Answer to the Letter to the Editor of Guoping Liao Et Al. Concerning “comparison of Unilateral Versus Bilateral Pedicle Screw Fixation in Lumbar Interbody Fusion: a Meta-Analysis” by W. Ding Et Al. (2014) Eur Spine J 23(2):395–403

Wenbin Ding,Hui Liu,Zhaomin Zheng
DOI: https://doi.org/10.1007/s00586-015-4095-7
IF: 2.721
2015-01-01
European Spine Journal
Abstract:We thank the readers for their interest in our manuscript [1], and we are glad to respond to the three main concerns. Firstly, we did not describe the search strategy in detail in the manuscript, but only mentioned the keywords ‘spinal fusion’, ‘unilateral’, ‘bilateral’ and ‘pedicle screw’. Actually, to expand the search range, we did not choose the method of ‘Mesh’ or ‘Text Word’. Taking PubMed for example, our search strategy was (((unilateral) AND bilateral)) AND ((spinal fusion) OR pedicle screw). Secondly, the readers pointed out that the studies in operative time, blood loss and hospital time were combined by using the method of inverse variance (IV) with the assumptions of a fixed-effects model. Two methods of analysis were available in Review Manager for meta-analysis of continuous data including the IV fixed-effect method and the IV random-effects method [2]. We have already used the IV random-effects model for analysis of these three outcomes. In our manuscript, we mentioned that ‘All of these three indexes might be related with the medical system and proficiency of the surgeons’ [1]. The subjectivity of these three indexes might be the source of heterogeneities. Finally, as suggested, the publication bias should be assessed by some statistical tests. Considering the small amount of included studies, we deleted the funnel plots in the manuscript. We selected the fusion rate, which was the primary outcome to assess the publication bias. As shown in Fig. 1, there was no obvious publication bias between these studies. We followed preferred reporting items for systematic reviews and meta-analyses (PRISMA) to perform this metaanalysis. It was with regret that we did not mention this in our manuscript. In conclusion, we would thank the readers again for taking time to give us their valuable comments.
What problem does this paper attempt to address?