[arc-in-space]

My point was that the data doesn't seem point at suicidal ideation in any particular way. Adverse event reports contain literally every outcome in connection with every drug, you need a strong trend to claim that you aren't just seeing report noise. (But you're right that a newly discovered side-effect would be more interesting than a well known one.)

[addicted]

That’s not how much research works. You never have data at the outset. You have to have anecdotal information and/or sometimes even just an unfounded hypothesis before you can collect useful data to prove/disprove the hypothesis.

This would fall into the anecdotal side of things, and should hopefully drive more rigorous research on whether any such concerns exist.

In reality, even if the research is proven, the only likely change is screening out a few people who are at an elevated risk of suicidal ideation and/or better monitoring, or even prescribing regular therapy along with the medication (which might be a good idea anyways).