ment area and most treatment is provided by general practitioners (GPs). In a series of papers since 1989, Rutz, Rihmer, and colleagues (Rihmer et al., 1995; Rutz, 2001; Rutz et al., 1989a; Rutz et al., 1989b) reported that educating the Gotland GPs about depression recognition increased the use of antidepressants and lowered suicide rates by 60 percent (see also Chapter 8).
Serotonin reuptake inhibitors (SSRIs) are used to treat depressive symptoms in the affective disorders as well as for symptom relief for those who have other diagnoses, or do not meet the criteria for the major affective disorders. SSRIs have gained great popularity in recent years, with the number of prescriptions increasing both in the United States and in other western nations (Isacsson, 2000; Lawrenson et al., 2000; Sclar et al., 1998). Although the SSRIs reduce depressive symptoms, their potency in reducing suicide is uncertain.
Verkes et al. (1998) found that patients with personality disorders and brief depression, but not major depression, had fewer suicide attempts when treated with paroxetine as compared with placebo. On the other hand, most studies failed to find statistically significant differences in suicide or suicidal behavior with SSRI treatments. Leon et al. (1999) followed 185 patients treated with fluoxetine (from among 643 patients as part of the NIMH Collaborative Depression Study). Using a mixed effects survival analysis, they found a decreased risk of suicide attempts and completions in the fluoxetine group, but this decrease did not achieve statistical significance, perhaps because the patients given fluoxetine were more severely ill than the comparison group before treatment. On the other hand, three meta-analyses failed to show effects of the SSRIs on suicide. Two (Khan et al., 2001; Khan et al., 2000) assessed FDA trials for efficacy and found that the major SSRI antidepressants were not significantly different than placebo with respect to suicides. Another meta-analysis of 17 clinical trials (Beasley et al., 1991) indicated that fluoxetine may reduce suicidal ideation but was not significantly different from either placebo or the tricyclic antidepressants in reducing suicides or attempts.
Several factors may enter into the interpretation of these results. Hirschfeld (2000) pointed out that these studies were time-limited. In addition, they mostly attempted to screen out those at risk for suicide. In most of the clinical studies, the base rate of suicide attempts was too low to determine effectively whether the antidepressant medications reduced the number of suicide attempts or suicides in comparison with placebo (Khan et al., 2001; Khan et al., 2000; Letizia et al., 1996; Montgomery et al., 1994; Tollefson et al., 1993). To a large extent, the low base rate of suicidal