Failure rates of stemmed metal-on-metal hip replacements – Authors’ reply

There are many brands of stemmed metal-on-metal prostheses used in England and Wales (see our original webappendix table 1),

particularly when different stem and cup combinations are taken into account. Even with the largest arthroplasty database in the world, the numbers for each brand are too small to show conclusively that every metal-on-metal design had inferior survivorship results to other bearings. However, we found no evidence that any particular brand was doing significantly better than any other once factors such as head size were accounted for.

There were seven stem and cup combinations that were used in at least 500 patients. Unadjusted 5-year revision rates for these varied from 2·0% (95% CI 1·1–3·6) for a mean head size of 32 mm to 7·2% (5·2–10·1) for a mean of 46 mm. None of these seven brands used 28 mm head sizes. The Metasul brand referred to by Harlan Amstutz and Michel Le Duff uses a 28 mm head, but was used in fewer than 400 cases in England and Wales in the study period; revision rates were not significantly different from those shown in our paper for all 28 mm head sizes.

We acknowledge that the large number of brands used makes it difficult to disentangle any brand effect from other factors, but this also means that the overall results are not distorted by the effect of any individual brand. Our analysis suggests that head size has a larger effect on revision rates than does brand design.

Amstutz and Le Duff state also that “the dominant modes of failure were not reported in this study, leaving the reader with only guesses as to why a difference in survivorship rate was seen”, but this is incorrect. Reasons for revision are reported in table 5 in the webappendix and discussed in the Results section of the paper. These clearly show that revisions for aseptic loosening and pain were more common with stemmed metal-on-metal than with other bearing surfaces.

Since there are many alternatives to stemmed metal-on-metal that have better implant survivorship and do not introduce the possible risks of exposure to increased chrome and cobalt concentrations, we would counsel surgeons and patients to strongly consider these alternatives. In view of these factors, we believe the stance taken by the Dutch Orthopaedic

Association is both reasonable and safe.

MP receives royalties from Depuy for a cemented total hip prosthesis. The other authors declare that they have no conflicts of interest.