It is a privilege to be with you today. My testimony here is an extension of my status, first, as a concerned citizen of this country, secondly as a Georgian, and thirdly as a professional who has invested countless hours studying the problem of suicide – not just amongst Veterans, but as an issue that transcends that distinguished status. I am a clinical psychologist and a researcher who has had the good fortune of collaborating, in some capacity, with VA researchers and clinical providers for the past six years. Prior to my current role as Director of Psychological Services with WellStar Health System, I was a post-doctoral fellow and clinical researcher with the VISN 19 Mental Illness Research, Education, & Clinical Center (MIRECC) in Denver, Colorado. The VISN 19 MIRECC is the nexus for clinical and translational research pertaining to Veterans and suicide. It is through my work there with that MIRECC’s director, Dr. Lisa Brenner, and a cohort of exceptionally bright and impassioned colleagues and that I understand, somewhat, the problem of suicide among our nation’s Veterans.
It is outside the scope of my discussion today to review the epidemiological factors pertaining to suicide and Veterans. Rather, it is my hope that, by reviewing the processes we have implemented at WellStar, a 5-hospital system in the metro-Atlanta area, we might shed light on how a clearly-defined and systematic effort, implemented with buy-in from leadership has helped to provide some safeguards regarding the management of patients presenting in our Emergency Departments and our inpatient psychiatric facility. To speak with you today affords not just the opportunity to describe our practices related to suicide prevention, but also as a reminder that clinically-relevant policies and procedures should exist as an extension of best clinical practices, and that these practices should be aligned with core organizational values.
I should note a priori that there are some key aspects of our work that should be clarified before we proceed. First, little of what we have implemented at WellStar to date was developed by our system. What success we have had is because we have implemented the most innovative suicide prevention processes to which we are informed. It is important that participants today understand that many of those processes and techniques were developed by (or in collaboration with) researchers at the Department of Veterans Affairs, and specifically the VISN 19 MIRECC. I have tremendous respect for what the VA is attempting to do to manage the problem of Veteran suicide. I also respect the enormity of that problem, and the inherent difficulty faced by VA administrators to implement the programs and processes that have been designed.
MANAGING ACUTE SUICIDE RISK
For 15 or more years, suicide has consistently ranked as one of The Joint Commission's five most frequently reported sentinel events (Joint Commission, 2010). There is well-documented evidence that suicide risk peaks in the days and weeks that immediately follow discharge from the Emergency Department (ED) and/or inpatient psychiatric care. Despite this understanding, the emergency management of suicide risk is often compromised due to the disconnect between standard hospital procedures and evidence-based, clinical practices (Knesper, 2010).
Accentuating the problem is the fact that those individuals most competent to provide suicide assessment and treatment (i.e. mental health clinicians) are often “not adequately trained to provide proper assessment, treatment, and management of suicidal patients” (U.S. Department of Health and Human Services, 2001, p. 79). Even when adequately trained, evidence suggests that a subset of suicidal patients are inclined to deny suicidal intent or minimize risk just prior to a suicide attempt (Knesper, 2010). Imagine, for example, that a patient arrives in a primary care physician’s office. He is a Veteran of the Army, and he presents as somewhat (and uncharacteristically) agitated. During the office visit, he casually mentions that his wife left him last week and his 13-year-old retriever died. Shaking his head, he says vaguely, “It’s just not worth it.” The astute physician alertly asks, “What do you mean?” The Veteran shakes his head again, “Nothing, forget it.” What now? To take the patient’s words at face-value (i.e. “Nothing, forget it”) can lead to problematic outcomes. What is the physician to do in this precarious situation, particularly given the research? In one study, Luoma, Martin, and Pearson (2002) indicated that up to 45% of adults who died by suicide visited with their primary care physician in the month preceding their death. This is anxiety-provoking for the physician. It is no less so for mental health providers. Although recommendations have been made to enhance mental health clinician’s competencies by offering education about suicide risk and warning signs, inadequate training remains a significant obstacle in many systems.
Apart from training deficits, a second problem, noted by former Secretary of the VA, Dr. James B. Peake, via the “Blue Ribbon Work Group on Suicide Prevention in the Veterans Population (2008), is the absence of uniform definitions of the terms (i.e. a common nomenclature or classification system) “suicide” and “suicide attempt.” Although seemingly simplistic, the failure to operationalize terms related to self-directed violence (SDV) has led to significant problems – impeding epidemiological efforts, and in clinical practice, making it difficult to provide evidence-based assessment and treatment of individuals who are high risk for suicide. In response to the former Secretary’s charge, the VISN 19 MIRECC reviewed existing literature regarding suicide risk, as well as prior SDV nomenclatures and classification efforts. In 2008, the VISN 19 MIRECC staff began working on a clinical nomenclature, and in 2009 members of this research team joined efforts with the Centers for Disease Control and Prevention (CDC) to finalize the Self-Directed Violence Classification System (SDVCS). The SDVCS was piloted at two VAMCs (Brenner et al., 2011), and in April 2010, the Department of Veterans Affairs announced its uniform adoption across the VA system. Although formally “adopted,” the implementation of it remains problematic.
Whereas there are some problems for which few solutions seem to exist, compelling are the notions that educating clinicians systematically and uniformly, improving screening and assessment processes, and initiating and motivating safety plan adherence at the time of an ED visit might culminate with a decrease in suicidal self-directed violence (SDV) behaviors post-discharge from hospitals and emergency receiving facilities. It was with this goal in mind that WellStar designed and implemented a 3-phase, stratified process to: (1) improve our clinicians’ understanding of suicide risk; (2) enhance our risk assessment screening tools; and (3) implement uniform safety planning for Behavioral Health patients admitted to our EDs.
A 5-hospital system, located in the northern suburbs of Atlanta, WellStar has 4 acute care Emergency Departments. One of these hospitals, WellStar Kennestone, is the busiest (by volume of patients seen) in the state; the second largest, WellStar Cobb, is the 6th busiest in the state. Emergency Rooms in all four acute care facilities are staffed with Masters-prepared Behavioral Health assessors, who in Fiscal Year 2013 conducted 8,556 Behavioral Health assessments, or an average of 713 each month. With mild fluctuations year to year, this rate of assessment has remained virtually unchanged for the past 5 years. In addition to the Behavioral Health assessment team, we have a 32-bed inpatient psychiatric facility, a team of consult-liaison and outpatient psychiatrists, and a small cadre of clinical psychologists in outpatient practice.
Phase I (PI), Clinician Trainings: In response to a system-wide quality improvement (QI) initiative to incorporate evidence-based practice into the assessment and treatment of individuals considered potentially high-risk for suicide, all Behavioral Health assessors were required to attend a 60-minute suicide prevention training. The training focused primarily on Thomas Joiner’s (2005) Interpersonal Theory of Suicide, as well as aspects of Fluid Vulnerability Theory (Jobes & Nelson, 2006), helping assessors understand some of the primary conceptual suicide risk domains. It was also in the context of Phase I training that clinicians were exposed to the SDVCS.
Phase II (PII), Safety Planning: In a subsequent 60-minute workshop, clinicians were trained to conduct collaborative Safety Planning for all ED Behavioral Health patients. Prior to that roll-out a small group of Super Users was formulated to ensure that within the group of assessors there were clinicians trained to help steer and champion the implementation effort. The Safety Planning approach was modeled after Stanley & Brown’s (2008) Safety Plan Treatment Manual, a method that in contrast to “No Harm Contracts” involves face-to-face collaboration with patients, during which time they are provided with specific strategies of “what to do” in the event they begin experiencing an acute crisis in the future. WellStar Behavioral Health providers were trained to develop Safety Plans with patients, walking systematically through the following steps:
1. Identifying warning signs (e.g. thoughts, feelings, behavioral risk factors, etc.)
2. Developing active coping strategies
3. Engaging with other people
4. Reaching out to social supports
5. Contacting an agency in the event the crisis has not passed
Patients are instructed to work through the steps sequentially, and are provided a hard copy to keep with them at all times. Providers process ways in which the patient will ensure that their environment is safe (e.g. eliminating access to firearms). Finally, the patient identifies a person in their life who they feel can help them manage future crises, and efforts are made to share the Safety Plan with them.
Phase III (PIII), Suicide Risk Screening: Because patient ratings of risk out-perform clinician’s ratings (Jobes et al., 2007), in a third and final 60-minute training, Behavioral Health Assessors were trained in the administration, scoring, and interpretation of the Habituation & Acute Risk Measure - Short (HARM-S) (Breshears, 2011). The HARM-S is a brief and modified version of the HARM, which is a theoretically-derived, multi-dimensional measure of suicide risk. The HARM-S is a 4-scale measure that includes a scale intended to detect potentially-concealed risk, as well as Habituation to Self-Directed Violence (SDV), Agitated Depression, and Suicidal Intent. It is intended not as a replacement, but as a supplement to the face-to-face clinical interview. Following training on the tool, mental health assessors were instructed to administer the HARM-S to patients and incorporate the data into their clinical assessments with ED Behavioral Health patients. The HARM-S incorporates a risk stratification protocol, informed by the literature, which is intended as an aid in treatment decision-making and triage to inpatient and/or outpatient treatment settings.
In the interest of full transparency, I know that in the three years since implementation of enhanced suicide training of our ED Assessors compliance with these best practices has waned somewhat largely due to changes in leadership. This affirms the need for strong leaderhip to maintain consistent, strong support for adherence to best practices in care for our patients suffering with mental illness.
CONCLUSIONS & RECOMMENDATIONS
While it is beyond the scope of reason to assume that all suicides can be prevented, there do exist viable processes and tools that can help to mitigate risk for Veterans and non-Veterans alike. Below I offer a handful of recommendations to participants of this committee, but I wish to do so humbly. I have great appreciation and respect for our Veterans and for the men and women who serve our Veterans. I also can empathize with the challenges of navigating the implementation barriers that exist in the context of a changing system. I do think that there are some opportunities going forward to improve care for our patients, and some of these could potentially be of help in managing and mitigating suicide risk going forward. I also want to take note of the various initiatives that have already been undertaken by the VA system. Although I am sure this list is not exhaustive, I want to specifically note the following:
1. Management of Acute Crisis: The suicidal crisis can last only moments or can persist for many days and weeks. I applaud the efforts of the VA to establish and maintain the 24-hour Suicide Hotline for Veterans. This effort alone has undoubtedly helped many Veterans manage problematic situations and keep them from becoming impossible.
2. Implementation of the SDVCS: The lack of a common language to differentiate suicidal from similar, yet different phenomenon, remains problematic. Although the DVA’s announcement of the SDVCS occurred 3 years ago, many facilities continue to struggle with its implementation. I would encourage administration to gauge the extent to which implementation barriers continue to exist in the system, and how these barriers can be eliminated.
3. Safety Planning & Stepped Care Models: Men with a mood disorder are 218.8 times more likely to die by suicide in the week immediately following inpatient psychiatric treatment (Qin & Nordentoft, 2005). They are 126 times as likely to die by suicide at 8-30 days post-discharge, 56 times more likely at 1-6 months, and still 30 times more likely from 7-12 months. This odds ratio continues to decline over time, but at 5 years post-hospitalization, they are still 5 times more likely to die by suicide than the non-psychiatrically hospitalized male. The gender trend for women is comparable (the rates are even slightly higher). Pinkis and Burgess (1998) reported that up to 41% of individuals who died by suicide were treated in an inpatient psychiatric facility within the year prior to death. The authors further noted that up to 9% who died by suicide had been psychiatrically hospitalized within 24 hours of death. The numbers are compelling. Data such as these speaks to the importance of developing and implementing safety plan measures, particularly in the days and weeks post-hospital discharge.
4. Care Letters: Already efforts have been made by the VA to send care letters to Veterans following a hospital discharge. I applaud these efforts and am hopeful they continue.
5. Suicide Consultation: The training and education of VA clinicians regarding suicide risk is essential. A novel approach is evidenced by the VA’s recent launch of the Suicide Risk Management Consultation Program at (846) 948-7880. This is an innovative program that offers clinicians the opportunity to consult with experienced suicidologists remotely to manage difficult cases, receive answers to complex questions, etc. I want to encourage my colleagues in Atlanta to make use of this service.
Table of Contents