Periprosthetic Joint Infection in Hip Arthroplasty: Is There an Association Between Infection and Bearing Surface...

Preliminary studies have raised the question of whether certain prosthetic biomaterials used in total hip arthroplasty (THA) bearings are associated with increased risk of periprosthetic joint infection (PJI). For example, some observational data suggest the risk of PJI is higher with metal-on-metal bearings. However, it is not known whether other bearings-including ceramic bearings or metal-on-polyethylene bearings-may be associated with a higher or lower risk of PJI.

The objective of this study was to use a national arthroplasty registry to assess whether the choice of bearings-metal-on-polyethylene (MoP), ceramic-on-polyethylene (CoP), ceramic-on-ceramic (CoC), or metal-on-metal (MoM)-is associated with differences in the risk of revision for deep infection, either (1) within 6 months or (2) over the entire period of observation, which spanned 15 years.

Data from primary THAs were extracted from the New Zealand Joint Registry over a 15-year period. A total of 97,889 hips were available for analysis. Inclusion criterion was degenerative joint disease; exclusion criteria were previous surgery, trauma, and any other diagnosis (12,566 hips). We also excluded a small group of ceramic-on-metal THAs (429) with short followup. The median observation period of the selected group of hips (84,894) was 9 years (range, 1-15 years). The mean age of patients was 68 years (SD ± 11 years), and 52% were women. There were 54,409 (64%) MoP, 16,503 (19%) CoP, 9051 (11%) CoC, and 4931 (6%) MoM hip arthroplasties. Four hundred one hips were revised for deep infection. A multivariate assessment was carried out including the following risks factors available for analysis: age, sex, operating room type, use of body exhaust suits, THA fixation mode, and surgeon volume. Because of late introduction of data collection in the Registry, we were unable to include body mass index (BMI, recording started 2010) and medical comorbidities according to the American Society of Anesthesiologists class (ASA, recording started 2005) in the multivariate analysis.

The rate of early PJI (< 6 months) did not differ by bearing surface. In contrast, we observed a difference over the total observation period. Within the first 6 months after the index surgery, CoC THAs were not associated with a lower risk of revision for PJI (p = 0.118) when compared with CoP (hazard ratio [HR], 1.31; 95% confidence interval [CI], 0.50-3.41), MoP (HR, 2.10; CI, 0.91-4.82), and MoM (HR, 2.04; CI, 0.69-6.09). When the whole observation period was considered, CoC hips were associated with a lower risk of revision for deep infection when compared with CoP (HR, 1.30; CI, 0.78-2.18; p = 0.01), MoP (HR, 1.75; CI, 1.07-2.86; p = 0.02), and MoM (HR, 2.12; CI, 1.23-3.65; p = 0.008).

Our finding associating CoC THA bearings with a lower risk of infection after THA must be considered very preliminary, and we caution readers against attributing all of the observed difference to the bearing surface. It is possible that some or all of the observed difference associated with bearing type may have been driven by other factors such as ASA and BMI, which could not be included in our multivariate analysis, and so future registry studies on this topic must assess those variables carefully.

Level III, therapeutic study.