J Knee Surg 2018; 31(06): 591-592
DOI: 10.1055/s-0037-1604441
Letter to the Editor
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.

Isolated Group of Failures without Denominator

Michael A. Mont
Department of Orthopaedics, Cleveland Clinic, Cleveland, Ohio
› Author Affiliations
Further Information

Publication History

14 June 2017

20 June 2017

Publication Date:
25 July 2017 (eFirst)

I read with interest the article entitled “Unusually high rate of early failure of tibial component in ATTUNE total knee arthroplasty system at implant-cement interface,” in the Journal of Knee Surgery.[1] Although it is interesting to bring to the attention of orthopaedic surgeons the mechanism of early failures of any device, I had several queries which are as follows:

  1. When one looks at a case series, we really do not know the denominator, and this certainly could have been a series of cases out of more than 100,000 done, which would make this a rare event. Therefore, simply publishing cases are not evidence-based medicine, as we do not know how many of these were implanted. For example, ceramic fractures occur very infrequently, maybe 1 in 10,000; and yet, a small series could easily be published. The key is to know the denominator.

  2. This further underscores the need to promulgate the American Joint Registry, as this would be an ideal topic for a registry study. A large registry can identify failures and further understand the nature of small series of cases. We could not only know how many failures for a particular device occurred but this could also be put into context of how many thousands of cases were performed. If level 4 data (detailed) were also obtained, one could even further not only understand the incidence of failures but also could potentially know the demographics and specific reasons for failure.

To summarize, although this report may raise some interest, I would like to think that it is only an isolated group of failures that can occur with any device in the field from any manufacturer.