Carpal Tunnel Syndrome

Carpal tunnel syndrome (CTS) is compression of the median nerve beneath the transverse carpal ligament at the wrist, and it is the most common peripheral nerve compression neuropathy. It affects 3-6% of the general population and accounts for approximately 90% of all entrapment neuropathies.

Who presents. Women outnumber men 3:1, attributed to hormonal factors and a smaller carpal tunnel, and incidence peaks at 40-60 years. Both hands are involved in 50-60% of cases, and the dominant hand is typically the more symptomatic. Repetitive wrist flexion and vibration exposure at work raise the risk.

What crowds the tunnel. Greater than 50% of cases are idiopathic, with no identifiable cause; these are the associations to ask about.

• Diabetes mellitus, the commonest systemic association - 10% prevalence in diabetics against 3% in the general population
• Pregnancy - 2-7% incidence, usually resolving post-partum
• Other oedematous states - congestive cardiac failure, renal failure
• Hypothyroidism - myxoedema deposition in the tunnel; amyloid and acromegaly belong on the same list
• Rheumatoid arthritis - synovial proliferation; also osteoarthritis
• Obesity - BMI greater than 30 increases the risk two-fold
• Masses in the tunnel - ganglion, lipoma, anomalous muscles
• Previous wrist fracture - distal radius fractures, especially malunion

The canal. A fibro-osseous tunnel at the wrist: the floor is the concave arch of the carpal bones, and the roof is the transverse carpal ligament (flexor retinaculum), running from the scaphoid tubercle and trapezium radially to the pisiform and hook of hamate ulnarly. The ligament is approximately 2.5-3cm long and 2-3mm thick, and it thickens with age and loading.

• Proximal row: scaphoid (radial), lunate, triquetrum, pisiform (ulnar)
• Distal row: trapezium (radial), trapezoid, capitate, hamate (ulnar)

Ten structures pass through it. Nine flexor tendons and one nerve, in layers: the four flexor digitorum superficialis (FDS) tendons to the index, middle, ring and little fingers superficially, the four flexor digitorum profundus (FDP) tendons deep to them, the flexor pollicis longus (FPL) tendon radial to the FDP tendons, and the median nerve.

The course of the nerve. It enters the forearm between the two heads of pronator teres, runs deep to FDS, and gives off the anterior interosseous nerve (pure motor) in the proximal forearm. It then passes into the carpal tunnel deep to the transverse carpal ligament, and beyond the tunnel gives off the recurrent motor branch to the thenar muscles before dividing into the digital sensory branches.

The palmar cutaneous branch. It arises 5-6cm proximal to the wrist crease and passes superficial to the transverse carpal ligament to supply sensation over the thenar eminence and central palm. Because it never enters the tunnel, palmar sensation is spared in true CTS, and that is how the level of the lesion is established.

The recurrent motor branch. Its relationship to the ligament is the variation that matters on the operating table.

• Extraligamentous - 50%. Exits distal to the ligament; the safest arrangement.
• Subligamentous - 30%. Travels under the ligament, then exits.
• Transligamentous - 20%. Pierces the ligament, and is the variant at risk during release.

View larger

View larger

View larger

View larger

View larger

Pressure first. Normal intracarpal pressure is 2-10 mmHg. In CTS it runs 30-110 mmHg, and 30 mmHg is the threshold at which symptoms appear; moving the wrist into flexion or extension raises it further. From there the sequence is predictable.

1. Venous congestion. Raised pressure obstructs venous outflow, oedema collects within the epineurium, and the oedema raises the pressure again - a vicious cycle.
2. Ischaemia and demyelination. Chronic ischaemia produces focal demyelination. This stage is reversible with decompression, and it is the one that gives sensory symptoms and prolonged latencies on nerve conduction studies.
3. Axonal damage. Severe chronic compression causes axonal loss with Wallerian degeneration. It is irreversible, it is what produces thenar atrophy, and recovery is limited even after decompression.

The history. Nocturnal paraesthesia - waking at night with numbness - is the classic story and occurs in 75% of patients, and the symptoms sit in the thumb, index, middle and radial half of the ring finger with palmar sensation preserved. Shaking or flicking the hand to relieve the symptoms is the flick sign, a useful question to ask rather than a pathognomonic finding. Progression runs sensory to motor, the motor complaints being weakness, clumsiness and dropping objects.

What makes it worse. Sustained wrist flexion (driving, reading, phone use), repetitive gripping or pinching, vibration exposure and cold weather. Pain radiates to the forearm or shoulder in 25%, and patients describe difficulty with buttoning and writing and weakness of pinch grip.

Inspection. Compare the thenar bulk of both hands palms-up: APB wasting shows as a depression over the thenar eminence. Trophic skin changes appear in severe cases.

Sensory testing. Light touch over the radial three and a half digits palmarly, compared against the little finger as an ulnar control, with Semmes-Weinstein monofilaments if objective testing is wanted. Two-point discrimination greater than 6mm is abnormal (normal 2-5mm), but it is a late finding reflecting large-fibre loss, and the JAMA systematic review found two-point testing of little diagnostic value - a normal result does not exclude CTS.

Motor testing. The abductor pollicis brevis is the muscle to test. With the hand palm-up on the table, ask the patient to point the thumb at the ceiling against resistance, palpate the muscle as it contracts, and grade it 0-5. Add opponens pollicis (thumb tip to little finger tip) and pinch strength, which is weak in advanced disease.

Provocative tests. Each sets out to reproduce the patient's symptoms in the median distribution; they differ in how reliably they do it.

Where each of those differentials is worked up in full. The distinction that carries the most weight in an exam is the level of the lesion, and the palmar cutaneous branch is the anatomical reason it works: a lesion proximal to the tunnel takes the palm with it. C6/C7 cervical radiculopathy and pronator (proximal median) syndrome both sit proximal to that branch and both spare nothing; thoracic outlet syndrome is more often ulnar-sided and positional. Cubital tunnel syndrome is the mirror-image compression at the other side of the hand and is the neuropathy most often confused with CTS when the patient reports "the whole hand". Where a release has failed, or symptoms outrun the territory, the double-crush pearl below and nerve injury and regeneration explain why a technically complete decompression can still disappoint.

Nerve conduction studies are the gold standard for confirming the diagnosis and grading severity. Request them when the presentation is atypical, when symptoms are bilateral and severe, to confirm the diagnosis before surgery, in a young patient (under 40), for medico-legal cases, and to put a number on severity and prognosis.

How it is done. Sensory studies are performed orthodromically (digit to wrist) or antidromically (wrist to digit), stimulating at the wrist and recording from the index or middle finger, measuring distal sensory latency and amplitude - normal is less than 3.5ms. The motor study stimulates the median nerve at the wrist and records from APB, with a normal distal motor latency under 4.5ms, and the ulnar nerve to ADM serves as the internal control.

What makes it diagnostic. Beyond the absolute latencies, the comparisons matter: a median-ulnar sensory latency difference greater than 0.5ms, or a motor difference greater than 1.0ms. Amplitudes fall as disease advances, a reduced motor amplitude indicating axonal loss. The study may be normal in early mild disease, so compare with the contralateral side and the ulnar nerve before calling it negative.

Ultrasound is used increasingly, and it answers structural questions electrophysiology cannot: it is non-invasive with no electrical stimulation, shows the nerve in real time, identifies structural causes such as a ganglion or a persistent median artery, and can guide an injection. It is operator-dependent, does not assess function, and cannot quantify severity the way nerve conduction studies do.

• Cross-sectional area greater than 10-12mm² at the tunnel inlet
• Flattening ratio - anterior-posterior against medial-lateral diameter
• Bowing of the retinaculum - displacement greater than 2mm
• Dynamic assessment - nerve movement with finger flexion

View larger

View larger

View larger

View larger

MRI is not routine. Reserve it for a suspected mass lesion, an atypical presentation, failed surgery, and research. It shows T2 hyperintensity and enlargement of the nerve, muscle oedema if denervation is acute, and thenar atrophy once the change is chronic.

View larger

Bloods, when the cause is not obvious. TSH for hypothyroidism, HbA1c for diabetes, rheumatoid factor and anti-CCP for rheumatoid arthritis, and serum protein electrophoresis if amyloid is suspected.

The decision. Thenar wasting, objective weakness, denervation on electrodiagnostics or severe persistent symptoms mean advanced compression, and those patients are offered release straight away rather than a conservative trial; counsel them that sensory symptoms improve but thenar strength may not fully return. Everyone else - mild to moderate disease with normal thenar muscles - starts with a splint, and the injection sits between the two as a temporising measure.

Algorithm

View larger

The night splint. Hold the wrist neutral (0-5 degrees of extension), not flexed and not hyperextended, worn during sleep for a minimum of 6-8 weeks. Full-time splinting is no better than night-only. It works best in mild disease, in pregnancy and in recent-onset symptoms, with a success rate of 30-50% at six months.

How that squares with the trial evidence. Gerritsen's randomised comparison against open release found splinting successful in 54% at three months - a little above the 30-50% quoted above - against 80% for surgery. By 18 months 41% of the splinted group had crossed over to surgery. Splinting is a reasonable first step, not a destination.

Activity modification. Avoid sustained wrist flexion (computer use, driving), take frequent breaks from repetitive tasks, adjust the workstation (keyboard height, mouse position), and modify the grip in cycling and tennis.

Corticosteroid injection. Both a treatment and a diagnostic test: relief within days that lasts weeks to months confirms the tunnel is the site of the problem. Give 1ml (10-40mg) methylprednisolone with 1ml of local anaesthetic, and splint for two weeks afterwards. Expect 70-80% short-term relief and 20-30% long-term benefit, with a maximum of 2-3 injections in total.

• Position the wrist slightly extended
• Enter 1cm proximal to the distal wrist crease, ulnar to palmaris longus
• Angle the needle 45 degrees distally
• Advance just past the retinaculum, felt as a loss of resistance
• Inject 0.5ml of saline first; it should meet no resistance, which guards against an intraneural injection

What the injection can and cannot buy. The Cochrane review found reliable benefit at one month against placebo but no demonstrated relief beyond that, and a second injection added nothing significant. The five-year randomised follow-up is blunter still: a single injection delayed surgery and modestly reduced the eventual surgery rate, but 84-97% of patients across all three arms had been operated on by five years. Most patients with moderate to severe disease will eventually need a release.

Everything else has thin evidence. Oral corticosteroids as a short course only; vitamin B6 with no proven benefit in trials; NSAIDs for symptom control without disease modification; diuretics with no evidence of efficacy; therapeutic ultrasound with inconsistent evidence; and yoga or stretching, with some evidence for symptom relief.

Pillar pain is the commonest complication and follows from the operation itself: dividing the ligament separates the thenar and hypothenar muscle origins from it. Patients describe an ache at the bases of the thumb and the hypothenar eminence, worse with grip and pinch. It typically resolves by 3 months, can persist across the 3-6 months given in the table above, and occasionally runs to 6-12 months; it is managed with reassurance, grip strengthening and thenar massage. Releasing the ligament only partially to avoid it is not the answer, because it risks an incomplete decompression.

Incomplete release is the commonest cause of persistent symptoms: a ligament not divided fully distally or proximally, an inadequate proximal release of the antebrachial fascia, or anomalous anatomy that went unrecognised. Repeat nerve conduction studies and ultrasound showing residual compression make the case, and the treatment is revision open release with an extended division.

Median nerve laceration is rare and devastating. It presents immediately after surgery with dense numbness and weakness, and it needs urgent exploration and primary repair, with a nerve graft if there is a gap.

Recurrent motor branch injury is commoner, particularly with the transligamentous variant, and presents as thenar weakness. Explore and repair if it is recognised within 72 hours; late recognition is managed conservatively, with tendon transfers if needed.

Complex regional pain syndrome brings disproportionate pain, stiffness and swelling with autonomic changes in temperature, colour and sweating. Recognising it early is what changes the course. Management is multidisciplinary - pain specialist, physiotherapy and occupational therapy, psychology - with gabapentin, vitamin C prophylaxis (some evidence) and sympathetic blocks if severe.

Refer for hand therapy when recovery is not following that course: severe pre-operative stiffness, pillar pain persisting beyond 6 weeks, symptoms of complex regional pain syndrome, slow functional recovery, grip weakness or a sensitive scar. The therapist works on oedema control (elevation, retrograde massage), scar desensitisation and massage, progressive grip and pinch strengthening, tendon gliding, and conditioning for return to work.

What release achieves. 85-95% of patients report an excellent or good outcome, 90-95% get relief of nocturnal paraesthesia, and 80-90% recover sensation completely if they were mild to moderate before surgery. Motor recovery is the variable one, depending on severity and duration of compression.

Where the lower figures come from. Bland's postal survey of 1268 decompressions returned a self-reported success rate of 69%, well below the figures above, and found middle-grade nerve conduction abnormality did better than either very severe or normal studies. Satisfaction is higher with appropriate patient selection, which is what the factors below describe.

Measuring the outcome. The Boston Carpal Tunnel Questionnaire (BCTQ) is the most widely used validated measure, with an 11-question Symptom Severity Scale and an 8-question Functional Status Scale; an improvement of 1.0 point or more is clinically significant. After release the symptom score improves from 3.2 to 1.5 out of 5 and the functional score from 2.8 to 1.4, with maximal improvement at 3-6 months. 85-90% of patients would have the operation again, and 10-15% are disappointed by persistent symptoms or complications.

Recurrence and persistence are different problems. True recurrence - scar tissue recompressing the nerve - affects 5-10% over ten years, is commoner in diabetes, rheumatoid arthritis and dialysis patients, and presents as a gradual return of symptoms after a period of genuine relief. Persistent symptoms, the commoner of the two, never had that symptom-free interval, and mean the ligament was not fully divided. Both are confirmed by repeat nerve conduction studies and ultrasound, and both are treated by revision release - with neurolysis and fat flap or vein wrap coverage where scar is the problem.

Carpal tunnel syndrome is the most common compression neuropathy worldwide and is managed in essentially every health system, but diagnostic pathways and thresholds for surgery differ by region and resource setting.

Global Epidemiology

The best population-based estimate comes from Atroshi et al (JAMA 1999, DOI), a Swedish general-population survey of 2466 responders: clinically certain CTS prevalence was 3.8%, electrophysiologically confirmed median neuropathy 4.9%, and combined clinical-plus-electrophysiological CTS 2.7%. Hand paraesthesia symptoms were far more common (14.4%), so symptoms alone over-estimate disease.

• Prevalence: roughly 3-5% of the general adult population, with a consistent female predominance (about 3:1 in most series).
• Occupational burden: markedly higher in jobs with forceful, repetitive or vibratory hand use (e.g. food/meat processing, assembly, construction) - work-related attributable fractions are concentrated in a small number of high-risk jobs (Roquelaure et al, BMC Public Health 2018, DOI).
• Systemic associations: diabetes, obesity, hypothyroidism, pregnancy, rheumatoid arthritis and amyloid raise risk across all populations.

Major Guidelines, Side by Side

Across guidelines the common threads are: (1) CTS is fundamentally a clinical diagnosis; (2) electrodiagnostic testing confirms the diagnosis, grades severity and predicts outcome but is not mandatory in classic cases; (3) splinting and a single corticosteroid injection are reasonable first-line measures for mild-moderate disease; and (4) surgical decompression is the definitive treatment, indicated urgently when there is thenar wasting or motor loss. The main genuine divergence is how readily electrodiagnostics are required before surgery (more emphasised in North America and for medico-legal/occupational cases) versus a more clinically driven pathway elsewhere.

Registry and Large-cohort Evidence

CTS is not an implant procedure, so it is not captured by the arthroplasty joint registries (NJR, AJRR, AOANJRR, SHAR). The closest "registry-scale" evidence comes from large national audit and administrative cohorts. Bland (Muscle Nerve 2001, DOI) analysed 1268 decompressions and reported a 69% self-reported success rate, with pre-operative nerve-conduction grade the strongest outcome predictor - very severe (axonal) and entirely normal studies both did worse than mid-grade abnormality.

Global Practice Variation

• Anaesthesia: wide-awake local anaesthetic no-tourniquet (WALANT) and field-block open release are increasingly standard worldwide and are particularly valuable where theatre and anaesthetic resources are limited.
• Technique: endoscopic release is more common in well-resourced centres for faster return to work, but Cochrane evidence (Vasiliadis 2014) shows no long-term advantage over open release.
• Imaging: high-resolution ultrasound (median nerve cross-sectional area) is used as a first-line confirmatory test in some European and Asian centres, reducing reliance on nerve conduction studies.

Medico-Legal and Consent Considerations (universal)

Informed consent should cover the natural history, conservative options and their success rates, and surgical risks: median or recurrent motor branch injury (0.1-0.5%), pillar pain (10-20%), incomplete relief (especially when thenar atrophy is already present), complex regional pain syndrome (rare), and the recovery/return-to-work timeline. Documentation of severity grading, thenar assessment, electrodiagnostic results and the duration of conservative treatment is good practice in any jurisdiction. The recurring litigation themes are nerve transection, incomplete release with persistent symptoms, and delayed diagnosis leading to irreversible thenar wasting.