Session VIII - Upper Extremity

Complications of Olecranon Osteotomies

Clifford B. Jones, MD; Sean E. Nork, MD; Julie Agel, MA; Douglas P. Hanel, MD; M. Bradford Henley, MD, University of Washington, Harborview Hospital, Seattle, WA

Purpose: To evaluate the results of olecranon osteotomies used in patients with AO Type C supracondylar humeral fractures treated according to AO protocols.

Material & Methods: 46 patients with AO Type C supracondylar humeral fractures treated at a level-1 trauma center from 1990-96, were evaluated retrospectively by chart and radiology review. Thirty-four patients of the 46 (73.9%) had olecranon osteotomies for surgical exposure of the elbow joint. Six had ipsilateral open olecranon fractures, 2 were skeletally immature with open physes, and 4 had follow-up for less than 6 months. The average age at injury was 40 years (range: 18 to 83). The dominant upper extremity was injured 53% of the time. Thirteen fractures (38.2%) were open (9 Type II, 3 Type IIIA and 1 Type IIIB-free flap).

Results: In 12 patients, the osteotomy technique was performed incorrectly. In 2 patients, multiple errors were documented. In 7 of these patients the chevron osteotomy was not perpendicular to the shaft of the ulna, although only 3 resulted in a malunion. Three osteotomies were malpositioned and extended either into the coronoid (n=2) or into the tuberosity (n=1). Four patients had the tension band wire placed dorsal to the proximal ulnar fixation although no complications were associated with this finding. Overall, 23 (68%) of the osteotomies healed with normal alignment or without complications. 17 complications were noted in the remaining 11 patients. Twelve (35%) reoperations were related to the olecranon fixation. Five (15%) osteotomies healed with >5 of sagittal plane deformity at the olecranon. The average amount of gapping and step-off at the osteotomy site was 2.3 mm (0 mm [n=5], 1-2 mm [n=18], > 2 mm [n=10]) and 1.0 mm (0 mm [n=18], 1-2 mm [n=11], > 2 mm [n=5]), respectively. Five (15%) osteotomies failed to heal within 6 months. The average 6.5 mm cancellous screw length (58 mm) in the failures was statistically shorter (p=0.03) than the average screw length in the healed fractures (69 mm). Three osteotomies had broken hardware (2 removed); 7 (20%) had hardware removed because of prominence and discomfort. The average amount of measured hardware prominence was 5.8 mm (range: 4-10mm).

Discussion / Conclusion: Since all patients had olecranon osteotomies with concomitant supra-intercondylar AO C-type fractures, this precluded correlation of elbow flexion and extension solely with the osteotomy. Despite this limitation, this patient group exhibits the morbidity associated with the transolecranon approach to the elbow joint. As a result of this analysis, an effort should be attempted to leave the extensor mechanism (bony and tendonous) intact by working on both sides of the triceps tendon. If an osteotomy must be performed in order to improve visualization, data suggest that the cut must be perpendicular to the ulnar shaft and that the fixation should be longer than 70 mm.