It's possible for more than one entity to behave badly.

The original editorial by the NEJM criticizing the VIGOR trial was just
odd -- it suggested the authors should have done something in the
analysis that really made no sense (reanalyze the data based on events
after the completion of the trial period). Then, it turned out that
those running the trial really had done something completely
inappropriate in the analysis (had different periods of time for
counting cardiac and GI adverse events), but it wasn't what the NEJM
editorial first mentioned. According to the Wall Street Journal piece
mentioned by another poster, the editors had not even spotted what the
VIGOR authors had done with the analysis until after the editorial, so,
if true, it's not just that the editorial was somewhat unclear.

In this pass, the NEJM did something I've never heard of before. They
published a correction to an article without the authors' agreement.
Normally, you only publish a correction under the authors' names. The
NEJM could conceivably have retracted the article about APPROVE, or they
could have written an editorial saying that the original article was
interpreted incorrectly and, had they known, they would have been
unwilling to publish it without changes, but it seems really strange to
issue a correction to an article written by someone else, even if
published in your own journal. Has anyone ever seen a major journal do
this? Perhaps it's more common than I realize.