 Selective outcome reporting

Reporting a favorable prespecified outcome while omitting a less favorable one can give an exaggerated impression of consistency of evidence. Multiple primary outcomes should be summarized transparently even when some lack statistical significance. A short abstract can prioritize results, but should not hide a materially different primary result; absence of information does not by itself establish deliberate misconduct.