The inferior lag screw position produced the highest axial and torsional stiffness. Anterior and posterior lag screw positions produced the lowest stiffnesses and load-to-failure. Inferior placement of the lag screw on the anteroposterior radiograph and central placement on the lateral radiographs is recommended.

Intramedullary nails are commonly used to repair femoral fractures. Fractures in normal healthy bone often occur in the young during motor vehicle accidents. Although clinically beneficial, bone refracture and implant failure persist. Large variations in human femur quality and geometry have motivated recent experimental use of synthetic femurs that mimic human tissue and the development of increasingly sophisticated theoretical models. Four synthetic femurs were fitted with a T2 femoral nailing system (Stryker, Mahwah, New Jersey, USA). The femurs were not fractured in order to simulate post-operative perfect union. Six configurations were created: retrograde nail with standard locking (RS), retrograde nail with advanced locking 'off' (RA-off), retrograde nail with advanced locking 'on' (RA-on), antegrade nail with standard locking (AS), antegrade nail with advanced locking 'off' (AA-off), and antegrade nail with advanced locking 'on' (AA-on). Strain gauges were placed on the medial side of femurs. A 580 N axial load was applied, and the stiffness was measured. Strains were recorded and compared with results from a three-dimensional finite element (FE) model. Experimental axial stiffnesses for RA-off (771.3 N/mm) and RA-on (681.7 N/mm) were similar to intact human cadaveric femurs from previous literature (757 + 264 N/mm). Conversely, experimental axial stiffnesses for AS (1168.8N/mm), AA-off (1135.3N/mm), AA-on (1152.1 N/mm), and RS (1294.0 N/mm) were similar to intact synthetic femurs from previous literature (1290 +/- 30 N/mm). There was better agreement between experimental and FE analysis strains for RS (average percentage difference, 11.6 per cent), RA-on (average percentage difference, 11.1 per cent), AA-off (average percentage difference, 13.4 per cent), and AA-on (average percentage difference, 16.0 per cent), than for RA-off (average percentage difference, 33.5 per cent) and AS (average percentage difference, 32.6 per cent). FE analysis was more predictive of strains in the proximal and middle sections of the femur-nail construct than the distal. The results mimicked post-operative clinical stability at low static axial loads once fracture healing begins to occur.

Optimal surgical positioning of cable-screw pairs in repairing periprosthetic femur fractures near the tip of a total hip implant still remains unclear. No studies in the literature to date have developed a fully three-dimensional finite element (FE) model that has been validated experimentally to assess these injury patterns. The aim of the present study was to evaluate the biomechanical performance of three different implant-bone constructs for the fixation of periprosthetic femoral shaft fractures following total hip arthroplasty. Experimentally, three bone-plate repair configurations were applied to the periprosthetic synthetic femur fractured with a 5 mm gap near the tip of a total hip implant. Constructs A, B, and C, respectively, had successively larger distances between the most proximal and the most distal cable-screw pairs used to affix the plate. Specimens were oriented in 15 degrees adduction, subjected to 1000 N of axial force to simulate the single-legged stance phase of walking, and instrumented with strain gauges. Computationally, a linearly elastic and isotropic three-dimensional FE model was developed to mimic the experimental setup. Results showed excellent agreement between experimental versus FE analysis strains, yielding a Pearson linearity coefficient, R2, of 0.90 and a slope for the line of best data fit of 0.96. FE axial stiffnesses were 601 N/mm (Construct A), 849 N/mm (Construct B), and 1359 N/mm (Construct C). FE surface stress maps for cortical bone showed maximum von Mises values of 74 MPa (Construct A), 102 MPa (Construct B), and 57 MPa (Construct C). FE stress maps for the metallic components showed minimum von Mises values for Construct C, namely screw (716MPa), cable (445MPa), plate (548MPa), and hip implant (154MPa). In the case of good bone stock, as modelled by the present synthetic femur model, optimal fixation can be achieved with Construct C.