This was a retrospective study looking at all olecranon fractures treated with surgical fixation in a single tertiary trauma center from 1 January 2005 to 31 December 2016. Ethics approval was obtained prior to the initiation of the study. The STROBE guidelines were used to ensure comprehensive reporting of this study(6).
Surgical records from 2005 to 2016 were reviewed from an electronic surgical database through the use of surgical procedures codes and the cases were identified based on the inclusion and exclusion criteria. The inclusion criteria were surgical treated olecranon fractures (AO/OTA 2U1B1(7)) for patients who were 16 years or older. The exclusion criteria were pathological fractures or peri-implant fractures. The radiographs and clinical notes were analyzed.
Variables collected included demographic data, medical history, injury factors, surgery details and clinical outcomes. Injury factors included proportion of low energy trauma or polytrauma, incidence of open fractures, associated elbow fractures such as coronoid or radial head fractures or use of CT scans for fracture purposes.
Clinical outcomes included complications and re-operation rates. Complications that were collected included failure (implant and bony), infection, wound complications, non-union, malunion, ulnar neuritis, heterotopic ossification and symptomatic implant (with or without implant failure). Implant failure was defined as implant migration or breakage with no loss of reduction such as K wire backing out in TBW or screw backing out in plate fixation. Bony failure in contrast was defined as a loss of reduction due to the failure of the fixation construct. Heterotopic ossification was defined by the formation of extra-skeletal bone in the elbow. Fracture union was defined as bridging cortices on at least 3 of the 4 cortices.
In terms of radiological assessment, the Mayo(8) and Schatzker(9) classifications (Figure 1) were used to categorize the fractures. The Mayo classification examines three main factors namely, the stability, displacement and comminution. Type 1 is undisplaced, type 2 is displaced and stable, type 3 is displaced and unstable. Each type is further divided into A and B depending on presence of absence of comminution. The Schatzker classification is a biomechanical-based classification that progressively moves from a stable transverse fracture (Type A, B) to increasingly unstable oblique fracture (Type C) or comminuted fracture (Type D) or oblique-distal fracture (Type E) to the highly unstable fracture dislocation (Type F).
Beyond the radiological classification, 2 additional radiological features namely fracture angle and number of fracture lines were included to provide a further basis for comparison. Fracture angle as a representation of fracture obliquity (Figure 2) was defined as the distal angle obtained from a line drawn along the ulnar shaft to a line drawn along the most prominent fracture line of the distal fragment. Building from the Schatzker classification, the more oblique the fracture, the more unstable it was deemed to be. In the event of multiple fracture lines, the most prominent fracture line was taken for measurement. Fracture angles were grouped (<60 degrees, 61-80 degrees, 81-100 degrees, 101-120 degrees, >121 degrees) in addition to mean fracture angle to facilitate comparisons. The number of fracture lines were identified as a measure of comminution.
Results were grouped into three 4-year intervals in order to demonstrate a trend of differences across a 12-year period. Group 1 were olecranon fractures that were surgically treated from 2005-2008, group 2 from 2009-2012 and group 3 from 2013-2016. Each group was analyzed separately and compared against each other.
Data were entered and analysed using the IBM SPSS Statistics for Windows (IBM Corp, Released 2010, Version 19, Armonk, NY: IBM Corp). For continuous variables, the descriptive statistics were presented in counts, minimum, maximum, mean (standard deviation), and median (interquartile range). As for the binary and ordinal/categorical variables, proportions and percentage were presented.
Comparisons among the three periods (2005-2008, 2009-2012, 2013-2016) were done using the chi-square test for categorical variables and ANOVA test for continuous variables. Results were presented in tabular format showing the change across the three period and the p-value. A p-value of less than 0.05 was considered to be significant. Data with missing information were excluded from the analysis.