Abstract
There is limited evidence in the literature suggesting that ceramic-on-ceramic (CoC) THA is associated with lower risk of revision for prosthetic joint infection (PJI) than other bearing combinations especially metal-on-poly (MoP) and metal-on-metal (MoM). Pitto and Sedel reported hazard ratios of 1.3 – 2.1 for other bearing surfaces versus CoC. Of interest, the PJI rate was not significantly lower in the first 6 months, when most infections occur, but only became significant in the long term. While factors such as patient age, fixation, mode, OR type, use of body exhaust suits, and surgeon volume were considered in the multivariate analysis, BMI, medical comorbidities, and ASA class were not. This is a major weakness that casts doubt on the conclusion, since those three factors are MAJOR risk factors for PJI AND all three factors are more likely to be unevenly distributed, and much more likely present in groups other than CoC. The data was also limited by the fact that it was drawn from a retrospective review of National Registry data, The New Zealand Joint Registry. While similar findings have recently been reported from the Australian Joint Registry, the danger in attributing differences in outcomes to implants alone is possibly the single greatest danger in interpreting registry results. While device design can impact implant survival, other factors such as surgical technique, surgeon, hospital, and especially patient factors have a far greater likelihood of explaining differences in observed results. A recent report from the same New Zealand joint registry reported that obesity, ASA class, surgical approach, and trainee operations all were associated with higher PJI and all would be more likely in non-CoC THAs. Accuracy of diagnosis is also a major concern. Revision for trunnionosis is more common in non-CoC THA and is frequently misdiagnosed as PJI.
Numerous non-registry studies and reviews have compared PJI in CoC vs. other bearings and none have concluded than the incidence of PJI differed significantly.
