syndrome, including the provision of splints and local corticosteroid injections, can usually be done in primary care, particularly if impairment of median nerve function is mild or moderate. Patients with severe impairment, symptoms that do not respond to splinting or an injection, or those with persistent median nerve deficit should be referred for nerve conduction studies to confirm diagnosis and for consideration of decompression surgery. Those with recurrent or persistent symptoms after surgery should be referred back to their surgeon for further assessment and repeat nerve conduction tests to elicit whether the features are due to incomplete decompression.

Conclusion

Carpal tunnel syndrome results from compression of the median nerve in the wrist and has a prevalence of around 2%. About 30% of patients have transient symptoms that recover completely after a few weeks or months without treatment, while some patients have progressive and eventually permanent median nerve deficit. The diagnosis can usually be made clinically, but should be confirmed with nerve conduction tests if diagnosis is in doubt or surgery is being considered. In mild cases, initial treatment should include the use of wrist splints. If this fails, then a local corticosteroid injection is appropriate, but it is not clear what the best regimen is to use. Patients with severe or persistent impairment, or symptoms that do not respond to splinting or an injection, should be referred for consideration of decompression surgery. There is no evidence that endoscopic surgery is more effective than standard open operation; however, it may permit earlier return to work, and may be associated with fewer complications.