AAGP Response to Request for Information (RFI): A Call to Identify Key Methodological Roadblocks and Propose New Paradigms in Suicide Prevention Research

Published Wednesday, April 18, 2012

April 18, 2012                                                                                                                                                                  

Attention: RFI on Suicide Prevention
National Institute of Mental Health
6001 Executive Boulevard, Room 7136 MSC 9635
Bethesda, MD 20892-9635

Dear Colleagues:

The American Association for Geriatric Psychiatry (AAGP) is the largest organization for geriatric mental health. The AAGP is submitting these comments in response to the call for “Request for Information (RFI): A Call to Identify Key Methodological Roadblocks and Propose New Paradigms in Suicide Prevention Research - Notice Number: NOT-MH-12-017.”  AAGP, established in 1978, is a membership association of nearly 2000 geriatric psychiatrists, geriatricians, nurses, family physicians, psychologists, neurologists, social workers, pharmacists and other health professionals interested in mental health and older adults. AAGP promotes the mental health and well-being of older people through professional education, public advocacy, and support of career development for clinicians, educators, and researchers in geriatric psychiatry and mental health. 

To realize the NIMH’s goal of reducing suicide, we make these recommendations:

Candidate Paradigms:

1. Care transition models for at-risk elderly in both mental health and general medical sectors: Much of preventable suicide in older adults is among individuals already under medical care and/or psychiatric care yet falling through the cracks at key transition points.  Older patients who die by suicide are far more likely to be seen by a primary care or specialty medical provider than by a mental health provider, therefore practitioners without mental health training should be considered in theoretical models of suicide and suicide prevention.  Specifically, physical illness, notably cancer, lung disease, and renal disease increase suicide risk in older adults. Suicides in patients with physical illness typically occur early in the course of treatment. For example, in patients with end stage renal disease, risk is highest within 3 months of dialysis initiation, and in cancer, risk is highest within 12 months of diagnosis. Primary care or specialty physicians may under-appreciate their patients’ mental health needs, not inquire about suicide, or not know how to respond to disclosures of suicide.  For medical specialists, suicide prevention research should address both 1) the “informal” care transitions which occur at the time of diagnosis, with emphasis on illnesses with poor prognosis and 2) formal care transitions, as described below.  Within the mental health system, significant attention is directed to suicide prevention at the time of illness diagnosis; however, opportunities for more effective suicide prevention include enhancing illness detection, bolstering care maintenance and improving care transitions.  Specifically, only 3% of depressed suicide victims had been taking adequate antidepressants at the time of death (Isometsa, 1994), ‘lost to follow-up’ rates are as high as 20-25% (Li, 2005), and rates of completed suicide in the elderly peak in the three months following discharge from psychiatric units (Karvonen 2008).

Interventions that promote continuity of care, such as the Care Transitions Intervention (CTI), Project Re-Engineered Discharge (RED), and the Transitional Care Model (TCM), are being promoted by Centers for Medicare and Medicaid (CMS) in their Integrating Care for Populations and Communities” aim and the CMS Innovation Center’s “Community-based Care Transitions Program” solicitation for applications as means to reduce medical rehospitalizations. Similar interventions research has been proposed for depression, but little work has been definitively completed, as evidenced by the absence of specific mental health concerns in large-scale reports such as the 2007 Transitions of Care Consensus Policy Statement.  Testing novel care transition interventions specifically for suicide prevention should be done in both general medical and mental health transitions from emergency rooms or hospitals to community, home, or long-term care.  Such interventions could span a spectrum of intensity ranging from automated telephone or postcard follow-up for individuals at minimal risk to face-to-face in-home or in-office interviews for those most at risk, such as individuals with a history of a prior suicide attempt.  Future models could be aligned with the CTI coaching approach in which at-risk individuals are provided education regarding medication self-management and problem solving at the time of hospital discharge, or could rely on a transitional care clinician to “bridge the gap” after hospital or emergency room care, similar to the successes shown by the TCM.  As has been demonstrated for medical patients, efforts which go beyond automated follow-up are likely to improve the process of patient-provider decision-making and communication at critical junctures, resulting in enhanced discharge planning with better coordination between inpatient/emergency and outpatient teams. 

2. A model to identify decision processes that lead to suicide and associated cognitive vulnerabilities: Poorly understood heterogeneity is one of the key obstacles for biological studies of suicidal behavior and for the development of effective interventions. Rather than search for a common biological substrate of all suicidal behavior, research should focus on cognitive vulnerabilities (e.g., loss of cognitive control, impaired reward learning) and other vulnerabilities of the suicidal diatheses such as impulsive-aggression that lead to impaired decision processes and then relate these processes to neural network abnormalities.  There is accumulating evidence that cognitive functions play an important role in the suicidal diathesis. For example, low IQ and other indicators of low cognitive reserve are associated with death by suicide and suicide attempts.  Depressed patients who attempted suicide show impaired cognitive control and specific impairments in learning and decision-making compared to non-suicidal depressed patients (Keilp, 2001, Jollant, 2005, Dombrovski, 2010, Clark, 2011). Elderly suffer from degenerative and vascular brain changes that affect circuits critical for decision-making. Age-dependent decline in cognitive control and decision competence may contribute to the accumulation of stressful life events (Denburg, 2007) and to the decision to take one’s life.  Studies using functional neuroimaging in tandem with decisional or neuroeconomic tests could elucidate the neurobiology of suicide in older adults.  Collaborative initiatives such as conferences and pilot study proposals that get neuroeconomic, neuroimaging, and suicide researchers together would be an important first step.

3. Test models used in accident research. One potential model is Cognitive Reliability and Error Analysis Method (CREAM), which has been successfully applied in process industries, air traffic, and more recently, in the analysis of traffic accidents. It highlights cognitive “overload” as a tipping factor, and could be particularly relevant in understanding late life suicidal behavior given reduced cognitive reserve in this age group.

Methodology Roadblocks:

1. Underrepresentation of older adults in suicide research.  Older adults are a large and fast-growing group at highest suicide risk, but they are underrepresented or excluded in too many studies of mental illness, substance abuse, and suicide risk, and in many efforts to implement suicide risk strategies.  Those aged 80+ are the fastest-growing age group with the highest suicide rates (among men) yet we know extremely little about this group.  Research must include older adults and have specific plans for including them in sufficiently high numbers.  For example, public service announcements must include media that reach older adults, particularly men aged 80+.  Studies must reach the homebound elderly, who are both hard to reach by usual means and at high risk for suicide.  There has been very little research on problematic alcohol/substance use – including medication misuse -- and suicide in older adults.  This is a “gathering storm” given the much higher use rates in the baby-boomer generation; a practical problem is the different presentation of substance abuse in this age group, and lack of insight or clarity into what constitutes misuse.  Similarly, the elevated risk of suicide in anxious older adults has received almost no study. 

2.  Development of measures that validly differentiate thoughts of death as a normal phenomenon of aging from thoughts of death and suicide that reflect psychopathology and warrant intervention. Unlike at other points in the life course, thoughts of death may reflect a normal developmental process in an older adult, stimulated by a range of common individual (illness, disability) and interpersonal (bereavement, isolation) stressors (Bartels, 2002, Scocco, 2001, Szanto, 1996). Our understanding of what are "normal" thoughts of death, vs. ideation at-risk for suicide, is crude. We need better, more nuanced and valid tools to make these distinctions.  Suicidal ideation is a risk factor for completed suicide. Thoughts of death are more common than thoughts of suicide in older adults. Studies indicate some overlap between death and suicidal ideation; that is, for some individuals thoughts of death may carry increased risk for suicide. For many older adults, however, thoughts of death are a normal aspect of aging. Currently we lack the tools with which to distinguish when thoughts of death are normal (not an indicator of increased risk for suicide) and when they are a warning sign that the individual is in danger of harming himself. Indicated preventive interventions can only be cost effective when the high risk group they are designed to treat can be reliably differentiated from others for whom the intervention is unnecessary or ineffective. Absent tools to make that distinction reliably and validly, we cannot target preventive interventions with sufficient accuracy to those for whom they are indicated and, therefore, most likely to respond. Under such circumstances, the intervention is far less likely to be cost effective (far greater number needed to treat).   Therefore, we need scale development and rigorous psychometric testing using samples of older adults with normal and pathological death and suicidal ideation.

3. Social connectedness and its decline in aging: Social connections are key to suicide risk in older adults and to its mitigation. Yet a recent review on social factors and suicidal behavior in older adults found only 16 papers across 14 samples (Fassberg, 2012).  We need to articulate and test a model of how social factors at the individual, interpersonal, and community levels elicit and mitigate suicidal behavior. Doing so would require meetings and/or proposals that spur collaboration between researchers with expertise at these various levels.

Sincerely yours,

Paul Kirwin, MD

Contributors to this statement:

Martha Bruce, Ph.D., M.P.H, Cornell University
Yeates Conwell, M.D., University of Rochester 
Christine deVries, Chief Executive Officer, AAGP 
Paul Duberstein, M.D., University of Rochester 
Gary Epstein-Lubow, M.D., Brown University 
Paul Kirwin, M.D., Yale University & AAGP President 
Helen Lavretsky, M.D., University of California Los Angeles 
Eric Lenze, M.D., Washington University & AAGP Research Chair 
Katalin Szanto, M.D., University of Pittsburgh 
Margda Waern, M.D., University of Gothenberg