Herbert screw fixation for scaphoid nonunions. An analysis of factors influencing outcome.

A retrospective review of 160 cases of scaphoid nonunion treated by internal fixation using a Herbert screw with bone grafting was conducted at an average followup of 24 months. Definite radiographic union was achieved in 90% of cases. Based on Cooney's clinical scoring system, 80 cases had an excellent result, 37 had a good result, 33 had a fair result, and 10 had a poor result. Failure of union was related to the existence of avascular changes of the proximal fragment, instability of the fracture fragment, the prolonged delay in surgery, and the location of the fracture site. In the united scaphoids, the lengthy period of postoperative immobilization, the existence of osteoarthritis, and the prolonged delay in surgery were significant factors in the patient's functional outcome. Overall, the results do not support the view that a residual flexion deformity of the scaphoid is less likely to yield a satisfactory outcome, although it seems worthwhile to correct excessive angulation at the time of repair to promote an anatomic union, thereby preventing early arthritis. A bone graft with internal fixation using a Herbert screw and a shorter period of immobilization may give a satisfactory functional result when the nonunion is treated before the onset of arthritic changes in the wrist.