Letter to the Editor: Oral risedronate increases Gruen zone bone mineral density after primary total hip arthroplasty: a meta-analysis

Dear editors,

Recently, we read a meta-analysis by Li et al. [1] entitled “Oral risedronate increases Gruen zone bone mineral density after primary total hip arthroplasty: a meta-analysis” with great interest. Periprosthetic bone loss after total hip arthroplasty is an inevitable phenomenon mainly due to stress shielding, which may predispose to aseptic loosening, periprosthetic fractures and challenges at revision surgery. We appreciate the authors’ work in this field, however, some issues in the article that may nullify the conclusion need to be mentioned.

Firstly, the authors declared that they had systematically retrieved electronic databases including PubMed, Embase, Web of Science, Cochrane Library, and Chinese Wanfang database. However, to our knowledge, a study by Yin et al. [2] in Wanfang database was eligible on the basis of inclusion criteria, which could be involved in this meta-analysis and beneficial to draw a more comprehensive and convincing conclusion.

Secondly, we noticed that four of the included studies with short-term follow-up (6–12 months) revealed significant efficacy of risedronate while the left one with relatively longer follow-up (4 years) drew the exact opposite conclusion. Although these studies showed excellent homogeneousness, it was improper to ignore the potential reasons for such difference and simply put them into pooled analysis, which without any doubt would bring extra bias and lead to an incorrect conclusion.

Thirdly, it is obvious that two of the included studies (Skoldenberg, 2011 and Kumar, 2011) were the same article. What’s more, both of them and Muren et al. [3] came from the same clinical cohort. Thus, extracting duplicate data from these three articles for analysis would be more likely to lead to an incorrect conclusion and misleading clinical practice. Given that four of the eligible RCTs were followed up no more than 1 year except Muren et al. (4-year follow-up), we recommend to rule out Muren et al. and conduct a short-term (≤ 1 year) meta-analysis in a reference of the work by Shi et al. [4]. Details of these eligible studies are shown in Table 1.