This meeting is the second in a series of five that will occur annually. The first meeting held one year ago focused on suicide in youth and young adults. Taken together, this series is designed to facilitate development of scientific consensus concerning the best practices for suicide prevention in specific segments of the population: youth and younger adults, older adults, men in their middle years, and women. The fifth meeting will be devoted to a summary assessment of the field’s progress to date.

The specific aims of the conference series are to:

1. Develop a consensus among clinical and public health researchers, and among key community and political leaders, including governmental agencies and non-governmental organizations, regarding those factors known to predict increased risk, or show protective benefit, for suicide or serious suicide attempts – across the age spectrum, within distinct population groups, and for each gender;
2. Systematically propose research to address areas of insufficient knowledge regarding risk factors specific to critical populations (i.e., those accounting for the greatest amount of burden due to morbidity and mortality), while examining as well those factors that protect similarly vulnerable individuals;
3. Utilize evidence-based approaches to review scientific data regarding the effectiveness of proposed or tested methods to reduce suicide, including population-oriented and high-risk approaches, as well as suicidal behaviors and their attendant morbidity;
4. Foster consensus on the best methods for population-oriented universal prevention, and preventive-clinical interventions to treat those individuals and groups who have been identified as being at highest risk;
5. Aggressively disseminate research findings and workshop recommendations to the scientific community, policy makers, and the public through a series of review publications and the use of electronic information and networking strategies.

B. Background of the Year 2 meeting: Suicide Prevention in Later Life

Suicide is a major public health problem in the United States, where 30,000 people take their own lives each year. Although the largest number of suicides occur among people in mid-life, elders have higher rates of suicide than other segments of the population. Older men in particular bear increased risk. Older adults are the fastest growing segment of the population, a trend that will continue in coming decades as the post-WWII "baby boom" enters late life. Further, the baby boom cohort has carried higher rates of suicide throughout life than previous or later birth cohorts at the same ages. This combination of demographic and cohort effects suggests that there will be an increase in the absolute number of suicides by senior citizens in the years to come.

Whereas the clinical approach to suicide prevention emphasizes the history, health conditions, and warning signs that precede suicide in an individual, the public health approach focuses on identifying patterns of suicide and suicidal behavior throughout a group or population. The public health prevention research model is constituted by five integrated steps – surveillance to define the problem, the identification of causal risk factors, development and preliminary testing of interventions, the implementation of interventions on a larger scale, and the evaluation of their effect. The application of the public health prevention model to the problem of suicide is reflected in the National Strategy for Suicide Prevention (NSSP): Goals and Objectives for Action. Published in 2001, the NSSP is intended to serve as a road map for moving suicide prevention efforts forward in the United States, guiding the coordinated efforts of an array of stakeholders in the public and private sectors. Next steps must include the development of more detailed plans for specific activities that correspond to each objective. Because the biological, psychological, and social characteristics and contexts of population subgroups differ, the NSSP’s objectives and the specific activities derived from the NSSP must be modified to address their particular needs. The effort this year focused on the needs of older adults.

C. Sponsors and Contributors

Support for the Year 2 meeting was provided by funding from NIMH, NIAAA, NIDA, NINR, CDC, and SAMHSA. Additional contributions were made by Pfizer Pharmaceuticals and Forest Laboratories.

D. Organization and Process

Meeting participants represented a broad array of perspectives and stakeholders integral to the consensus process. They included representatives of government agencies and non-governmental organizations, researchers, primary and mental health care providers, clergy, and others. Meeting activities included both plenary addresses and breakout groups. Introductory plenary sessions provided a common knowledge base regarding the epidemiology of, and risk and protective factors for, suicidal behaviors in later life, definitions of the public health model and terms, and preliminary results of a national survey of elder suicide prevention initiatives.

Breakout groups were then formed a) to consider the strength of the evidence for risk and protective factors and for the effectiveness of implemented interventions, and b) to review and refine the proposed elder-specific NSSP goals and objectives.

A final series of plenary sessions was then held to reach large group consensus on a) modifications to the proposed elder-specific goals and objectives, b) priorities for implementation of preventive interventions for later life suicide, and c) priorities for future research.

E. Goals and Outcomes

The meeting had three primary goals. Each is listed below with a description of the issues raised and consensus reached by participants.

1) Our first goal was to examine the evidence base for current knowledge regarding elder suicide at each step of the prevention research spectrum, including surveillance, identifying causes, translation of research to practice, and program evaluation.

• Surveillance: Surveillance for completed suicide in later life is well established. Vital statistics reporting is systematic nationwide and provides important data with which to monitor trends and inform policy and research priorities. However, some experts believe that elder suicide may be underreported due in part to social stigma leading families and officials to misclassify cases, and in part to the lack of broadly accepted, rigorously standardized definitions and reporting requirements for suicide. Also, elders are a heterogeneous group. Insignificant attention has been given to the collection and reporting of surveillance data for suicide in specific subsegments of the older adult population defined by age (e.g. the oldest-old), geographic region (including rural vs. urban), and race and ethnic background. As suicide in older adults is a relatively uncommon event, the small number available for epidemiological analysis poses a special challenge.

Another serious deficiency is the lack of a nationwide or even comprehensive regional surveillance system for non-fatal suicidal behaviors in older adults. Creation of a reliable and valid system for tracking suicide attempts among elders would provide critical information to augment our understanding of the epidemiology of this important form of preventable morbidity. As well, it would provide the infrastructure for targeting and evaluating the outcome of preventive interventions, and facilitate research to define the similarities and differences between populations of elders who attempt and those who complete suicide.

• Risk factor definition: Much progress has been made in recent years in the definition of factors that place older adults at increased risk for suicide. Evidence from a recent series of carefully conducted retrospective case control (psychological autopsy) studies was reviewed. It is clear that psychiatric illness (especially major mood disorders), a prior history of suicidal behaviors, physical illness and functional impairment, social isolation and stressful life events (especially physical illness and family or interpersonal discord) represent risk factors for suicide in later life. The knowledge base, however, remains limited in a number of areas. In particular, there have been no large-scale prospective cohort studies conducted in older adults that have included careful documentation and tracking of putative risk factors and suicidal behaviors as predictors and outcomes of primary interest. The existing case control studies all lack sufficient power to examine adequately potential moderators of risk factors for late life suicide. Few data are available to explain how people with significant risk factor ‘burdens’ may be protected from suicide. Discerning such protective factors could inform the design of prevention programs. Finally, virtually no research has examined possible contributions of biological factors to suicide risk with increasing age.
• Implementation of preventive interventions: Despite limits to our knowledge, the existing evidence provides sufficient basis for the development of elder-specific suicide prevention strategies. However, very few programs have been implemented. A survey of initiatives in the U.S. identified 13 agencies or organizations reporting suicide prevention programs pertinent to elders. Further examination revealed that several were only in the planning stages, and of those in operation, only five specifically target older adults. None could be characterized as community wide or as offering a complete array of universal, selective, and indicated preventive interventions. The Link Plus-Life Crisis Line in St. Louis Missouri, the Center for Elderly Suicide Prevention in San Francisco California, and the Spokane Mental Health Center’s Elder Services Program in Spokane Washington are particularly well developed examples of promising approaches.
• Program evaluation: These three programs are also the only ones to have published program evaluation data. The virtual absence from the literature of outcomes assessment for late life suicide prevention programs is a significant limitation.
• Barriers to progress: A number of barriers were identified to the translation of pre-intervention research findings to the design, implementation, and evaluation of later life preventive interventions for suicide. These include lack of funding; the lack of researchers, clinicians, and other staff members with sufficient training in these areas; high staff turnover in those programs that have piloted programs; lack of resources (expertise, infrastructure and funding) for program evaluation; and poor coordination of the agencies involved.

2) Based on that evidence base, we next sought consensus on NSSP goals and objectives tailored specifically to address the challenges of suicide prevention in older adults. Breakout groups were charged with reviewing the current NSSP goals and objectives and adapting them as indicated by our current knowledge base to the more specific task of older suicide prevention. The NSSP goals and objectives were divided into universal prevention approaches, selective prevention, and indicated interventions. Each of three breakout groups then reviewed the goals, commenting and suggesting revisions to make them more applicable for this age group. These suggested revisions were in turn reviewed in a final plenary session to arrive at consensus goals and objectives, which are attached as Appendix 1.

3) Our third goal was to develop consensus on priorities for next steps in a) preventive intervention strategies and b) research.
The discussion of priorities for next steps took place in the final plenary session in two parts.

• Priorities for preventive intervention: The group first considered priorities for prevention programming, responding to this question: "Given what we do know about risk for, and prevention of, suicide in older adults, where should we put available prevention resources?"

  Discussion regarding priorities for prevention programming reflected first and foremost a sense of urgency that the general public, elders themselves, their families, and their clinical and community providers lack an adequate understanding and appreciation that suicidal behavior in older people is ordinarily a sign of illness and a preventable form of morbidity and mortality. There were repetitive calls and clear consensus for placing highest priority on the development of educational programs targeting each of these groups. Educational content should include aging, ageism, and life cycle developmental issues; the assessment and treatment of suicidal states in elders; the assessment and treatment of more proximal risk factors, including depression and substance abuse; and the support of healthy lifestyles, including physical activity and social connectedness. The approaches advocated included both standard curricula and the application of newer information system-based pedagogy. Use of mass media was advocated -- public education campaigns utilizing public service announcements and high profile spokespeople to reduce ageism and stigma regarding mental health care for older people, and supporting positive models of aging, including among those who are dependent on others for care. As well, media campaigns should target professional schools and postgraduate education for healthcare and social service providers at all levels. There was a consistent call as well for the establishment, where feasible, of accreditation processes and the delineation of required competencies in each of these content areas.

  A second area of clear consensus was on the need to establish partnerships between stakeholder groups so that prevention programs reflect the necessary degree of interdisciplinary collaboration. More specifically, participants called for the empowerment and incorporation of survivors in the process, and more explicit roles for representatives of faith communities, home care providers, social service professionals, and the primary care delivery system. A common theme, consistent with the emphasis on education programming, was that a broad range of people and groups throughout society should have an understanding of, investment in, and responsibility for suicide prevention in older people. Though not discussed explicitly during the meeting, there was an inherent recognition that the prevention of suicide and suicidal behavior is not recognized as a high priority for many major community action groups. Future efforts should be devoted to educating community stakeholders about the common risk factors that underlie suicide and other major social problems.

• Priorities for research: The second phase of the final plenary session addressed the question, "Given an infusion of funds for research, what priorities would you set for their use?" Participants stressed three areas. First, they pressed for the development of more and better tools for the study of suicide and its prevention, including screening measures and pertinent scales for use in subgroups of the elderly population (e.g., the medially ill). They called for an expansion of fundamental risk/protective factor research that included sufficient sample sizes to examine subgroup differences and more complex models of risk determinants. A major focus in this area, however, was on protective factors, addressing issues such as the impact of resiliency and the role of formal and informal helping networks for older adults. Participants expressed interest in means to measure and influence physician behavior, particularly as it relates to the recognition and treatment of depression in older adults. As well, they stressed the need for more extensive treatment research for late life mood disorders.

  Finally, there was clear consensus that conducting more extensive evaluation research was of high priority. Needs include the development of tools to evaluate current and proposed interventions, and resources and infrastructure for dissemination of research findings.

F. Summary

The meeting brought together a broad array of participants representative of the stakeholders central to later life suicide prevention programming. The different perspectives represented were not always easily reconciled. Nonetheless, common themes and consensus emerged. The process reinforced that our knowledge base concerning risk factors for suicide in later life is rapidly expanding. Additional research is needed to refine further our understanding of who is at risk and why, and thus inform the design of more effective preventive intervention strategies. However, even our current knowledge base is sufficient to form the basis of program design and implementation. Greatest emphasis should be placed on universal prevention strategies, stressing educational interventions, although their combination with selective and indicated strategies in complex interventions offers most hope for effecting significant reductions in late life suicide rates.