A 2004 outpatient study was done with patients suffering from minor depression who had much lower rates of suicidal ideation. These patients were split into double-blind groups, one being treated with Prozac and the other with the placebo, and they were checked up on for 12 weeks. Basically, they had those questionnaires that you get and they ask you about your suicidal ideation, rated on a scale of zero to four. And this paper analyzed the rates that people in the Prozac and placebo groups had increases of one or two points. They found that people in the placebo group had higher rates of increased suicidal ideation, but the amount was statistically insignificant. However, there was a wrinkle. There was no statistical significance in the difference in groups in terms of who had an increase of one or two points. But if you only included the people who had baseline suicidal ideation at the beginning of the study, people in the group treated with Prozac had significantly less of an increase. For patients without suicidal ideation before the study, 5.1% of the placebo group had an increase of two points, whereas the Prozac group was 3.5%. That's not that big of a difference. For patients who had initial suicidal ideation, 40.9% of the placebo group had increases of two points compared to 12.5% with the Prozac group.