J Knee Surg 2019; 32(07): 702
DOI: 10.1055/s-0038-1660842
Letter to the Editor
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.

Response to: Reliability of Roentgenographic Knee Alignment Measurements in Gonarthrosis: Methodological Issues to Avoid Misinterpretation

Chang-Chin Wu
1  Department of Orthopedics, College of Medicine, National Taiwan University Hospital, Taipei, Taiwan
› Author Affiliations
Further Information

Publication History

09 May 2018

09 May 2018

Publication Date:
06 August 2018 (eFirst)

Reliability of Roentgenographic Knee Alignment Measurements in Gonarthrosis: Methodological Issues to Avoid Misinterpretation

We appreciate the opinions and the methodologies proposed by the author of the letter,[1] but in the beginning, we searched for the proper statistical methods for this kind of analysis between radiological calculations of angles of high tibial osteotomy and interpretations of different observers. There is no related literature on this topic, and we thus turned and consulted a paid expert of biostatistics because of the limited capability of all four orthopaedic doctors in this kind of analysis. A generalized estimating equation model was chosen after many discussions as each observation had eight repeated measures from the four investigators and two methods. We did not apply intraclass correlation coefficient single measure or Bland–Altman's plot for different regress objects as suggested by the consulted expert and therefore to compare the differences between these two methods, another consultation to test the data should be done again.

We agree that statistically significant and clinically important are two completely different methodological issues for the shortage and incompetence of present clinical data before the perfect clinical results were gathered and published. This is our limitation that we did not mention in the article and we apologize for this missed point.

We sincerely thank the author for the kind and valuable comments.