During a revision of the database that had been used for the above-referenced article, the authors realized that a mistake had been made in the retrieval of events from one of the trials included in the analysis. The trial (unpublished NCT00698932, saxagliptin vs. placebo) had been incorrectly reported to have observed 2 fractures in control versus 0 in saxagliptin, whereas the result had not been disclosed. The mistake was made with a cut-and-paste command when arranging the database for statistical analysis. The authors controlled all the other trials, finding no further error. They repeated the meta-analysis with the amended database, with the following results: for all trials, the MH-OR for bone fractures with DPP-4 inhibitors is 0.62 (0.38–1.03), P = 0.063 and 0.61 (0.40–0.95), P = 0.029, without and with continuity correction, respectively. For placebo-controlled trials, the corresponding figures are 0.43 (0.22–0.86), P = 0.017 and 0.43 (0.24–0.79), P = 0.007, respectively.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.