Introduction: Aseptic loosening of the acetabular cup is the commonest cause for revision surgery of total hip replacements (THR). Whereas a sound bone cement micro-interlock is believed to contribute most to the stability, most surgeons prefer to enhance their fixation by adding a macro-interlock. Drilling of additional keyholes creates cement pegs intended to resist rotational forces at the acetabular bone-cement interface. Only a few attempts have been made to investigate the effect of number, configuration and shape of these keyholes. Following the limited experiments by Oh (1983) on beechwood blocks, Mburu (1999) systematically optimised diameter, depth and number of keyholes using the same beechwood model. He subsequently developed a novel drill, bell-shaped in cross-section, aiming to minimise the stress concentration observed at the base of cement pegs.
Methods: This study compares the novel drill against a conventional drill using the same beechwood model and the previously optimised number, diameter and depth of keyholes, the shape of the drills being the only difference. The tests were performed on twenty beechwood blocks, half of the blocks allocated to the novel drill and half to the conventional drill. Since the three keyholes were not equidistant, it was also tested whether the direction of torque had an influence on the results. This was done by applying torque clockwise in half of the specimens and anticlockwise in the other half.
Results: The results showed that the static torque to failure was superior in the novel drill (mean: 163 Nm) compared to the conventional drill (mean: 127 Nm), but this was not statistically significant (p= 0.12) due to the wide variation of results. There was no difference for the direction of torque (p= 0.8) and the type of drill and the direction of torque did not show any interaction (p= 0.5).
Conclusions: Results suggest that there is an improvement with the novel drill in static torque to failure. Further testing appears therefore to be worthwhile. Results also suggest that the difference or improvement achieved with the novel drill is less pronounced than it is for number, depth and position of keyholes (Mburu, 1999). However, more variability than expected was encountered. Despite recognition as a possible model for keyhole testing, beechwood blocks testing three keyholes at the same time may not be appropriate when the load to failure for cement pegs is approximating the load to failure for beechwood as encountered in our experiment. Simplification of the system like testing single pegs should be considered.
