Introduction: Compression staples are becoming increasingly popular for osteotomies and arthrodesis. Their design can be divided into “Mechanical Compression” or “Shape Memory”. However, there are no publications investigating the actual compressive forces achieved or the ideal limb-length to staple width ratio.

Methods and Materials: Compression was compared using a load cell mounted within a previously validated simulated fusion site. Two designs each of “mechanical compression” and “shape memory” staples were tested and filmed. The effect of altering limb length on compression was noted.

Results: Both designs of “mechanical compression” staple splayed open causing either no net compression or even distraction. Distractive forces of up to 23N were recorded. The “shape memory” staples all achieved compression at the fusion site of between 5 and 25N. Limb length did not appear to alter the compression force achieved. The outcome was not affected by the material used.

Discussion: “Mechanical compression” staples act in a similar manner to the AO principle of a 2-hole compression plate used without a lag screw or pre-bending. Although there is compression of the cis-cortex, the limbs of the staple splay open with a fulcrum around the bridge-limb intersection resulting in distraction of the trans-cortex. “Shape memory” staples compress both the cis-and trans-cortices along the length of the limb leading to adequate stability and compression forces across the fusion site.

Conclusion: “Mechanical compression” staples cause a distractive force rather than a compressive force and we therefore recommend that they are not relied upon for fusion and the manufacturers need to modify the product or it’s indications for use. The “shape memory” staples do provide compression and the length-to-width ratio of the staple does not appear to be important.