Antidepressants, suicidality and rebound effect: evidence of similitude?