Abstract
Introduction
A recent review of the literature on metal-on-metal total hip arthroplasties (THA) revealed the lack of comparative clinical studies with a sufficient sample size and the inclusion of patient-reported outcomes as well as patient activity levels.
Methods
We conducted a prospective cohort study including all metal-on-metal and conventional polyethylene (PE)-ceramic THAs with an uncemented cup (Morscher press-fit cup), a 28mm head and operated upon via a lateral approach at our University hospital between 1/1999 and 12/2008. Only THAs for primary osteoarthritis were included. The study population is part of the Geneva Hip Arthroplasty Registry, a prospective cohort followed since 1996. The following outcomes were compared between the two groups (metal-on-metal=group 1 vs. PE-ceramic bearing=group 2): (1) Complication rates with respect to infection, dislocation and revision, (2) Radiographic outcomes (presence of linear or focal femoral osteolysis, loosening), and (3) Clinical outcomes (Harris Hip score increase, SF-12, activity and patient satisfaction evaluation, presence of groin pain). Patients operated between 1/1999 and 12/2004 were evaluated five years postoperatively by an independent assessor. Cox regression analysis was used to compare incidence rates while adjusting for differences in baseline characteristics.
Results
1988 THAs were included, 544 with a metal-on-metal and 1444 with a PE-ceramic bearing. The two groups differed significantly with respect to gender distribution (men 56% vs. 41%), mean age (66 vs. 73 years), co-morbidities and type of stem (uncemented 16% vs. 2%). Crude incidence rates for complications were: 0.16 vs. 0.11 cases/100 person-years for infection; 0.37 vs. 0.35 cases/100 person-years for dislocation; and 0.29 vs. 0.16 cases/100 person-years for all-cause revision (incidence rate ratio 1.8, 95% CI 0.7; 4.6). After adjusting for baseline differences the slightly higher risks for infection and revision in group 1 were attenuated. Osteolytic lesions were found in 3.7% of metal-on-metal vs. 4.7% of PE-ceramic THAs. After adjustment for age, gender and activity the OR was 0.6 (95% CI 0.2;2.1). Five years postoperative, 176 THAs of group 1 and 695 THAs of group 2 were seen at follow-up. Clinical outcomes were similar among the two groups with a mean Harris Hip score increase of 39.2 vs. 40.9 points. SF-12 mental and physical health, patient satisfaction (9.3 vs. 8.9 on visual analogue scale) as well as the incidence of groin pain (1.7% vs. 1.2%) was also similar among the two groups. The activity level was significantly higher in group 1 (6.4 vs. 5.4, p<0.001).
Conclusion
Mid-term results with respect to complications, revision rates, presence of osteolysis and clinical outcomes were similar among patients with metal-on-metal and PE-ceramic total hip arthroplasties.
