populations. Dr. Comtois noted that it is important to know what patients are actually receiving via health services in order to assess opportunities for improved outcomes. In addition, to develop prevention efforts that can be applied in real-life settings and have efficacy outside of the research setting, the intervention’s setting needs to be considered.

Dr. Comtois reported that most of the studies were based on epidemiological data, primarily from Europe where more comprehensive public data bases are kept. As with the treatment efficacy literature, she reported the body of data to be woefully small and fraught with methodological problems. Often the “treatment-as-usual” condition was not measured. This is a particular problem in studies of outpatient services. Where treatment is described, there are vast variations in treatment services across and within countries and over time.

Suicide and Mental Illness. There is a high but not complete overlap between suicide and mental illness. Dr. Comtois questioned if “treating the psychiatric disorder is not only necessary, but also sufficient to reduce suicide.” In some studies, according to Dr. Comtois, “there is a close relationship between a reduction in the symptoms and a reduction in suicidality, but that is not universal.” The workshop attendees discussed studies showing a reduction in suicidality without a reduction in depression or vice versa.

The workshop attendees also discussed pharmacological studies showing significant reductions in suicide, even when depressive symptoms are not similarly reduced. Dr. Jamison described a recent meta-analysis on the efficacy of lithium. They found an over 8-fold reduction in suicide risk with prophylactic lithium treatment. Other studies of antipsychotic medications have also found significant reductions in suicide, according to Dr. Comtois. However, not all psychiatric medications are efficacious in reducing suicide, as indicated by some recent meta-analyses of anti-depressants. The lack of efficacy is again confounded by the exclusion of those at high risk for suicide.

Behavioral Interventions. Dr. Comtois reported variability in the behavioral intervention literature. Some interventions reported significant reductions in suicide, others reported trends, and still other studies reported no effect. She underscored the difficulty in comparing studies because of methodological differences across studies, inconsistencies in nomenclature, and exclusion of people at high risk.

Dr. Comtois described the only study reporting a significant reduction in completed suicide. Patients who refused treatment after a suicide attempt were entered into a study in which the “experimental” group received letters (contact) while the “control” group did not (non-contact). Dr. Comtois reinforced the importance of the non-demanding quality of this intervention. People in the contact group received a letter once per month for the first year, every 2 months in the second year, and one every third month for years 3 through 5. After the fifth year, the letters were discontinued. The patients were followed for 15 years. Dr. Comtois described that at year 2 the contact group had significantly reduced suicide rates compared to the non-contact group. This difference continued until year 14, at which point the rates in both groups were similar. Dr. Comtois posed the question: would continuing contact past the 5 years continue to reduced suicide rates?

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement