Scaphoid Fracture — Diagnosis & Treatment

Scaphoid fractures are the most common carpal fractures, accounting for approximately 70% of all carpal bone injuries. Despite their frequency, they remain diagnostically and therapeutically challenging — they are frequently missed on initial X-ray, have a precarious vascular supply that predisposes to avascular necrosis (AVN), and can develop non-union if not treated appropriately. The consequences of missed or inadequately treated scaphoid fractures — non-union, AVN, and scaphoid non-union advanced collapse (SNAC wrist) — can be devastating and lead to progressive wrist arthritis. Prompt and accurate diagnosis, and appropriate management based on fracture location, is therefore essential.

• Epidemiology: peak incidence in young males (15–30 years) following a FOOSH; one of the most commonly missed fractures in the ED (approximately 20–30% are occult on initial X-ray); accounts for the largest litigation risk in emergency orthopaedics in the UK for missed fractures
• Vascular anatomy: the scaphoid receives its blood supply in a retrograde pattern — the primary blood supply (70–80%) enters the scaphoid through the dorsal ridge via branches of the radial artery; these vessels enter distally and supply the proximal pole retrogradely; the remaining supply enters through the distal tubercle; this retrograde vascularity means the proximal pole is the most vulnerable to AVN — a proximal pole fracture disrupts the only blood supply to the proximal fragment, resulting in AVN rates of approximately 40–50% for proximal pole fractures
• Anatomy: the scaphoid bridges the proximal and distal carpal rows; it is the critical link between the rows; fractures through its waist disrupt the normal kinematics of the wrist; the scaphoid sits in the radial fossa of the distal radius and articulates with the lunate (scapholunate joint), trapezium, trapezoid, and capitate

• Clinical diagnosis: tenderness in the anatomical snuffbox (ASB) — the depression on the dorsoradial wrist between the EPL and EPB/APL tendons; ASB tenderness has high sensitivity (~90%) but low specificity (~40%) for scaphoid fracture; scaphoid tubercle tenderness (on the palmar aspect at the base of the thenar eminence) adds specificity; pain on axial compression of the thumb (Watson`s scaphoid compression test) is additional; a combination of all three clinical signs has sensitivity ~100% — if all three are negative, scaphoid fracture is effectively excluded
• Plain radiographs: standard wrist series (PA, lateral, scaphoid view — PA in ulnar deviation to profile the scaphoid); scaphoid fractures are occult on initial X-ray in approximately 20–30% of cases; the fracture becomes visible on X-ray at 10–14 days (resorption at the fracture margins makes it visible); a negative X-ray does NOT exclude scaphoid fracture — further imaging is mandatory if clinical suspicion is high
• MRI: the gold standard for diagnosing occult scaphoid fractures; sensitivity and specificity >95%; shows bone marrow oedema (STIR/T2 high signal) at the fracture site within 24 hours of injury; also identifies AVN (T1 low signal in the proximal pole), scapholunate ligament injury, and other carpal injuries; MRI should be obtained within 72 hours if the initial X-ray is negative but clinical suspicion is high (to avoid prolonged unnecessary immobilisation); CT scan is the best modality for assessing fracture union, displacement, and the presence of humpback deformity (flexion deformity of the scaphoid at the fracture site)
• CT scan: used for — (1) defining fracture displacement (displacement >1 mm is an indication for fixation); (2) assessing fracture union in the healing phase; (3) identifying humpback deformity (intrascaphoid angle >35° on lateral view); (4) planning surgical approach; CT is NOT as sensitive as MRI for acute occult fractures

• Clinical scaphoid (suspected, X-ray negative): immobilise in a scaphoid cast (below-elbow thumb spica extending to the IP joint of the thumb); repeat X-ray at 10–14 days; if still negative X-ray but persistent clinical signs, obtain MRI; treat as confirmed fracture until imaging clears the diagnosis
• Conservative management: undisplaced waist fractures (Herbert A2, B2 undisplaced) in low-demand patients or those declining surgery; below-elbow scaphoid cast including the thumb (thumb spica) for 8–12 weeks for waist fractures; repeat CT at 8 weeks to confirm union; union rates for undisplaced waist fractures in cast are approximately 90–95%; however, RCT evidence shows equivalent functional outcomes at 10 years for operative vs non-operative management of undisplaced waist fractures, with earlier return to work and lower non-union rate with surgery
• Surgical fixation — Herbert screw: a headless compression screw (Herbert, or variants — Acutrak, HCS) is the standard fixation device; the screw is entirely buried within the bone (headless design), providing compression across the fracture; percutaneous antegrade (from distal pole through the trapezium) or retrograde (from proximal pole through a dorsal approach) approaches are used; arthroscopic-assisted fixation is increasingly used; open ORIF is required for displaced fractures or where percutaneous techniques are not possible; indications for fixation — displaced fractures (>1 mm), proximal pole fractures, professional athletes/manual workers requiring early return to activity, non-union, and perilunate fracture-dislocations
• Non-union management: established scaphoid non-union requires bone grafting to restore the scaphoid`s anatomy and provide biological stimulus for union; for non-union without AVN — Russe graft (corticocancellous iliac crest graft packed into the non-union site + internal fixation); for non-union with AVN (avascular proximal pole) — vascularised bone graft is required (the 1,2-ICSRA — 1st and 2nd intercompartmental supraretinacular artery pedicle graft from the dorsal distal radius; or the free medial femoral condyle vascularised corticoperiosteal graft for large defects)

• SNAC wrist (Scaphoid Non-union Advanced Collapse): the natural history of untreated scaphoid non-union; the scaphoid flexes at the non-union site (humpback deformity — intrascaphoid angle >35°); the flexed distal scaphoid fragment abuts the radial styloid → radioscaphoid arthritis (Stage I); the arthritis extends to the scaphocapitate joint (Stage II); then the entire midcarpal joint (Stage III); SNAC wrist follows a predictable progression over 5–10 years; the management of SNAC depends on the stage — radial styloidectomy (Stage I); proximal row carpectomy or partial wrist fusion (4-corner fusion — capitate-lunate-triquetrum-hamate) for Stages II–III; total wrist fusion for Stage IV (pancarpal OA)
• Humpback deformity: the distal scaphoid fragment flexes relative to the proximal fragment at the non-union site due to the flexion force of the radioscaphocapitate ligament; this creates a DISI (dorsal intercalated segment instability) pattern — the lunate extends (tilts dorsally) as the scaphoid loses its stabilising role on the proximal row; DISI on lateral X-ray (scapholunate angle >70°) in the context of non-union indicates significant carpal instability requiring restoration of normal scaphoid alignment with graft (structural iliac crest graft or vascularised graft to restore length) + fixation
• Radiological assessment of union: clinical assessment (absence of tenderness) is unreliable for confirming union; CT scan through the scaphoid is the standard for union assessment; union is confirmed by trabecular bridging across the fracture site on CT; MRI shows persistent oedema at the fracture site for months even after union — MRI is not reliable for confirming union

• Scaphoid: most common carpal fracture (70%); young males; FOOSH; 20–30% occult on initial X-ray; anatomical snuffbox tenderness — high sensitivity, low specificity; negative X-ray does NOT exclude fracture
• Vascular supply: retrograde (dorsal ridge enters distally, supplies proximal pole retrogradely); proximal pole fracture = highest AVN risk (~40–50%); waist fracture = moderate AVN risk; distal pole = lowest AVN risk
• MRI: gold standard for occult fractures; >95% sensitivity/specificity; within 24 hours; CT: best for union assessment, displacement, humpback deformity
• Herbert classification: A = stable (A1 tubercle, A2 incomplete waist); B = unstable (B2 complete waist, B3 proximal pole, B4 trans-scaphoid perilunate); C = delayed union; D = non-union
• Indications for fixation: displacement >1 mm; proximal pole fracture; non-union; professional athlete; trans-scaphoid perilunate dislocation (emergency); Herbert headless compression screw — entirely buried, headless design
• Undisplaced waist in cast: 90–95% union; 8–12 weeks thumb spica; CT at 8 weeks to confirm union; equivalent long-term outcomes to operative treatment but slower return to work
• SNAC wrist: untreated non-union → humpback deformity → radioscaphoid OA (Stage I) → scaphocapitate OA (Stage II) → midcarpal OA (Stage III); Stage IV = pancarpal; radial styloidectomy (Stage I); 4-corner fusion or proximal row carpectomy (Stage II–III); total wrist fusion (Stage IV)
• Humpback deformity + DISI: intrascaphoid angle >35°; scapholunate angle >70°; requires structural graft (not just cancellous) to restore length + fixation
• Non-union without AVN: Russe graft (corticocancellous) + fixation; Non-union with AVN: vascularised graft (1,2-ICSRA pedicle or medial femoral condyle free flap)