Cubitus varus is the most frequent complication following the treatment of supracondylar humeral fractures in children. We investigated data from publications reporting on the surgical management of cubitus varus found in electronic searches of Ovid/MEDLINE and Cochrane Library databases. In 894 children from 40 included studies, the mean age at initial injury was 5.7 years (3 to 8.6) and 9.8 years (4 to 15.7) at the time of secondary correction. The four osteotomy techniques were classified as lateral closing wedge, dome, complex (multiplanar) and distraction osteogenesis. A mean angular correction of 27.6º (18.5° to 37.0°) was achieved across all classes of osteotomy. The meta-analytical summary estimate for overall rate of good to excellent results was 87.8% (95% CI 84.4 to 91.2). No technique was shown to significantly affect the surgical outcome, and the risk of complications across all osteotomy classes was 14.5% (95% CI 10.6 to 18.5). Nerve palsies occurred in 2.53% of cases (95% CI 1.4 to 3.6), although 78.4% were transient. No one technique was found to be statistically safer or more effective than any other.

Cite this article: Bone Joint J 2014;96-B:691–700.

The glenopolar angle assesses the rotational alignment of the glenoid and may provide prognostic information and aid the management of scapula fractures. We have analysed the effect of the anteroposterior (AP) shoulder radiograph rotational offset on the glenopolar angle in a laboratory setting and used this to assess the accuracy of shoulder imaging employed in routine clinical practice. Fluoroscopic imaging was performed on 25 non-paired scapulae tagged with 2 mm steel spheres to determine the orientation of true AP views. The glenopolar angle was measured on all the bony specimens rotated at 10° increments. The mean glenopolar angle measured on the bone specimens in rotations between 0° and 20° and thereafter was found to be significantly different (p < 0.001). We also obtained the AP radiographs of the uninjured shoulder of 30 patients treated for fractures at our centre and found that none fitted the criteria of a true AP shoulder radiograph. The mean angular offset from the true AP view was 38° (10° to 65°) for this cohort. Radiological AP shoulder views may not fully project the normal anatomy of the scapular body and the measured glenopolar angle. The absence of a true AP view may compromise the clinical management of a scapular fracture.

Cite this article: Bone Joint J 2013;95-B:1114–20.

Between 1998 and 2007, 22 patients with fractures of the scapula had operative treatment more than three weeks after injury. The indications for operation included displaced intra-articular fractures, medialisation of the glenohumeral joint, angular deformity, or displaced double lesions of the superior shoulder suspensory complex.

Radiological and functional outcomes were obtained for 16 of 22 patients. Disabilities of the Arm, Shoulder, Hand (DASH) and Short form-36 scores were collected for 14 patients who were operated on after March 2002. The mean delay from injury to surgery was 30 days (21 to 57). The mean follow-up was for 27 months (12 to 72). At the last review the mean DASH score was 14 (0 to 41). Of the 16 patients with follow-up, 13 returned to their previous employment and recreational activities without restrictions. No wound complications, infection or nonunion occurred.

Malunion of the scapula can be prevented by surgical treatment of fractures in patients with delayed presentation. Surgery is safe, effective, and gives acceptable functional results.

Gaining stable fixation in cases of recalcitrant non-unions can be challenging. These cases can be accompanied by a segmental bone defect and disuse osteopenia. One strategy to gain stable fixation is the use of allografts. Both cortical struts and intramedullary fibular allografts have been used for this purpose in the femur, tibia and humerus. The present study aims to compare the mechanical properties a locking plate, an intramedullary fibular strut allograft and a cortical strut allograft in a femur model of segmental bone defect.

A transverse mid-shaft osteotomy was performed in fifteen third generation large composite femurs. Twelve millimeters of bone was resected to create a segmental bone defect. Fixation was undertaken as follows: Construct F (Fibula): Lateral Non Locking plate and Intramedullary Fibula Allograft Construct LP (Locking Plate): Lateral Locking Plate Constrcut S (Strut): Lateral Non-Locking Plate and Medial Cortical Strut Allograft Axial, Torsional and Bending Stiffness as well as Load-to-Failure were determined using an Instron 8874 materials testing machine.

Overall, construct S was the stiffest, construct F intermediate and construct LP the least stiff. Specifically, the S construct was significantly (p< 0.05) stiffer than the two other constructs in the axial, coronal plane bending, sagital plane bending and torsional modes. Construct F was significantly stiffer than construct LP in the axial and coronal plane bending modes only. Both the S construct (6108 N) and the F construct (5344 N) had a greater Load-to-Failure than the LP construct (2855 N) (p=0.005 and 0.001 respectively).

The construct with a lateral non-locking plate and a medial allograft strut was stiffer and had a higher load-to-failure than the construct consisting of a stand-alone locking plate. An intramedullary fibular allograft with a lateral non-locking plate had intermediate characteristics. Other factors, such as anatomic and biologic considerations need to be considered before choosing one of the above constructs. The allograft procedures should only be used once soft tissue coverage has been obtained and any infection eradicated.