Abstract
Following the work of Avenell et al. [1, 2] that has raised concerns about the integrity of the Yamaguchi Osteoporosis Prevention Study (YOPS) conducted by Ishida and Kawai, [3] we would like to issue an adjustment to our systematic review on the effectiveness of vitamin K on bone mineral density and fractures [4].
The work of Ishida and Kawai that has been questioned was included in our systematic review and was considered to be at high risk of bias. The study was included in four meta-analyses within our results. We feel there is a need for rapid correction when concerns, such as this one, are identified [5].
We do not think the removal of the YOPS study changes the overall results of our paper or the conclusions we drew regarding fractures or bone mineral density. We have provided recalculated estimates for each of these meta-analyses without the YOPS study.
Total Clinical Fractures - Figure 2.
This meta-analysis was performed with a subgroup analysis based on the risk of bias assessment. As the study by Ishida and Kawai was deemed to be at high risk of bias, the results of the low risk of bias studies remain unchanged. Removal of the YOPS study changes the overall odds ratio (OR) from 0.72 (95%CI 0.55 to 0.95) to 0.74 (95%CI 0.56 to 0.97).
Total Vertebral Fractures - Figure 3.
This meta-analysis was performed with a subgroup analysis based on the risk of bias assessment. As the YOPS study was deemed to be at high risk of bias, the results of the low risk of bias studies remain unchanged. Removal of the Ishida and Kawai study changes the overall odds ratio from 0.96 (95%CI 0.83 to 1.11) to 0.98 (95%CI 0.85 to 1.14).
This also does not change the conclusions or interpretation of our review.
Mean difference in Bone mineral Density of the Radius at 12 Months - Figure 4, part 3.
The YOPS study had a weighting of 37.49% in this meta-analysis and thus the removal of this study did change the estimate of the effect for this analysis. The estimate changed from 0.14 (95%CI − 1.04 to 1.33) to − 0.58 (95%CI − 1.24 to 0.08). This means that the result now favours the control arm over vitamin K.
Mean difference in Bone Mineral Density of the Radius at 24 months - Figure 5, part 3
Only two studies had data available for this, YOPS and Cheung et al. Taking the results only from Cheung et al. [6], the estimate would be 0.00 (95%CI − 0.85 to 0.85).