Two months ago, I wrote a post about a New Yorker article that reported that 67% of the depressed patients in the STAR*D trial "recovered." As I noted in that post, the 67% figure was a highly exaggerated number. Only 51% of the 3,671 patients who entered the trial ever remitted, even for a short period. Furthermore, only about 20% of the patients remitted and then reported to STAR*D investigators, at some point during a 12-month follow-up period, that they were still doing well.

But this left an obvious question, one that I hadn't been able to find an answer to in the published STAR*D reports. How many of the 3,671 people who entered the trial remitted and then stayed well and in the trial throughout the entire 12-month follow-up? That number would provide a documented long-term recovery rate for patients in the trial.

A few days ago, Allan Leventhal sent me a 2009 article he coauthored with David Antonuccio, and in it, they successfully identified this number (finding it in a confusing graphic I hadn't been able to decipher.) In their computations, they relied on STAR*D reports that told of 4,041 initial participants (3,671 was the number of "enrolled" patients counted in the analysis of drug-remission rates), and then they came to this bottom-line conclusion about the documented long-term recovery rate:

"Although the study's reports make no mention of this outcome, their data show that after a year of continuation treatment following remission, of the 4,041 patients who entered the program only 108 (3%) had a sustained remission — all the other patients either dropped out or relapsed. Yet STAR*D's authors and the NIMH have publicized the study as showing a 67% success rate for antidepressants."