I would like to thank Dr. Labek for his interest in the commentary I wrote related to unicondylar knee arthroplasty (1).

As stated in my commentary, and as known to all of us in the surgical community, the success of any surgical procedure is dependant on numerous parameters, most of important of which may be appropriate patient selection and proper execution of the surgery. I am in agreement with Dr. Labek in that studies performed in high volume centers or by innovators of a particular procedure may not be generally reproducible in the community. Hence discordance in data, including survivorship of a particular prosthesis, may exist between studies from single centers and the Registry data. I do not, however, accept the proposition that Registry data provides more “representative” information than well-designed and executed studies. In fact I believe some Registries have numerous shortfalls both related to variability in data collection and also lack of adequate data that should caution anyone from generalizing their findings.

Reference

1. Parvizi J. Unicompartmental knee replacement did not differ from total knee replacement with regard to clinical outcomes at 15 years. J Bone Joint Surg Am. 2009;91:2012.

To the Editor:

In his commentary (1), Javad Parvizi compared a great variety of studies and datasets. Clinical studies like the one by Newman et al. (2) are based on samples and try to draw conclusions on the general population from the results. The results of such studies implicate a multitude of influencing factors like the expertise of the surgeons involved, specific surgical techniques, the implants used or patient selection. These factors may have a substantial impact on the outcome.

Register data, by contrast, comprise virtually all surgeries performed in a certain area and thus provide a good reflection of average patient treatment and its outcome. Here, too, divergences in the respective areas from the individual surgeon’s standard certainly are a potential confounder for data assessment, but they are usually smaller than in sample-based studies.

Therefore, the differences between the results of the study by Newman et al. and register studies are not contradictory. A hospital publishing a study of such high quality like Newman et al. can well be supposed to have a particularly high interest and extraordinary expertise. Also, the number of surgeries performed has an influence on the outcome (3).

It would therefore have to be questioned whether the preconditions of the group examined by Newman et al. correspond with the average standard in patient treatment. The data from registers show that it is obviously impossible in average practice to obtain the same revision results with unicompartmental knee replacement as are achieved with total knee replacement. This is also true for countries such as Australia, a fact that clearly supports this conclusion on a wide geographical scale (4).

The outcome measuring end-points of the study by Newman et al. and of register studies are ordinal and nominal data requiring relatively large numbers of cases for calculations of statistical significance. Under these circumstances, the statistical power of the study is relatively low. It should therefore be not be concluded from the absence of a statistically significant difference in the sample that there is no difference.

A commentary on evidence-based orthopaedics is usually aimed at making general recommendations. Javad Parvizi’s conclusion that it is possible "with stringent criteria for patient selection and technically proficient execution of the procedure" to achieve results with UKA that are comparable to TKA outcomes is certainly correct for the study collective. However, the results seem to be irreproducible in general practice.

It is the objective of high-quality study designs such as blinded randomized controlled trials to minimize the inevitable impact of the sample in the case of general conclusions. The worldwide development of national arthroplasty registers enables orthopaedists – as a positive exception in medicine – to obtain comparative values based on the entire population and all surgeries performed. Thus, in the field of arthroplasty, real outcome values are available from registers – data that, in other areas of medicine, have to be acquired by approximation on the basis of complex study designs and structured meta-analyses.

In the assessment of diverging results these differences should be taken into account.

The author did not receive any outside funding or grants in support of his research for or preparation of this work. Neither he nor a member of his immediate family received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which the author, or a member of his immediate family, is affiliated or associated.

References

1. Parvizi J. Unicompartmental knee replacement did not differ from total knee replacement with regard to clinical outcomes at 15 years. J Bone Joint Surg Am. 2009;91:2012.

2. Newman J, Pydisetty RV, Ackroyd C. Unicompartmental or total knee replacement: the 15-year results of a prospective randomised controlled trial. J Bone Joint Surg Br. 2009;91:52-7.

3. Robertsson O, Knutson K, Lewold S, Lidgren L. The routine of surgical management reduces failure after unicompartmental knee arthroplasty. J Bone Joint Surg Br. 2001;83:45-9.

4. National Joint Replacement Registery of the Australian Orthopaedic Association. Annual Report 2008. http://www.dmac.adelaide.edu.au/aoanjrr/publications.jsp?section=reports2008. Accessed 2009 Dec 5.