Why We Need to Enhance Suicide Postvention

Abstract

Suicide prevention efforts are increasing to enhance capabilities and better understand risk factors and etiologies. Postvention, or how clinicians manage the postsuicide aftermath, strengthens suicide prevention, destigmatizes the tragedy, operationalizes the confusing aftermath, and promotes caregiver recovery. However, studies regarding its efficacy are minimal. The Psychopathology Committee of the Group for the Advancement for Psychiatry surveyed a convenience sample of psychiatrists to better understand postvention activities. Ninety psychiatrists completed the survey; they were predominantly men (72%) with an average of 24.6 years of experience (SD, 16.7 years). Most had contact with the patient’s family within 6 months of the suicide, and most psychiatrists sought some form of support. Few psychiatrists used a suicide postvention procedure or toolkit (9%). No psychiatrists stopped clinical practice after a patient suicide, although 10% stopped accepting patients they deemed at risk of suicide. Postvention efforts, therefore, should be improved to better address survivor care.

Tragically, suicide is becoming more commonplace in the United States, now at its highest level in the past 30 years and the 10th leading cause of death among Americans (Tavernise, 2016). Almost half of all individuals who die by suicide were seen by a health care provider in the month before their death—and a quarter of them had a behavioral health encounter (Ahmedani et al., 2014). Attempted and completed suicides impact clinician-caregivers, and there is a robust literature regarding the significant emotional effects upon clinical practice (Seeman, 2015). To better address the widespread increase in suicide rates, resources for suicide prevention have been bolstered. These bench-to-bedside efforts range from attempts to identify neurobiological risk and protective markers to Zero Suicide campaigns and social media app development. There are considerably fewer resources, guidelines, or rigorous studies as to how clinicians manage the postsuicide aftermath.

The term “postvention” is attributed to Edwin Shneidman in 1968 at the first conference of the American Association of Suicidology. Shneidman defined postvention as follows: interventions to address the care of bereaved survivors, caregivers, and health care providers; to destigmatize the tragedy of suicide and to assist with the recovering process; and to serve as a secondary prevention effort to minimize the risk of subsequent suicides due to complicated grief, contagion, or unresolved trauma (Shneidman, 1975). Postvention efforts also enhance suicide prevention by providing behavioral health, psychosocial, spiritual, and public health services to the survivors (Aguirre and Slater, 2010). Each suicide, noted Dr Thomas Insel, former director of the NIMH, created 11 victims: the person who died and the 10 caregivers/ family/friends who are at risk themselves (Insel, 2014).

Postvention efforts are proactive and may help prevent further suicides (Andriessen, 2009; Jordan, 2008). However, for the past five decades, postvention efforts have been limited largely to a focus on specific groups (adolescents or co-employees) because of concerns regarding contagion, which may not be generalizable across different settings. Concerns regarding contagion and increased suicide risk behaviors among children and adolescents after the suicide of a peer, for example, is well documented, but standardized approaches are lacking due to the heterogeneity of such tragic incidents (Cerel et al., 2005). Consequently, postvention efforts tend toward specific approaches such as “first responder,” “managing liability,” and “preventing contagion” to name a few (Erlich and Psychopathology Committee of the Group for the Advancement of Psychiatry, 2016).

Moreover, there has been a dearth of rigorous testing of the effectiveness of postvention programs. Evidence of “what works” is mostly anecdotal. Many of the immediate and long-term psychological treatments (i.e., trauma-informed therapies, psychological first aid, bereavement counseling, among others) are effective for some individuals after someone has died by suicide, but lack the rigor (and specialization) that may be needed for postvention interventions. According to a recent RAND report for the US Department of Defense, there is a dearth of scientific evidence on how to best respond to suicide, manage survivors' grief, and monitor caregivers' own risk for suicide and/or self-harm. In addition, there is no clear understanding of which resource to use, when to deploy it, and with whom (Ramchand et al., 2015).

There are several thoughtful, recovery-oriented, and reportedly effective postvention resources to assist survivors and their families coping with a suicide. Programs with the best evidence use structured procedures and follow-up with survivors that actively engage them during periods of care transition (Bryan et al., 2014; Luxton et al., 2013). There are several examples: Texas' postvention toolkit, evidence-based practices endorsed by the Suicide Prevention Resource Center, and the clinician- and peer-led outreaches of the LOSS Program (Loving Outreach to Survivors of Suicide Program). These postvention programs are manualized, have toolkits, and are easily accessible. Many incorporate practical interventions and best practices available to lay and professional caregivers alike. In addition to the goal of decreasing suicide contagion risk, they foster relationships among survivors and build community resources for both community and professional survivors.

To better understand psychiatrists' current postsuicide/postvention behaviors after their client commits suicides and to deepen understanding how psychiatrists practice afterward, the Psychopathology Committee of the Group for the Advancement of Psychiatry (GAP) surveyed a convenience sample of psychiatrists to understand the prevalence and content of postvention interventions. The survey had three aims: first, for those clinicians who experienced the suicide of a patient, to elicit their responses including determining what suicide postvention interventions were utilized; second, to better gauge what realms of practice were most challenging to the clinician in the aftermath of a suicide; third, to better assess what postvention interventions could help address this complex clinical situation that is often traumatizing to clinicians-caregivers. We were particularly interested in whether psychiatrists used available postvention toolkits defined as a formalized set of protocols and resources to assist the survivors of a suicide with references, guidelines, and expert recommendations. In addition, we undertook a second Web-based convenience sample to compare and confirm the results of the psychiatrists' survey.

METHODS

Overview

We administered a Postvention Survey at the GAP fall 2015 meeting held November 12 to 14, 2015, in West Harrison, NY. The survey was designed by the GAP Psychopathology Committee. All participants were psychiatrists attending the GAP meeting. The survey took between 10 and 15 minutes to complete. This survey was deemed exempt from institutional review board (IRB) review in that no individually identifiable information was collected. Moreover, no subjects received any reimbursement for completing this survey.

Recruitment

Individuals were recruited from the semiannual meeting of the GAP, a think tank of psychiatrists seeking to inform and educate mental health professionals, policy makers, and the public at large on current issues in psychiatry.

Survey Development

The questionnaire was developed using a modified Delphi process by an expert committee composed of members of the Psychopathology Committee of GAP based on the most recent literature about psychiatrists' response to patient suicide. Domains assessed by the survey included demographics, types of clinical contact with surviving caregivers, barriers to utilization of postvention responses, support(s) sought in the aftermath of patient suicide, and how clinical practice was altered in response to patient suicide. The survey consisted of both closed-ended (either binary questions or on a 5-point Likert scale) and open-ended questions to allow for respondents to share additional information. This survey was reviewed by an institutional research ethics committee, which found that because this study did not collect any identifiable data from the participants or need informed consent, it was determined that this study was excluded from IRB review and oversight.

Data Analysis

Responses to paper surveys were entered into Excel and analyzed using STATA version 12.1. Summary descriptive statistics for each domain listed above were first calculated. The subsequent analyses included psychiatrists who experienced at least one suicide of a patient (n = 59). For each domain, differences in responses based on years of experience, sex of psychiatrist, and number of patient suicides were evaluated using the analysis of variance statistical test.

RESULTS

Description of the Sample

A total of 90 (47%) of the 190 psychiatrists attending the meeting completed the survey. Seventy-two percent of the respondents were men with a mean age of 56.8 years (SD, 15.9 years). Clinicians had an average of 24.6 years of experience (SD, 16.7 years). Thirty-four percent (n = 31) of the sample had experienced no patient suicides, 38% (n = 34) had one patient suicide, 25% (n = 22) had two to three patient suicides, and 3% (n = 3) had four or more suicides. With 90 responses and more than a third experiencing more than one suicide, we asked psychiatrists to report the setting where the suicide occurred. Respondents were representative of diverse overlapping settings where the suicide(s) occurred: private outpatient practice (36%), academic outpatient setting (35%), academic inpatient setting (28%), public outpatient setting (22%), public inpatient setting (10%), and private inpatient setting (10%). Psychiatrists with more years in practice were more likely to have more patient suicides (χ² = 17.3081, df = 8, p < 0.03). There were no other significant relationships detected among sex of psychiatrist, number of suicides, setting, and years in practice.

Contact by Psychiatrists With Suicide Survivors

Most psychiatrists (n = 49, 83.1%) had some form of contact with the deceased patient's family, friends, or caregivers within 6 months after the suicide as shown in Table 1. The most common type of contact was to call the family and friends (n = 47, 79.7%), whereas the least common was use of a suicide prevention toolkit (n = 5, 9.1%). Those with more experience were more likely to call family or friends (p < 0.04). Male psychiatrists were more likely to offer condolences (p < 0.03) and attend the funeral (p < 0.03) than female psychiatrists. Individuals with more patient suicides were more likely to use a postvention procedure or toolkit, but the number was small (p < 0.01). Otherwise, psychiatrist sex, experience, and number of patient suicides were not associated with type of contact.

TABLE 1.

Types of Contact Psychiatrists Had With Suicide Survivors

	Psychiatrists Experiencing ≥1 Patient Suicide n (%)	Sex			Years of Experience				No. Suicides
		Male n (%)	Female n (%)	χ2, df, p	<5 yrs n (%)	6–20 yrs n (%)	>20 yrs n (%)	χ2, df, p	1 Suicide n (%)	2–3 Suicides n (%)	4–5 Suicides n (%)	>5 Suicides n (%)	χ2, df, p
Any contact	49 (83.1)	35 (85)	11 (73)	1.08, 1, 0.298	3 (43)	6 (86)	40 (89)	9.16, 2, 0.01	28 (82)	18 (82)	1 (100)	2 (100)	0.648, 3, 0.885
Called family or friends	47 (79.7)	34 (82.9)	10 (66.7)	1.72, 1, 0.189	3 (42.9)	6 (85.7)	38 (84.4)	6.65, 2, 0.036	26 (76.5)	18 (81.8)	1 (100)	2 (100)	1.04, 3, 0.791
Offered condolences	41 (71.9)	32 (80)	7 (50)	4.65, 1, 0.031	3 (43)	6 (86)	32 (74)	3.72, 2, 0.156	22 (69)	17 (77)	0 (0)	2 (100)	3.81, 3, 0.282
Apologized	12 (21.1)	9 (23)	3 (21)	0.007, 1, 0.934	1 (14)	1 (14)	10 (23)	0.511, 3, 0.774	6 (19)	6 (27)	0 (0)	0 (0)	1.41, 3, 0.702
Offered formal therapeutic session	23 (39.0)	17 (41)	4 (27)	1.0, 1, 0.311	1 (14)	3 (43)	19 (42)	2.04, 2, 0.361	13 (38)	7 (32)	1 (100)	2 (100)	5.18, 3, 0.159
Allowed family or friends to vent	23 (39.0)	15 (38)	7 (50)	0.67, 1, 0.413	1 (14)	3 (43)	19 (44)	2.26, 2, 0.324	13 (41)	7 (32)	1 (100)	2 (100)	5.10, 3, 0.165
Suggested family or friend join support group	14 (24.6)	12 (30)	2 (14)	1.33, 1, 0.248	2 (29)	0 (0)	12 (28)	2.60, 2, 0.273	10 (31)	3 (14)	1 (100)	0 (0)	5.91, 3, 0.116
Attended patient’s funeral	14 (23.7)	14 (34)	0 (0)	6.8, 1, 0.009	0 (0)	1 (14)	13 (29)	3.2, 2, 0.203	8 (24)	5 (23)	0 (0)	1 (50)	1.09, 3, 0.780
Used a suicide postvention procedure or toolkit	5 (9.1)	4 (11)	0 (0)	1.6, 1, 0.206	0 (0)	1 (14)	4 (10)	0.95, 2, 0.622	1 (3)	2 (10)	1 (100)	1 (50)	15.3, 3, 0.002

Support Sought by Psychiatrists After the Suicide of a Patient

Most psychiatrists sought some form of support after a patient suicide (Table 2). Most reviewed their notes (n = 47, 81%), whereas only 7.1% (n = 4) consulted an attorney preemptively and 3.6% (n = 2) consulted an attorney due to initiated litigation. The psychiatrist's sex, experience, and number of patient suicides were not associated with types of support sought.

TABLE 2.

Types of Support Sought by Psychiatrists After Patient Suicide

Types of Support Sought	na (%)
Reviewed their notes	47 (81.0)
Obtained informal supervision with a colleague	42 (71.2)
Discussed with a family member or friend	38 (70.4)
Obtained formal supervision with a colleague	12 (21.1)
Consulted the risk assessment department	18 (31.6)
Consulted an attorney preemptively	4 (7.1)
Consulted an attorney because of initiated litigation	2 (3.6)

Unless otherwise indicated, nonsignificant for sex of psychiatrist, experience of psychiatrists, or number of patient suicides.

^an varies from 53 to 57 due to blank responses.

Barriers

Participants were asked about concerns that might impede contact with the caregivers of the deceased patient (the number responding varied from 52 to 57 of the 60 respondents). Most felt that neither fear of litigation (n = 11, 19.3%) nor issues of confidentiality or HIPAA regulations (n = 9, 17.3%) were barriers to interacting with survivors. Only a small number of psychiatrists (n = 10, 17.5%) felt their own emotional difficulties interfered with their ability to deal with the suicide of a patient. Most psychiatrists (n = 50, 82.5%) were motivated by their concern for the well-being of the family or friends of the patient—and contacting the family was not deemed a barrier per se. None of these barriers were significant when analyzed by sex of psychiatrist, years of experience, or number of patient suicides.

Changes in Clinical Practice Post Suicide

Half of the surveyed psychiatrists reported no change in their clinical practice after a patient suicide. Of those who reported a change in their clinical practice (n = 27, 50.9%), a third sought increased supervision or review, a quarter of psychiatrists began using formal measures to assess suicidality, and only 9.1% began using any available suicide postvention toolkit or formalized set of protocols. No psychiatrists stopped clinical practice as a result of the patient suicide, although 9.8% (n = 5) reported they stopped accepting new patients they deemed as at risk of committing suicide (women were more likely than men, but the small number of respondents preclude significance).

Follow-up Survey Results

In a follow-up to this study, we were able to collect additional data through a Web-based convenience sample survey of 282 clinicians. The survey was distributed through the Psych Times Web site (Erlich and Psychopathology Committee of the Group for the Advancement of Psychiatry, 2016). As this was an online questionnaire, response rates are difficult to determine because of how the response rate is calculated without knowing the entire “denominator,” that is, the readership of Psychiatric Times. For the purpose of this sample, we report percentages of survey respondents (n = 283). The sample consisted of a broader variety of mental health professionals, including psychiatrists (28%), psychologists (25%), physician assistants (16%), nurse practitioners (13%), social workers (10%), and counselors (8%). As with the GAP survey, most clinicians had some type of contact with the decreased patient's family, friends, or caregivers (n = 102, 69%). Again, the least common type of contact used among this diverse group of providers was the use of a suicide prevention toolkit (n = 10, 7%). A minority, however, began using a suicide postvention toolkit or formalized set of protocols after a patient suicide (n = 13, 8%). No clinicians reported stopping clinical practice after the suicide of a patient.

DISCUSSION

This preliminary study is an important step in helping to understand how psychiatrists respond in the aftermath of patient suicide. It is one of the few studies to report psychiatrists' responses after a patient suicide. Prior investigations have focused on how select populations cope with the aftermath of a suicide: trainees (Cazares et al., 2015), copatients (Seeman, 2015), consultation psychiatrists (Takahashi et al., 2011), and health care providers in school and/or employment environments (Cox et al., 2016; HEMHA, 2014). To our knowledge, there are no published studies that assess a cross-section of psychiatrists, from academic to community settings, from recent graduate to later career practitioners regarding awareness of postvention and limits to its implementation. The findings of the Web-based convenience sample survey largely confirmed the results from the current survey and reinforced the importance of postvention efforts—and the dearth of a standardized approach.

Regardless of the scope of practice, the duration of the relationship, and the experience of the physician, the suicide of a patient is a difficult—and even traumatic—time for the psychiatrist. This study indicated that psychiatrists sought out some type of support, either informally from colleagues and/or friends and family or more formally via supervision—and this positively correlated with the number of suicides experienced in practice. Although it is not surprising that psychiatrists sought support postsuicide, we highlight four findings. First, there was an inverse relationship between years of experience and supervision sought postsuicide, although the number of suicides experienced was not a significant factor. Second, legal concerns were not considered an impediment to contact with surviving caregivers, even though such concerns are often cited as a one of the main reasons for litigation in behavioral health care. Third, there were sex differences that should be further studied. Fourth, although psychiatrists seek support after a suicide, less than 1 in 10 implemented an evidence-based intervention such as a postvention toolkit. This small number highlights the realities that in our field, postvention interventions are sparse, variable, and underutilized, even in this motivated group (Ramchand et al., 2015). As the surveyed sample reflected psychiatrists who tend to be more experienced and academic, the lack of use of protocols is important, as these would be the practitioners one would expect to be most aware of these interventions. It is possible that because of the sample’s median age, suicide postvention kits were not available when some of these suicides occurred.

To better understand what was most challenging to psychiatrists in the aftermath of suicide, we assessed changes in ongoing practice. Individuals continued practicing during and after the trauma of a patient suicide; however, half became more cautious in their care. They sought additional supervision and use more formal measures for suicide assessment. That 10% stopped accepting patients at risk of committing suicide (women more than men per our findings) speaks to physician burden and the risk of burnout. As with many concerns in behavioral health, there is also worry that the stigma of suicide upon its survivors may hinder the application of postvention interventions (Andriessen and Krysinska, 2012).

This study has important implications for psychiatric residency training. A national survey examining suicide education in psychiatry residency programs indicated some formalized teaching, but a lack of standardization to the approach and, importantly, a dearth of preparation for the management of suicide’s tragic aftermath (Melton and Coverdale, 2009). Introducing a standardized approach to postvention to residents during their training is critical and opportune because most psychiatrists will experience a suicide during the course of their careers. Learning a best practice regarding postvention in training, such as a postvention toolkit, can provide the combination of supervision, structured support, and access to hospital and outpatient resources that is critical to the well-being of the clinician and survivors alike (Ellis et al., 1998).

Postvention interventions need to be calibrated to best suit the clinician’s scope of a practice, experience, and resources (and obstacles) to implementation. Setting is also critical in regard to garnering resources. Public settings tend to have forums to provide postvention support for clinicians and family members. For example, sustained postvention effort would include allowing for clinical time to reach out to caregivers, providing resources for the clinician to learn from the event, and insuring that caregivers have the option to follow-up with behavioral health care after the suicide. Individuals in private practice setting have fewer system resources upon which to draw upon.

There are limitations to this study. First, the convenience sample was composed of invitees to an academic psychiatric meeting; accordingly, the membership differs from psychiatrists in general and limits the generalizability of the results. Second, the sample size is small. Third, assessing physicians’ perceptions (i.e., barriers and challenges to postvention) depends on the surveyed physicians’ perceptions of them, and it is possible that what is identified in the survey may not accurately reflect how it is observed in practice. Finally, the survey asked the participant to answer questions retrospectively—and how long ago a patient suicide occurred differs among the participants and accurate reflection of the event may also affect the response. A larger study following community practice prospectively is necessary to better understand both postsuicide practice and continuing clinical barriers.

CONCLUSION

There are no 100% effective treatments for preventing suicide. How psychiatrists manage the postsuicide aftermath, destigmatize the tragedy, and promote caregiver recovery are all part of the postvention process. Data regarding the efficacy of postvention activities and programs are often lacking due to its lack of widespread use and heterogeneity when it is deployed. This survey supports the need for further research to better assess what postvention interventions need to be enhanced and integrated into everyday practice in an effort to mitigate the risk to caregivers.