This paper by Palmer and colleagues (BMJ 2021;372:m4573, doi:10.1136/bmj.m4573, published 13 January 2021) has been corrected to account for an error in the data for all cause death extracted from the DECLARE-TIMI-58 trial. When the correct data for the DECLARE-TIMI-58 trial are used, the network estimated odds ratio for all cause death for a sodium-glucose cotransporter protein-2 (SGLT-2) inhibitor compared to placebo is 0.85 (95% confidence interval 0.79 to 0.92). The corrected network estimate for the odds of all cause mortality with an SGLT-2 inhibitor versus a glucagon-like peptide-1 (GLP-1) receptor agonist is 0.95 (95% confidence interval 0.86 to 1.06). A data entry error occurred that was not identified by double checking.
This correction changes the finding in the original publication (that SGLT-2 inhibitors reduced all cause mortality compared with GLP-1 receptor agonists) to: SGLT-2 inhibitors and GLP-1 receptors had similar effects on all cause mortality (moderate to high certainty evidence). The absolute number of deaths and evidence certainty did not change substantively. The article will be updated in due course. The recommendation in the associated guideline (BMJ 2021;373, doi:10.1136/bmj.n2331) will be reviewed.