Does placing screws off-centre in tubular bone alter their pullout strength?

Screws are used to fix broken bones either directly or through plates. Surgeons sometimes find that a screw they have used is not quite in the centre of the bone but to one side or maybe even the edge. It has been postulated that screws catching the edge of the bone do not give good fixation and may even predispose to fractures. We conducted the present experiment using porcine femora to see if a screw's transverse plane position in the bone made any difference to its pullout strength. 20 cortical screws were inserted into the cortical segments of 5 pig femora (4 screws per femur) using the standard AO technique. The screws were inserted in one of 5 randomly chosen positions-centre, medial off-centre, lateral off-centre, medial edge and lateral edge. The screws were tested to failure in axial pullout using a Losenhausen universal testing machine. We found that 4 of the 8 'edge' screws failed with fractures developing around the screw track during pullout testing. These 4 screws were noted on cross-section to have 100% bone contact with their threads completely embedded in the cortical bone. They also had significantly lower pullout resistance than the 4 'edge' screws without fractures (p=0.05) and the 12 'non-edge' screws (in the central 75% of the bone) (p=0.03). This was most likely due to the associated fractures. There was a statistically significant association between the 'edge' screw position (i.e. within 12.5% of the medial or lateral edge of the bone) and the likelihood of fracture (p=0.000). We conclude that in the transverse plane, cortical screws either on their own or through plates should be inserted in the central three-fourths of the bone. Screws placed outside this zone carry a higher risk of fixation failure due to fractures around the screw track with axial loading.