It's rather shocking to someone who's not familiar with the inner workings of the pharma industry and medicine in general.
In particular in psychiatry, we have working theories about how the various agents work to achieve anti-depressive or anti-psychotic effects, which are taken as the gospel or proven fact. And then those models turn out absolutely wrong. And we have to face the fact we're prescribing drugs to sick people we have no clue how and why they work. And whether they work at all, often.
It's especially the case when a given drug must be taken for months before any effect is seen. It's essentially impossible to control for other variables over MONTHS of someone's life. You can easily convince them they feel slightly better now because they've been months on end taking Zoloft. They may've easily been better if they hadn't taken it.