LAUSR

Wirries et al. [1] have written an interesting paper on the relationship between lateral joint space width (JSW) and later total hip arthroplasty (THA) after arthroscopically arthrotomy for treatment of femoroacetabular impingement (FAI). In this retrospective cohort study, the authors underline the importance of pre-operative JSW for the prognosis of patients with FAI. Therefore, the success of a jointpreserving mini-open arthrotomy seems to be dependent on the status of the radiological JSW and the intra-operative cartilage status of the lateral edge. However, we would like to offer the following points for better clarification. Firstly, we noticed that seventy patients (47.0%) were not available for follow-up, and they were excluded from the case series. Whatever, the author did not report the reason for the high percentage of loss. In addition, the author did not report the types of FAI or other vital information corresponded to included 77 patients. Loss to follow-up is inevitable in most cohort studies and commonly leads to bias and loss of statistical power [2]. However, we should control the rate of loss within 10–20% generally [3]. Otherwise, there will be seriously biased estimates of the odds ratios. Loss to follow-up in cohort studies rarely occurs randomly, and we should attempt to achieve the maximum follow-up rate possible. If there is a high loss rate, sensitivity analysis should be carried out. The outcomes of the lost cases are calculated according to the worst and best results, respectively. Otherwise, the reliability of the conclusion is poor. Secondly, the surgeons were an essential intro-operative factor to consider in the logistic regression. On the one hand, different levels of surgeons have various surgical techniques. The conversion rate to THA was reported to vary from 1% in Cvetanovich et al. [4] to 7% in Hevesi et al. at 2 years [5]. The average rate was 3.78% in a systematic review from different surgeons [6]. On the other hand, the study did report the surgical methods in the section of results. Still, it did not introduce the sub-groups as the resection, acetabular rim trimming, and both procedures. The patients should be classified into sub-groups as the surgeons or the surgical methods in the analysis. However, the sample in this study is not large enough, and the additional work might be challenging to carry out. In fact, the number of the included variables is more than ten factors. Thus, the total sample is relatively short. Even though this study had some limitations, we believe it is a meaningful retrospective cohort study. It would undoubtedly bolster further large samples to explore the risk factors.