We hypothesized that locking plates are not superior to non-locking plates in regards to mechanical and wound complications. To test this hypothesis, we conducted a single-center retrospective chart review at a tertiary university medical center. Our clinical information system was searched to retrieve files from all patients receiving internal fixation of a lower leg fracture (ICD-10 S82) in combination with treatment by internal fixation with a standard (non-locking) one-third tubular plate or an anatomical locking plate (OPS 5-793.3R, 5-793.KR, 5-794.2R, 5-794.KR) between January 1, 2015 and December 31, 2021. All records were pseudonymized and this retrospective analysis was approved by the local ethics committee (approval number A 2024-0049). Apart from baseline anthropometric characteristics (age at time of fracture, sex), the presence of diabetes, arterial hypertension, osteoporosis, peripheral artery disease as well as smoking behavior and alcohol abuse at the time of internal fixation were gathered by chart review. Revision surgery due to either wound or mechanical complications within one year after internal fixation were considered as primary endpoint, whereas duration of surgery (cut-seam interval) served as a secondary measure of outcome. Wound-related complications were identified according to the criteria of Metsemakers et al. [7] whereas failure of fixation, implant loosening or breakage, or non-union were summarized as mechanical complications.

Patients with a fibula fracture unrelated to a malleolar fracture, i.e., one of the AO/OTA regions 42 and 43 were excluded. Similarly, those receiving an antibacterially coated locking plate or being transferred from another hospital for revision surgery were not eligible.

An anatomically shaped locking plate for the lateral aspect of the distal fibula with combi-holes in the proximal segment and holes for multidirectionally angular stable fixation in the distal segment (2.7 mm/3.5 mm LCP Lateral Distal Fibula Plat, Johnson & Johnson Medical, Depuy Synthes, Konrad-Zuse-Strasse 19, 66459 Kirkel, Germany) and a 3.5 mm one-third tubular plate (2.7 mm/3.5 mm LCP Lateral Distal Fibula Plate, Johnson & Johnson Medical, Depuy Synthes, Konrad-Zuse-Strasse 19, 66459 Kirkel, Germany) were regularly available at our institution (Fig. 2). The decision for a particular system was left to the discretion of the surgeon although the manufacturer’s recommendations were followed in case of osteopenic bone.

Plates used throughout this study for treatment of distal fibular fractures

(A) Top view of the 3.5 mm one-third tubular plate (DePuy Synthes) (top) and the LCP Lateral Distal Fibula Plate (DePuy Synthes) (bottom). (B) Frontal view of the 3.5 mm one-third tubular plate (DePuy Synthes) (left) and the LCP Lateral Distal Fibula Plate (DePuy Synthes) (right)

All operations were performed by specialized orthopedic surgeons themselves or under their supervision. Preoperative preparation ensured, that the extent of soft tissue swelling allowed for definitive fracture repair. Clinical signs of the reduction in swelling included the absence of new tension blisters, the appearance of small skin folds, the so-called wrinkle sign, and a decrease in circumference. The prep routinely used in the study period was BD ChloraPrep™ (Becton Dickinson, Heidelberg, Germany). Either the direct lateral approach or the posterolateral approach was chosen. In all cases, an open reduction and lateral plate osteosynthesis was performed using either a conventional one-third tubular plate or an anatomically shaped locking plate in a lateral position. If the syndesmosis complex was unstable, a syndesmotic screw was used. As our study criteria excluded Maisonneuve fractures, this applied to selected fractures of AO types 44-B and 44-C1. To determine or exclude instability of the syndesmosis, the stability of the syndesmosis complex was routinely tested intraoperatively after stabilization of all fractures by posterolateral traction on the proximal fibula with a bone hook. The layer-by-layer wound closure including skin suturing was performed by the surgeon himself.

All information was summarized in a spreadsheet (Microsoft® Excel Version 16. 0.5422.1000, Microsoft Corporation, USA) and the mode of osteosynthesis, i.e., non-locking plate or anatomical locking plate, was used for categorization while the need for revision surgery and the duration of the index surgery served as primary and secondary endpoints. Assumptions of normality were tested by means of the Shapiro-Wilk test and evaluated graphically using histograms and Q-Q-plots. That data was described as mean and standard deviation or median and interquartile range (IQR), as appropriate. Analysis of variance was used to compare the metric characteristics. With respect to the primary outcome the Student’s t-test and Welch’s t-test were applied to check for significant differences between groups with normally distributed data and similar variance respectively dissimilar variance, whereas Mann-Whitney U test was performed for non-normally distributed data. Categorical variables are presented as percentages and the Chi-square test was applied for comparison between groups. A two-tailed p value ≤ 0.05 was considered statistically significant. Analysis was performed using SPSS version 27 (SPSS GmbH, Germany) and Sigma Plot 13 (Systat Software GmbH, Germany) were used for visualization of results.

As it is still a matter of discussion, whether and which patients at all benefit from locking vs. non-locking plates in terms of clinical outcome, all patients being at least 15 years of age at time of fracture were initially included. However, and as to be expected locking plate osteosynthesis was preferentially used in the elderly, i.e. a subgroup with a high incidence of comorbidities. In addition, the absolute number of patients treated with locking plates is significantly lower than the number of patients treated with conventional plates. To account for this apparent dysbalance between patients treated with locking and non-locking osteosynthesis, we performed a 2:1 matching for the parameters age and gender.