Carpal Tunnel Syndrome — Body Handbook

Musculoskeletal · §146

Carpal Tunnel Syndrome

Your hand wakes you at 3 AM, every night, for weeks — that's carpal tunnel syndrome, the median nerve being squeezed where it passes through your wrist, not aging or a bad pillow. It is one of the most reliably treatable conditions in medicine. A thirty-dollar wrist splint resolves it for many people. A fifteen-minute operation under local anaesthesia resolves it for almost everyone else. The catch is recognising it early enough that the nerve recovers fully.

Respond · As-needed Evidence Strong Chapter Musculoskeletal

The headline is sleep — the nightly waking ends. Within weeks of a splint or an injection, the numbness fades and the grip comes back. If it gets to surgery, that is a day-procedure with a thin palm scar that fades, and roughly eight or nine in ten people are done with the condition a year later. Wait too long and some of that hand never fully returns; the one thing this article asks is that you act before that point.

Inside your wrist, the median nerve runs through a tight tunnel — a curve of bone for the floor, a thick ligament for the roof, nine flexor tendons crowding alongside the nerve. The tunnel has no give. Anything that swells the contents — irritated tendon sheaths, fluid retention, an inflamed lining — pushes the pressure up against the nerve.

Normal resting pressure inside is essentially zero — two or three millimetres of mercury. A symptomatic carpal tunnel runs ten times that, and bending the wrist drives it past the level at which the nerve's blood supply gets pinched off. That is the early symptom phase: numbness only when you sleep with your wrist curled, only when you grip the steering wheel too long. Over months the nerve's insulation breaks down. Over years the actual nerve fibres start dying. The early stage reverses completely once the pressure comes off. The late stage does not (Werner & Andary 2002).

Who gets it: women about three times more often than men, mostly between forty and sixty, and anyone with something extra inside the tunnel — pregnancy fluid retention (usually third trimester, usually gone within months of delivery), obesity (Shiri et al. 2015), diabetes (around one in seven) (Atroshi et al. 1999), an underactive thyroid, rheumatoid arthritis.

How a doctor confirms it

Three things, in sequence. First, the story — nightly numbness in the thumb, index, and middle fingers, relieved by shaking the hand out. That pattern alone, drawn out on a paper hand diagram, is roughly 80% sensitive and 90% specific (Katz & Stirrat 1990). Second, three quick bedside tests — the doctor taps over the wrist, has you press the backs of your hands together for a minute, then presses firmly on the tunnel — looking for any of them to reproduce the tingling. None is conclusive on its own; together they raise or lower the suspicion.

Third, the test that actually confirms it: nerve conduction studies. Small electrical pulses are sent down the median nerve at the wrist; how fast and how strong the signal arrives on the other side is measured. A delay across the tunnel — compared to the ulnar nerve in the same hand as a control — is the objective fingerprint. A joint practice parameter from the major neurology and physical-medicine societies established this as the standard (Jablecki et al. 2002). The test takes about half an hour, costs a few hundred dollars or is insurance-covered, and feels mildly uncomfortable rather than painful.

Ultrasound is increasingly used as a quick alternative — a swollen median nerve where it enters the tunnel is visible on the screen — with sensitivity around 89% and specificity around 83% (Wong et al. 2004). It is cheaper, faster, and involves no electrical stimuli; the major orthopaedic guideline endorses it as a reasonable adjunct (Graham et al. 2016).

One useful caveat: roughly one in five people with classic clinical carpal tunnel have entirely normal nerve conduction studies. That does not mean they do not have it. It means the disease is mild enough to be hard to capture electrically, and they should still be treated.

What most guides get wrong

It is not from the keyboard. This is the deeply held belief that will not die, and the best evidence does not support it. A systematic review of occupational studies (Thomsen et al. 2008) and a large two-cohort prospective study of computer workers (Mediouni et al. 2015) both failed to find a consistent link between computer or mouse use and carpal tunnel syndrome. The jobs where rates are clearly elevated are heavy repetitive grip with real force (meat processing, vehicle assembly), hand-held vibrating tools (construction, mining), and sustained extreme wrist postures. Office work is not on that list.

Surgery is not a last resort. The treatment ladder is not "splint forever and operate only if you have no choice." For moderate-to-severe disease, surgery is the right answer up front. Waiting too long with a compressed nerve costs you nerve fibres that do not grow back (Graham et al. 2016).

Wrist exercises do not cure it. Stretches, tendon glides, nerve glides, yoga — evidence for any of these as standalone treatment is weak and short-term. They are reasonable add-ons. They are not substitutes for the splint, the injection, or the operation once disease is confirmed.

The little finger is not part of this. The thumb, index, middle, and the thumb-side half of the ring finger are median territory. The little finger and the other half of the ring finger are ulnar — a different nerve, in a different tunnel, with a different operation. If only your pinky is going numb, that is cubital tunnel syndrome at the elbow, not carpal tunnel at the wrist.

What you give up by waiting

Tonight: three in the morning, you are shaking your hand to get the feeling back. You will do it for ten minutes, fall asleep again, do it again at four-thirty. That is the seventh week in a row.

A few months in: people start asking if you are tired. You are — you have not had an unbroken night in months — but you do not quite want to explain that the problem is your hand. Coffee cups slip out because you did not feel them slipping. Typing feels deniable, but the typos have been adding up.

A year or two: look at the pad of muscle at the base of your thumb on both hands and compare. The affected side is visibly flatter. Pinch grip is weak — opening a jar takes both hands now. That muscle wasting can be stopped by decompressing the nerve. Mostly it does not come back (Padua et al. 2016).

Five years untreated severe carpal tunnel: permanent numbness across three and a half fingers, permanent weakness in pinching anything between thumb and finger. Hand dexterity is how you interact with almost every object in your life. Advanced carpal tunnel quietly takes a portion of that off the table.

What to do, in order

For most people, the order is splint, then injection, then surgery — and the further you have to go, the better the result tends to be.

Splint at night. A rigid wrist brace that holds your wrist straight. Worn from lights-out to morning. The Cochrane review puts it modestly: better than no treatment for short-term symptoms (Page et al. 2012). About thirty dollars over the counter. Almost no downside.

Steroid injection if the splint is not enough — your doctor places methylprednisolone into or just outside the tunnel.

Surgical release. A small cut in the palm; the thick ligament forming the roof of the tunnel is divided; the nerve has room. The whole operation takes ten to fifteen minutes under local anaesthesia. Open and keyhole versions have essentially the same long-term result; keyhole gets you back to work a little faster (Vasiliadis et al. 2014). Multiple randomised trials and a Cochrane review confirm surgery outperforms continued conservative care for both symptoms and hand function at a year (Gerritsen et al. 2002) (Jarvik et al. 2009) (Verdugo et al. 2008).

Carpal tunnel itself is not contraindicated in any group — the question is which treatment fits which situation.

Where this goes wrong

Almost always: operating on the wrong problem. A pinched nerve root in the neck, the median nerve compressed up at the forearm, the ulnar nerve at the elbow, and diabetic nerve damage in both hands can all produce hand numbness that looks similar from the outside but is distributed differently and will not respond to wrist surgery. Nerve conduction studies before surgery are the protection against this. Insisting on them is the single highest-yield thing a patient can do.

The other big one: waiting too long. Severe long-standing disease with established muscle wasting recovers only partially. Surgery stops the damage; it does not always restore what is already lost (Padua et al. 2016).

Smaller things to know: post-surgery palm tenderness ("pillar pain") that lingers for months and then resolves; a 3-5% recurrence rate; roughly one in ten people whose symptoms do not fully clear even after a technically successful operation. A failed steroid injection does not mean you do not have carpal tunnel — it tends to mean you are someone who needs the operation.

What changes when you treat it

First night with the splint: most people sleep through. The thing is obvious — a brick of plastic on the back of the hand — but the wrist stays straight, the pressure inside the tunnel stays flat, and the 3 AM wake-up does not come.

Two weeks after a steroid injection: numbness gone, grip strong, your hand feels like your hand again. By month two or three the symptoms tend to creep back; you are either splinting through it or queueing for surgery.

Two months after surgical release for mild-to-moderate disease: roughly eight or nine in ten people report full or near-full resolution (Atroshi et al. 2013) (Jarvik et al. 2009). The scar in the palm is a thin pink line that fades to invisible over months. You stop thinking about your hand.

A year out from surgery: for the vast majority of people the condition is essentially over. The exceptions are people who waited until the nerve was severely damaged — they keep some numbness, some weakness (Padua et al. 2016). The whole lesson of the entry is in the timing.

Specific populations

Pregnancy. Symptoms in the third trimester are common — somewhere between one in ten and one in three pregnancies — driven by fluid retention rather than a permanently swollen tunnel. Most resolve within months of delivery. The default plan is splint plus wait; if symptoms are severe, a steroid injection is reasonable; surgery is deferred unless motor function is rapidly deteriorating (Padua et al. 2016).

Diabetes. About one in seven diabetics has carpal tunnel (Atroshi et al. 1999). It coexists with diabetic nerve disease, which complicates the diagnostic picture, but moderate-to-severe carpal tunnel still warrants surgical decompression in this group.

Underactive thyroid. Treating the thyroid sometimes resolves the carpal tunnel without local intervention. A thyroid panel is worth checking before anything invasive.

Obesity. A BMI of 30 or more roughly doubles the risk (Shiri et al. 2015). Weight loss is not a primary treatment for active carpal tunnel — the timescale does not fit — but it shifts the long-term risk picture.

Adjacent conditions worth recognising as separate from this one:

Cubital tunnel syndrome — the ulnar nerve compressed at the elbow. Pinky and ring-finger numbness, worse with the elbow bent.
Cervical radiculopathy — a pinched nerve root in the neck, hand symptoms in a broader pattern, often with neck or arm pain.
De Quervain tenosynovitis — thumb-side wrist pain, tendons rather than nerve, different treatment.
Diabetic peripheral neuropathy — both hands and both feet, glove-and-stocking pattern, not focal.
Workstation ergonomics — not actually a cause of carpal tunnel, but a real factor in hand fatigue and posture.

Related in the handbook

— Keyboard and mouse position is one of the modifiable loads behind wrist nerve symptoms — straighten the setup first.
— Carpal tunnel and frozen shoulder share the same risk soil - diabetes and an underactive thyroid. Having one raises the odds of the other.
— An underactive thyroid makes carpal tunnel more likely — worth checking if it's bilateral or unexplained.
— Diabetes roughly doubles carpal tunnel risk — high blood sugar leaves the median nerve more vulnerable to compression.
— Hand and wrist pain isn't always the nerve — overuse tendinopathies can mimic or sit alongside carpal tunnel.

Substance and claimed effects

Carpal tunnel syndrome (CTS) is compression of the median nerve at the wrist, where the nerve passes beneath the transverse carpal ligament alongside nine flexor tendons. It is the most common entrapment neuropathy in humans, with general-population prevalence of approximately 3-5% in adults rising to 14% in diabetics (Atroshi et al. 1999). The classic clinical syndrome is nocturnal paresthesia of the thumb, index, middle and radial half of the ring finger, relieved by shaking the hand (the "flick sign"); advanced disease produces thenar muscle atrophy, loss of two-point discrimination, and weakness of thumb abduction and opposition. Untreated severe CTS causes permanent sensory and motor deficit. The entry covers diagnosis (clinical features and electrodiagnostic confirmation), the conservative treatment ladder (night splinting, ergonomic modification), corticosteroid injection, and surgical decompression outcomes. Consequences scored holistically: substantial short-term health improvement (symptom relief is often dramatic), sleep restoration (CTS is one of the most reliable medical causes of nightly waking), downstream energy and focus, mood lift from chronic pain relief, modest cost and effort burden, and an extraordinarily strong evidence base.

Evidence by addressing question

mechanism

The carpal tunnel is a fibro-osseous canal at the volar wrist. Carpal bones form a concave floor; the transverse carpal ligament forms a fixed, non-distensible roof. Through this rigid space pass nine flexor tendons (four flexor digitorum superficialis, four flexor digitorum profundus, one flexor pollicis longus) and the median nerve. Because the tunnel volume is fixed, any increase in contents — synovial swelling, tendon thickening, edema, ganglion, mass, or tenosynovitis — directly raises intratunnel pressure (Werner & Andary 2002; Padua et al. 2016).

Normal resting intracarpal pressure is roughly 2-3 mmHg. In symptomatic CTS, pressures of 30 mmHg are typical at rest, with wrist flexion or extension driving pressure past 90 mmHg. Sustained pressure above ~20-30 mmHg compromises epineurial venous drainage; sustained pressure above ~60-80 mmHg blocks arterial inflow. The resulting microvascular ischemia produces the early symptom phase (intermittent paresthesia, position-dependent). Persistent compression then causes focal demyelination of the nerve at the tunnel and ultimately axonal loss — the late phase, when sensory deficit becomes fixed and motor units in the thenar muscles die off (Werner & Andary 2002).

Risk factors fall into three buckets:

Anatomic / demographic: female sex (3:1 ratio; smaller tunnel cross-section), middle age (40-60), genetic predisposition.
Systemic disease that swells the tunnel contents: obesity (BMI ≥30 roughly doubles risk per meta-analysis of 58 studies; Shiri et al. 2015), diabetes (~14% prevalence vs 3-5% baseline; Atroshi et al. 1999), hypothyroidism (myxedematous infiltration), pregnancy (third-trimester fluid retention, usually transient), rheumatoid arthritis, hemodialysis (β₂-microglobulin amyloid deposition), acromegaly.
Occupational: high-force repetitive gripping, hand-arm vibration, and sustained extreme wrist postures (meat processing, vehicle assembly, construction) elevate risk meaningfully. Routine computer keyboarding does not: a systematic review (Thomsen et al. 2008) and a large two-cohort prospective study (Mediouni et al. 2015) both found no consistent causal association between computer/mouse use and CTS.

evidence

Diagnosis rests on three signals. First, symptom pattern. The Katz hand diagram — patient marks a hand outline where symptoms occur — has sensitivity around 80% and specificity around 90% when the pattern is "classic" or "probable" (Katz & Stirrat 1990). Nocturnal numbness in the thumb-index-middle distribution relieved by shaking is the highest-yield single historical feature.

Second, provocative physical-exam tests: Tinel sign (percussion paresthesia over the tunnel), Phalen test (sustained wrist flexion reproducing symptoms within 60 seconds), Durkan carpal compression test (direct manual pressure over the tunnel). Reported sensitivity ranges are wide — Phalen 51-91%, Tinel 23-67%, Durkan 64-87% — and individual specificities likewise variable. The carpal compression test has the most consistent diagnostic accuracy of the three, but no single bedside test is sensitive or specific enough to confirm or exclude CTS on its own (Padua et al. 2016).

Third, electrodiagnostic confirmation: nerve conduction studies (NCS), with or without needle electromyography (EMG), document prolonged median sensory and motor distal latencies across the wrist relative to ulnar control segments. The American Academy of Neurology / AAEM / AAPMR practice parameter on electrodiagnostic studies in CTS established NCS as the reference standard (Jablecki et al. 2002). The Padua neurophysiological classification — minimal, mild, moderate, severe, extreme — stratifies severity and predicts surgical outcome (Padua et al. 1997).

Ultrasound has emerged as an accessible adjunct. Increased median nerve cross-sectional area at the tunnel inlet (threshold typically >10 mm²) correlates well with NCS findings: Wong et al. 2004 reported sensitivity ~89% and specificity ~83%. The AAOS evidence-based clinical practice guideline endorses NCS as the primary confirmatory test, with ultrasound as a reasonable adjunct (Graham et al. 2016). MRI is rarely used clinically (cost, accessibility, no clear added value).

A diagnostic caveat: 10-25% of patients with classic clinical CTS have normal NCS, particularly in early or mild disease. This is "clinically defined CTS with normal electrodiagnostics" and is treated based on clinical findings; responsiveness to splinting or injection is reasonable confirmation in this subset.

protocol

The treatment ladder is one of the best-established in MSK medicine (Padua et al. 2016; Graham et al. 2016).

First-line: nocturnal wrist splinting in neutral position. A rigid prefabricated splint that holds the wrist near 0° prevents the sleep-related wrist flexion/extension that drives intracarpal pressure spikes. A Cochrane review found splinting more effective than no treatment for symptom and functional improvement at short follow-up (Page et al. 2012); long-term durability beyond a few months is uncertain. Cost $15-40 over-the-counter. Worn every night until symptoms resolve; resumed if symptoms recur.

Second-line: local corticosteroid injection. Methylprednisolone 40-80 mg injected proximal or into the tunnel produces substantial short-term symptom relief. The pivotal Atroshi double-blind placebo-controlled RCT randomized 111 patients with idiopathic CTS to 80 mg methylprednisolone, 40 mg methylprednisolone, or saline placebo; both active arms produced significantly greater symptom improvement than placebo at 10 weeks, with 80 mg superior to 40 mg, but by 1 year ~73% of the 80 mg group, 81% of the 40 mg group, and 92% of the placebo group had undergone surgery (Atroshi et al. 2013). The Cochrane review confirms a real but transient benefit at 1-3 months (Marshall et al. 2007). Useful as a bridge in pregnancy CTS (avoiding surgery while awaiting postpartum resolution), as a diagnostic confirmation (responders are good surgical candidates), or as a stalling measure where surgery is contraindicated.

Definitive: surgical carpal tunnel release. The transverse carpal ligament is divided, decompressing the nerve. Open release uses a 2-4 cm palmar incision; endoscopic release uses one or two small portals and an endoscopic blade. A Cochrane review of endoscopic versus open CTR found equivalent long-term symptom outcomes; endoscopic produced slightly faster return to work and slightly higher transient nerve-injury rates (Vasiliadis et al. 2014). Both are typically 10-15 minute day-procedures under local or regional anesthesia. Surgical decompression outperforms splinting and conservative care for symptom relief and functional outcome at 12-18 months (Gerritsen et al. 2002; Jarvik et al. 2009; Verdugo et al. 2008). Complete or near-complete symptom resolution at 1 year is reported in 70-90% of patients; numbers are higher in mild-to-moderate disease and lower (around 50-70%) in severe long-standing disease where axonal loss limits recovery (Padua et al. 2016).

Indications for proceeding directly to surgery, per AAOS and Padua: moderate-to-severe NCS findings, thenar atrophy, motor weakness, failed 6-week conservative trial, or rapidly progressive symptoms (Graham et al. 2016).

contraindications

Treatment-specific rather than condition-specific.

Corticosteroid injection: active skin/joint infection at the site; uncontrolled diabetes (expect transient hyperglycemia); rare but documented inadvertent intraneural injection causing nerve worsening.
Surgical release: active local infection; uncorrectable coagulopathy or anticoagulation that cannot be safely interrupted; severe peripheral vascular disease.
Pregnancy: splinting and observation are first-line; injection considered for severe symptoms; surgery deferred unless motor function is rapidly deteriorating (most pregnancy-related CTS resolves within months postpartum).

misconceptions

"Computer keyboards cause carpal tunnel." The best available evidence — a systematic review of occupational studies (Thomsen et al. 2008) and a large two-cohort prospective analysis (Mediouni et al. 2015) — finds no consistent causal link between routine computer or mouse use and CTS. The persistent occupational signals are with high-force repetitive grip, hand-arm vibration, and extreme sustained wrist postures.
"Surgery is a last resort." AAOS guideline framing and the Padua review both treat surgical release as appropriate first-line for moderate-to-severe disease, not as a fallback (Graham et al. 2016; Padua et al. 2016). Delay risks irreversible axonal loss.
"Wrist exercises cure CTS." Tendon-gliding, nerve-gliding, and yoga programs have weak short-term evidence and no demonstrated durable effect; they are not a substitute for splinting, injection, or surgery in confirmed cases.
"It will go away on its own." True for some pregnancy-related CTS and some very mild idiopathic CTS. Moderate-to-severe untreated CTS progresses; thenar atrophy, once present, does not reliably reverse.
"Pinky numbness is CTS." The pinky and ulnar half of the ring finger are ulnar-nerve territory; numbness there is cubital tunnel syndrome (ulnar entrapment at the elbow), not CTS. A patient with pure pinky symptoms has the wrong diagnosis.

audience

Population breakdown matters for both prevalence and treatment selection.

Women: approximately 3:1 female-to-male ratio; smaller carpal tunnel anatomy and hormonal contributions to flexor tenosynovium. Women also present at younger ages on average (Atroshi et al. 1999).
Pregnant women: third-trimester onset is common; symptomatic CTS in roughly 10-30% of pregnancies. Most cases resolve within months of delivery. Splint + watchful waiting is standard first-line; injection reserved for severe; surgery deferred (Padua et al. 2016).
Older adults: cumulative compression damage; surgery is still effective for symptom relief but motor recovery is slower and often incomplete when thenar atrophy is established.
Diabetics: roughly 14% prevalence; CTS coexists with diabetic peripheral neuropathy and outcomes are somewhat worse, but surgery remains recommended for moderate-to-severe disease.
Hypothyroid patients: treating the thyroid sometimes resolves CTS without local intervention.

alternatives

Tendon-gliding and nerve-gliding exercises: weak short-term evidence; reasonable adjunct, not substitute.
Yoga: small RCT support for Iyengar-style programs; not in mainstream guideline recommendations.
Therapeutic ultrasound, low-level laser: small studies show inconsistent benefit; insufficient evidence to recommend.
Oral NSAIDs: no demonstrated benefit over placebo for CTS.
Oral corticosteroids: short-term symptom benefit but inferior to local injection and not sustainable.
Ergonomic devices (wrist rests, vertical mice, split keyboards): may reduce hand fatigue in symptomatic users but do not prevent CTS in unexposed populations.
Acupuncture: small positive RCTs; not in major guideline recommendations.

failure-modes

Operating on the wrong diagnosis. Cervical radiculopathy (C6-C7), proximal median neuropathy (pronator syndrome), peripheral diabetic polyneuropathy, and thoracic outlet syndrome can mimic CTS. Each produces a different distribution; none responds to carpal tunnel release. Pre-surgical NCS is the protection against operating on a mimic.
Operating too late. Severe long-standing CTS with established thenar atrophy and axonal loss recovers poorly: decompression stops further damage but does not always restore lost function. Time-to-surgery matters.
Recurrent CTS post-surgery (3-5%): causes include incomplete release of the ligament, scar tissue around the nerve, and persistent occupational exposure.
Pillar pain: persistent tenderness in the thenar and hypothenar eminences after open release; usually self-limited over months but a real source of patient dissatisfaction.
"Failed steroid trial" misread as ruling out CTS: non-response to injection does not exclude CTS; it identifies a patient who likely needs surgical decompression.

practicalities

Diagnosis pathway: GP recognises the syndrome, refers to neurologist or hand surgeon for NCS. NCS takes ~30 minutes; mildly uncomfortable (small electrical stimuli, occasional needle EMG of thenar muscles).
Splint: $15-40 OTC; standard widely available. No prescription needed.
Steroid injection: out-of-pocket cost ~$200-500 in the US without insurance; covered by most insurance with diagnosis.
Surgery: US self-pay typically $1500-3000 for open and $2000-4000 for endoscopic; insurance covers both with appropriate diagnosis. Return to light desk work in 1-2 weeks; manual labor 4-6 weeks. The scar fades to a faint line.

stakes

Felt-experience forecast of untreated CTS over time, anchored to the established natural history (Padua et al. 2016; Werner & Andary 2002).

Tonight: 3 AM wake-up with a hand that feels like it has been slept on, except this is the seventh week in a row. Shake it out, lose another forty minutes of sleep.
Six months in: cumulative sleep loss surfaces as foggy afternoons that look like ordinary tiredness but aren't. Buttons get harder. Dropping things you can't quite feel slipping.
Two years untreated: visible flattening of the muscle pad at the base of the thumb. Pinch grip weak — jar lids defeat you. Once the muscle is wasted, surgery stops further loss but rarely fully restores it.
Five-plus years untreated severe CTS: permanent sensory deficit across thumb-index-middle and permanent loss of thumb opposition. Hand dexterity is what mediates almost every motor interaction with the world; advanced CTS quietly removes it.

payoff

Felt-experience forecast of treatment, anchored to RCT and Cochrane evidence (Atroshi et al. 2013; Gerritsen et al. 2002; Jarvik et al. 2009).

First night with a splint: most patients sleep through. The splint is obvious — a brick of plastic and Velcro — but the wrist stays neutral, intracarpal pressure stays flat, and the 3 AM wake-up doesn't come.
Two weeks after a steroid injection: numbness gone, grip strength back. Catch: 1-3 months and the symptoms tend to return. A bridge, not a cure.
Two months after surgical release (mild-to-moderate disease): 80-90% report full or near-full symptom resolution. The palm scar is a thin pink line that fades.
One year post-surgery: the condition is largely a non-event for the vast majority. Severe long-standing cases retain residual numbness or weakness — the lesson is don't wait that long.

out-of-scope

Adjacent and frequently confused conditions worth flagging:

Cubital tunnel syndrome (ulnar nerve compression at the elbow): pinky and ring-finger numbness, elbow-flexion provocation.
Cervical radiculopathy (C6 or C7 root): neck-and-arm pain, broader distribution, no nocturnal-shaking relief.
De Quervain tenosynovitis: radial-side wrist pain, distinct anatomy, distinct treatment.
Diabetic peripheral neuropathy: bilateral stocking-glove distribution, length-dependent, not focal.
Ergonomic workstation setup: real effect on hand fatigue and posture even if it does not cause or prevent CTS.

The credibility range

Optimist case

Carpal tunnel syndrome is one of the most clearly characterised peripheral neuropathies in medicine. Pathophysiology is mechanistically resolved (compressive ischemia → demyelination → axonal loss). Diagnosis is confirmable with objective measurements (NCS, ultrasound). Splinting has Cochrane support; corticosteroid injection has placebo-controlled RCT support; surgical release has multiple RCTs and Cochrane reviews demonstrating superiority over conservative care for symptom relief and functional outcome at 12-18 months. Surgical release in well-selected patients with mild-to-moderate disease produces 80-90% complete or near-complete symptom resolution, durable beyond a decade. Few conditions in MSK medicine have a diagnostic-to-curative pathway this well-supported (Padua et al. 2016; Graham et al. 2016).

Skeptic case

The diagnostic boundary is fuzzy. 10-25% of clinically symptomatic patients have normal NCS, and asymptomatic individuals can show abnormal NCS; the diagnostic gold standard is imperfect. Some surgical "successes" may reflect natural-history fluctuation or regression to the mean rather than treatment effect, particularly in milder presentations. The headline 80-90% surgery success rate is from heterogeneous study populations; severe long-standing cases see materially worse outcomes (~50-70% incomplete recovery). Steroid injection benefits are real but largely transient — most injected patients reach surgery within 1 year regardless (Atroshi et al. 2013). Recurrence rates of 3-5% post-surgery plus a 10% incomplete-resolution rate mean surgery is not 100% curative. Many "CTS" diagnoses made in primary care without electrodiagnostic confirmation are misattributed hand-pain syndromes that will not respond to any CTS-specific treatment.

Author's call

The evidence base is among the strongest in MSK medicine. The condition is real, diagnosable, and effectively treatable across the full severity spectrum. The skeptic case's strongest point — that diagnosis is fuzzy at the margins — argues for NCS confirmation before any surgical decision, not against the treatment paradigm. Practical guidance: classic nocturnal-hand-numbness symptoms in median-nerve distribution warrant a clinician visit, NCS where available, a nightly neutral-position splint at minimum, and escalation to injection or surgery based on severity and response. Do not wait years. Do not accept "it's just my keyboard"; the workplace evidence does not support that attribution. Meta scoring lands at evidence = 5 (multiple Cochrane reviews plus AAOS guideline), controversy = 1 (broad agreement on the paradigm; minor disagreement on optimal injection dose, ultrasound's role, and exact timing of surgery).

Stakeholder and incentive map

Hand surgeons and orthopaedic groups: surgical-volume incentive partially balanced by the durability of the operation (no repeat business per patient).
Ergonomic device manufacturers: commercial reliance on the "keyboards cause CTS" narrative that the occupational literature does not support; the products themselves can still reduce hand fatigue in symptomatic users.
Insurers: cover surgery readily once NCS is documented; sometimes require extended conservative trials first.
Physical therapists and hand therapists: offer tendon-gliding and ergonomic programs; evidence weak but low-risk and often integrated into pre- and post-operative care.
Employer / workers'-compensation systems: stake in occupational-causation attribution; jurisdictional politics around what counts as work-related CTS.
Patients with chronic hand pain seeking diagnosis: real-world incentive to receive a treatable diagnosis can pull diagnostic thresholds toward over-attribution.

Population variability

Sex: female:male ~3:1 (Atroshi et al. 1999).
Age: peak incidence 40-60; uncommon under 20.
Pregnancy: ~10-30% of pregnancies produce symptomatic CTS; most resolve within months postpartum.
Obesity: BMI ≥30 roughly doubles risk (Shiri et al. 2015).
Diabetes: prevalence ~14%, often coexisting with diabetic polyneuropathy that complicates diagnosis.
Hypothyroidism, rheumatoid arthritis, acromegaly, amyloidosis: secondary CTS that may improve with treatment of the underlying condition.
Occupational: meaningfully elevated in high-force repetitive grip and hand-arm vibration jobs; not meaningfully elevated in computer-based work (Thomsen et al. 2008; Mediouni et al. 2015).
Hemodialysis: β₂-microglobulin amyloid deposition in long-term dialysis is a known cause.
Severity at presentation: predicts surgical outcome — mild-moderate disease recovers near-completely; severe with axonal loss recovers partially.

Knowledge gaps

Optimal timing of surgery — at what severity threshold does waiting cost too much axon? No consensus, despite obvious clinical relevance.
Ultrasound as primary diagnostic versus NCS — ultrasound is cheaper, faster, and patient-tolerable, but long-term outcomes from ultrasound-only-confirmed diagnosis are less studied.
Optimal corticosteroid dose — Atroshi 2013 favoured 80 mg over 40 mg at 10 weeks, but dose-response data remain thin.
Open versus endoscopic at very long follow-up (10+ years) — both look equivalent at 1-5 years; longer-term durability data are limited.
Whether ergonomic interventions prevent CTS in low-risk populations — population baseline risk is low enough that prevention trials are underpowered.
Mechanistic basis for symptomatic CTS with normal NCS — a meaningful subgroup whose pathophysiology is not yet fully characterised.

Scope matches the brief end to end — symptoms, nerve-conduction confirmation, splinting and ergonomic management, steroid injection, surgical decompression outcomes. No narrowing relative to the input description.

Action type. Scored as respond rather than know or decide. The entry's centre of gravity is "you have these symptoms, here's what to do next", and the treatment ladder is well-defined enough that decide would understate how clearly the path is laid out. Could be argued either way.
Cadence. as-needed rather than course. Definitive surgery is a one-time bounded course, but the entry covers the whole symptom-triggered response — which is episodic, not scheduled. Splinting may recur indefinitely; this maps to as-needed better than to course.
Cost burden 2. Treatment costs are bounded one-time-ish rather than annual ongoing — splint ~$30, injection ~$200-500 self-pay, surgery ~$1.5-4k self-pay with most insurance covering. The 0-5 ladder is calibrated to annual ongoing cost; a one-time bounded course of mid-three-figures-to-low-four-figures lands at 2.
Energy / focus / mood scored 2 each (downstream effects). No CTS RCT measures these directly; the scores reflect mechanism — sleep restoration and chronic-pain relief drive them. Could plausibly be 1 each if scored strictly on direct evidence; the holistic-scoring rule in meta §5a §7 pushes toward 2.
Controversy 1. Broad guideline alignment on the paradigm. Active debates exist at the margins (ultrasound vs NCS as primary diagnostic, optimal steroid dose, exact severity threshold for skipping ahead to surgery) but none threaten the core ladder.
Keyboard misconception deliberately foregrounded. The biggest patient-side barrier to treatment in office workers is the belief that they need to "fix their ergonomics" instead of seeing a clinician. The literature (Thomsen et al. 2008; Mediouni et al. 2015) is unusually clear on the absence of a causal link for routine computer work; the article spends a paragraph on it.
Future-link candidates: cubital-tunnel-syndrome, cervical-radiculopathy, de-quervain-tenosynovitis, workstation-ergonomics, diabetic-peripheral-neuropathy. Cross-link from out-of-scope once any of these exist.
Separate-entry candidates: Ergonomic workstation setup deserves its own entry — real effect on hand fatigue and posture even if it does not prevent or cause CTS. Pregnancy carpal tunnel could justify a standalone but is well-served by the audience block here.
Hard call on injection coverage. Some patient resources frame the steroid injection as a "cure" — the Atroshi 2013 trial data are unambiguous that it is not, and the article reflects that. Worth flagging because it contradicts what many readers will have heard from a primary-care doctor.
What stays out: cervical-radiculopathy differential, full Padua neurophysiological grading detail, β₂-microglobulin amyloid in dialysis, mini-incision and wide-awake CTR variants. All are in the research dossier as superset material but would clutter the reader-facing article.