Abstract
Tension-band wiring is commonly used to treat transverse patella fractures. The most common configuration has parallel Kirschner (K) wires and a stainless steel wire loop placed in a vertically oriented figure of eight. Early mobilisation is important but can cause detrimental displacement at the fracture site. A wooden model of a transversely fractured patella was used to compare different fixation constructs. The fracture was fixed using two K wires and a stainless steel wire loop. One or two wire twists were used, placed either arbitrarily or at corners of the loop. The loop was oriented either vertically or horizontally. Interfragmentary compression at the point of wire breakage and permanent displacement on cyclic loading was measured for each construct. Placement of the figure of eight in a horizontal orientation with two wire twists at the corner improved interfragmentary compression by 63% (p<0.05, Tukey post hoc test). On cyclic loading, all constructs with vertical figure of eight but none with a horizontal construct failed (p=0.01; Fisher's exact test). Permanent fracture displacement after cyclic loading was 67% lower with horizontal figure of eight constructs (p<0.05; t test). Placing wire twists at the corner and a horizontal placement of figure of eight improves stability of the construct.