We performed a biomechanical test simulating clinical situation as far as was practicable. A metal acetabulum model was created with an inside diameter of 52 mm (Figure 1). It had 3 holes equidistant from each other, with threads cut in the holes where 3 diaphragmatic type piezoelectric pressure transducers (Figure 2) were screwed in. The tips of the transducers were flush with the surface of the acetabulum. The other end of each transducer was connected to a separate lead which all fed into a computer with software to generate the pressure-time graphs. The transducers were colour coded to correspond with the colours of the curves made by each of the 3 transducers.
3 different cup types were compared, with similar outer diameters of around 48 mm to ensure a uniform cement mantle of at least 2 mm all around. These were, the Charnley Ogee flanged cup (size 47 mm with an outer diameter of 47 mm selected as 48 mm cup is not produced by the manufacturers), the Exeter low profile cup (size 48 with an outer diameter 48 mm selected) which was unflanged, and the Exeter contemporary cup (size 52 mm cup which corresponds with an outer diameter of 48 mm) which was flanged with PMMA beads on the outer surface of the cup designed to prevent bottoming out of the cups. These sizes were chosen to be representative of the cup sizes that would normally be used in actual clinical situation for an acetabulum size of 52 mm, which was the size of our model acetabulum. The Charnley Ogee and the Exeter low profile are currently two of the most commonly used cups in hip arthroplasty, representative of the flanged and unflanged varieties respectively and hence these two were used for the comparison. The Exeter flanged cup is a relatively new product which claims to have additional benefits with the inclusion of beads to prevent bottoming out of the cup and a firmer flange compared to the Charnley cup to prevent eversion of the flange and consequent extrusion of cement during cup insertion and pressurisation. We wanted to observe whether these changes to the Exeter contemporary cup resulted in better pressure generation.
3 sets of experiment were performed using each cup type with Simplex cement giving a total of 9 readings. The whole experiment was then repeated using CMW1 cement. Simplex and CMW1 are the 2 types of cement used commonly in clinical practice in our hospital. We wanted to ensure that our readings were not influenced by cement properties and hence the experiment was performed using both types.
A Lloyd Instruments 6000R plus universal testing machine was used to generate a static load of 70 N for the purpose of the initial cement pressurisation as well as subsequent cup insertion and pressurisation (Figures 3 and 4). A series of pilot readings were obtained whereby an arthroplasty surgeon manually cemented the cups using the set up and 70 N was found to be the force that could be comfortably maintained throughout the process of cement polymerisation without bottoming out of the cups. This pressure figure was hence chosen for our experiment. Besides, previous similar studies had also come up with similar figures for cement pressurisation [7, 8].
The temperature of the lab was maintained at a uniform 20 degrees throughout the experiment. The cement was vacuum mixed according to the manufacturers' instructions (45 seconds for both cement types), and inserted into the metal acetabulum. The cement was then pressurised, until cup insertion, by a cement pressuriser fitted onto the universal testing machine. The acetabular cup was then inserted and pressurised using a cup pressuriser mounted on the universal testing machine. In the case of Simplex, the cup was inserted at 5 and 1/2 minutes and in the case of CMW the cup was inserted at 4 and 1/2 minutes (as per manufacturers' recommendations and clinical protocol), due to the increased viscosity of the CMW1. After each experiment, the residual cement was cleaned out from the acetabular model.
Continuous pressure recordings were obtained with each of the 3 pressure transducers simultaneously and these were represented both graphically on a pressure time curve as well as digitally at 10 seconds intervals till the end of cement polymerisation which was identified as the point where the pressure graphs plateaued near the baseline pressure and this corresponded to the expected in-vivo polymerization times for the Simplex and Palacos cements. This also corresponded to the time that a sample of extruded cement from the experiment had turned hard. The data was analysed using the SPSS 11.0 (Statistical Package for the Social Sciences, SPSS Inc., Chicago, Illinois). The pressures generated by the cups were compared within each cement group and not between cement groups, in order to prevent multiple variables from affecting the readings. A Kolmogorov-Smirnov test was initially used and this confirmed that the data was normally distributed. Independent sample t-tests were then performed within each cement sub-group comparing two types of cup at a time. As the sample sizes were small we also calculated the median and the interquartile range.