Retrospective Review of 234 Scaphoid Fractures and Nonunions Treated with Arthroscopy for Union and Complications

The purpose of this paper is to retrospectively review 234 consecutive cases of scaphoid fractures and nonunions treated using arthroscopy with the dorsal percutaneous implantation of a headless compression screw for healing and complications. Solid union of fracture is determined by CT scan. We identified 126 acute injuries, including 65 proximal pole fractures; 67 grossly displaced fractures; 12 trans-scaphoid perilunate dislocations including four trans-scaphoid trans-capitate fractures; and ten combined scaphoid and distal radius fractures. 108 scaphoid nonunions were identified. 98 were correctly aligned and ten had a humpback deformity which was correctable using arthroscopic assisted reduction techniques at the time of surgery. 82 presented with a fracture gap 2mm or greater requiring percutaneous bone grafting. 12 cases of avascular necrosis (AVN) were identified by MRI. 20 nonunions had surgery performed at other institutions. The mean time to surgery for the nonunions was 20 months. 99% union rate of acute scaphoid fractures was obtained by 12 weeks, as determined by CT scan. Two complications were identified (3%). One case of delayed healing was identified. This delayed union was treated with percutaneous bone grafting and continued on to heal uneventfully. The other complication was a case of volar trans-scaphoid peri-lunate dislocation. While the fracture healed, the patient developed a traumatic dislocation requiring a capitate-lunate arthrodesis. Treatment of scaphoid nonunions resulted in ten cases of delayed healing, which were treated with repeat percutaneous bone grafting. This represented a 9% complication rate. Of the ten cases of delayed unions that were re-bone grafted, four failed to heal by nine months. This resulted in a 96% union rate of our nonunion group by nine months. When acute fracture healing was compared to nonunions the average healing of acute fractures as determined by CT scanning measuring trabecular bridging was 12 weeks, while the average healing of nonunions was 22 weeks. We conclude that the dorsal percutaneous treatment of scaphoid fractures and nonunions using arthroscopy is safe and effective. CT scans to evaluate scaphoid healing by measuring trabecular bridging at the fracture site was determined to be an excellent modality to evaluate scaphoid healing. While not witnessed in this series, the potential for complications requires proper training.