Herbert Classification of Scaphoid Fractures

View larger

View larger

The Herbert system was introduced alongside the Herbert dual-pitch compression screw in 1984. It classifies scaphoid fractures into four types based on stability, chronicity, and displacement.

Each main type has subtypes that refine the classification further.

• The A2 versus B2 distinction hinges on displacement and completeness. An A2 hairline fracture must be genuinely non-displaced and incomplete on CT. Any displacement (even one millimetre of gapping or translation) upgrades the fracture to B2, which warrants fixation. Over-treating A2 with surgery is acceptable; under-treating B2 with a cast is not.
• Proximal pole fractures (B3) are inherently unstable. Even if non-displaced, the tenuous retrograde blood supply means prolonged casting carries an unacceptably high nonunion rate. Most hand surgeons now recommend operative fixation for all proximal pole fractures regardless of displacement.
• CT is the imaging modality of choice for classification. Plain radiographs underestimate displacement and cannot reliably distinguish complete from incomplete waist fractures. A dedicated scaphoid CT (sagittal and coronal reconstructions along the scaphoid axis) is essential before committing to non-operative treatment.
• Time to diagnosis matters. Scaphoid fractures diagnosed after a delay of more than 4 weeks have a higher nonunion rate. A patient presenting late with snuffbox tenderness and a radiographic waist fracture may already be a type C (delayed union) rather than an acute type B.
• The Herbert screw has largely been superseded by cannulated headless compression screws (e.g. Acutrak, Synthes HCS), but the classification remains the decision-making framework. The principle of rigid internal fixation with a headless compression screw spanning the fracture is unchanged.
• Type D2 nonunions with a humpback deformity and dorsal intercalated segment instability (DISI) require correction of the lunate angle in addition to bone grafting and fixation. A vascularised pedicle graft from the 1,2 intercompartmental supraretinacular artery (1,2-ICSRA) is a standard option for sclerotic proximal pole nonunions.
• Patient factors influence the decision. An elite athlete or manual worker with a type A2 fracture may reasonably choose percutaneous fixation to return to activity sooner, even though casting would likely succeed. Conversely, a low-demand patient with significant comorbidities may be managed more conservatively even for some type B fractures.

• No single society guideline replaces the classification. BOA/BSSH (UK), AAOS (US) and AO Foundation teaching all use the Herbert stability framework to triage cast versus fixation; the AAOS appropriate-use criteria and BOAST scaphoid pathways emphasise CT for equivocal injuries and early specialist referral for proximal-pole and displaced fractures.
• A widening role for early fixation. Internationally there is a trend toward percutaneous fixation of non-displaced waist fractures in high-demand patients (athletes, military, manual workers) to shorten time to union and return to work, balanced against the cost and small surgical risk — union rates are equivalent to casting, so the decision is functional, not about achieving union.
• Resource-setting variation. Where CT/MRI access is limited, presumptive casting with clinical follow-up and repeat radiographs remains standard; in well-resourced centres, immediate advanced imaging reduces both over-immobilisation of non-fractures and missed unstable fractures.
• Salvage for failed nonunion (proximal row carpectomy, four-corner fusion) is consistent across regions for SNAC-wrist arthritis when union cannot be achieved.