Retraction of meta-analysis on statin loading pre-PCI for ACS
RETRACTION 7 Out of 13 studies included in a meta-analysis of RCTs on atorvastatin loading prior to PCI for ACS should have been excluded based on the study’s selection criteria. A corrected analysis substantively alters the results.
News - Mar. 1, 2019A meta-analysis of RCTs on atorvastatin loading prior to PCI for ACS that we recently covered has been retracted by the European Heart Journal.
Turgeon and Althouse write in the European Heart Journal that they identified several errors when reading the article of Ye et al .. They state that at least 7 of the 13 studies included in the meta-analysis should have been excluded based on the study’s selection criteria.
For two ACS studies, the reason was that none of the patients underwent PCI. Four included studies were not conducted primarily in ACS patients, and in most cases they were excluded entirely. The last study that should not have been included was a non-randomised prospective registry study, rather than an RCT.
A corrected analysis on the six remaining trials yields weaker effect sizes than originally reported based on the 13 trials. In the corrected analysis, there is insufficient evidence that a statin loading dose prior to PCI for ACS improves outcomes. Moreover, the results are largely driven by those of SECURE-PCI: the other, smaller, studies offer little information beyond SECURE-PCI. Because the corrected analysis substantively alters the results and conclusions, Turgeon and Althouse recommend that the article by Ye et al . be withdrawn for correction.
That is indeed what happened, write editor-in-chief Lüscher and Davies . The European Heart Journal editorial team became aware of the situation after Turgeon and Althouse expressed their concern on Twitter, after which they were invited to write a contribution to the journal’s Discussion Forum. The editorial team, after review, reached the same conclusion about the inappropriate inclusion of 7 studies in the meta-analysis. Ye et al. have acknowledged mistakes in their work, apologised, and will submit a revised version.