A Small Ratio of Proximal to Distal Bone Fragments Is a Risk Factor for Scaphoid Nonunion: A Volumetric Analysis of Preoperative CT.

Small proximal bone fragments are a known risk factor for nonunion, but it is not known what ratio of proximal to distal bone fragments actually results in nonunion.  We hypothesized that a small proximal ratio of proximal scaphoid fragment volume to distal scaphoid volume as measured by preoperative computed tomography (CT) would be a risk factor for postoperative scaphoid nonunion.  We retrospectively examined the factors that led to postoperative nonunion in 78 patients who underwent surgery using free bone grafts for scaphoid nonunion. Nonunion was defined as no evidence of union on plain radiography or CT more than 3 months after injury. Three-dimensional models of scaphoid fractures were created from the preoperative CT of all cases, and volumes were measured. Other patient characteristics, surgical methods, and imaging were investigated.  Persistent nonunion was observed in 13 patients after surgery, a rate of 16.7%. A multivariate analysis showed that only the proximal fragment ratio was an independent factor (union group 0.94, persistent nonunion group 0.54,  = 0.03). Receiver operating characteristic analysis showed that a proximal fragment ratio less than 0.66 was associated with persistent nonunion.  Treatment of scaphoid nonunion with a small proximal fragment ratio should be based on a thorough understanding of the anatomy, blood flow, and carpal kinematics of the scaphoid bone itself.  Level IV, prognostic study.