End caps are intended to prevent nail migration (push-out) in elastic stable intramedullary nailing. The aim of this study was to investigate the force at failure with and without end caps, and whether different insertion angles of nails and end caps would alter that force at failure.

Simulated oblique fractures of the diaphysis were created in 15 artificial paediatric femurs. Titanium Elastic Nails with end caps were inserted at angles of 45°, 55° and 65° in five specimens for each angle to create three study groups. Biomechanical testing was performed with axial compression until failure. An identical fracture was created in four small adult cadaveric femurs harvested from two donors (both female, aged 81 and 85 years, height 149 cm and 156 cm, respectively). All femurs were tested without and subsequently with end caps inserted at 45°.

In the artificial femurs, maximum force was not significantly different between the three groups (p = 0.613). Push-out force was significantly higher in the cadaveric specimens with the use of end caps by an up to sixfold load increase (830 N, standard deviation (SD) 280 vs 150 N, SD 120, respectively; p = 0.007).

These results indicate that the nail and end cap insertion angle can be varied within 20° without altering construct stability and that the risk of elastic stable intramedullary nailing push–out can be effectively reduced by the use of end caps.

Cite this article: Bone Joint J 2015;97-B:558–63.

We investigated the static and cyclical strength of parallel and angulated locking plate screws using rigid polyurethane foam (0.32 g/cm³) and bovine cancellous bone blocks. Custom-made stainless steel plates with two conically threaded screw holes with different angulations (parallel, 10° and 20° divergent) and 5 mm self-tapping locking screws underwent pull-out and cyclical pull and bending tests. The bovine cancellous blocks were only subjected to static pull-out testing. We also performed finite element analysis for the static pull-out test of the parallel and 20° configurations. In both the foam model and the bovine cancellous bone we found the significantly highest pull-out force for the parallel constructs. In the finite element analysis there was a 47% more damage in the 20° divergent constructs than in the parallel configuration. Under cyclical loading, the mean number of cycles to failure was significantly higher for the parallel group, followed by the 10° and 20° divergent configurations.

In our laboratory setting we clearly showed the biomechanical disadvantage of a diverging locking screw angle under static and cyclical loading.

A cavovarus foot deformity was simulated in cadaver specimens by inserting metallic wedges of 15° and 30° dorsally into the first tarsometatarsal joint. Sensors in the ankle joint recorded static tibiotalar pressure distribution at physiological load.

The peak pressure increased significantly from neutral alignment to the 30° cavus deformity, and the centre of force migrated medially. The anterior migration of the centre of force was significant for both the 15° (repeated measures analysis of variance (ANOVA), p = 0.021) and the 30° (repeated measures ANOVA, p = 0.007) cavus deformity. Differences in ligament laxity did not influence the peak pressure.

These findings support the hypothesis that the cavovarus foot deformity causes an increase in anteromedial ankle joint pressure leading to anteromedial arthrosis in the long term, even in the absence of lateral hindfoot instability.