Crisis Issue 1, 2019 by Hogrefe

Crisis

Volume 40 / Number 1 / 2019

Editor-in-Chief Jane Pirkis Associate Editors Maria A. Oquendo Ella Arensman

The Journal of Crisis Intervention and Suicide Prevention

Published under the auspices of the International Association for Suicide Prevention (IASP)

Symposium Partners

Hilton Trinidad & Conference Centre

Crisis The Journal of Crisis Intervention and Suicide Prevention

Volume 40 / Number 1 / 2019 Published under the Auspices of the International Association for Suicide Prevention (IASP)

Editor-in-Chief

Jane Pirkis, Centre for Mental Health, Melbourne School of Population and Global Health, University of Melbourne, Melbourne 3010, VIC, Australia (Tel. +61 3 8344-0647, Fax +61 3 9348-1174, E-Mail j.pirkis@unimelb.edu.au)

Associate Editors

Maria A. Oquendo, MD, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA (Tel. +1 215 662-2818, Fax +1 215 662-6911, E-mail moquendo@pennmedicine.upenn.edu) Ella Arensman, PhD, National Suicide Research Foundation, Department of Epidemiology and Public Health, University College Cork,Western Gateway Building, Room 4.34,Western Road, Cork, Ireland (Tel. +353 21 420-5551 or +353 87 052-2284, E-mail earensman@ucc.ie)

Editorial Board

A. Apter, Tel Aviv, Israel R. Battegay, Basel, Switzerland M. P. Battin, Salt Lake City, UT, USA A. L. Beautrais, Christchurch, New Zealand A. L. Berman, Washington, DC, USA S. S. Canetto, Fort Collins, CO, USA J. Cutcliffe, Amherstview, ON, Canada D. De Leo, Mt. Gravatt, Australia O. Ekeberg, Oslo, Norway A. Erlangsen, Aarhus, Denmark M. Goldblatt, Boston, MA, USA D. Gunnell, Bristol, UK K. Hawton, Oxford, UK L. M. Hayes, Mansfield, MA, USA K. van Heeringen, Gent, Belgium H. Hjelmeland, Trondheim, Norway E. Isometsa, Helsinki, Finland G. Jessen, Odense, Denmark N. Kapur, Manchester, UK N. Kapusta, Vienna, Austria A. J. F. M. Kerkhof, Amsterdam, The Netherlands M. M. Khan, Karachi, Pakistan

Editor Emeritus

Diego De Leo

Past Editors-in-Chief

Raymond Battegay, Annette L. Beautrais, David C. Clark, John F. Connolly, Diego De Leo, Ad J. F. M. Kerkhof, Hermann Pohlmeier

Responsible Organization Publisher

The journal is published under the auspices of the International Association for Suicide Prevention (IASP) (IASP Central Administrative Office, National Centre for Suicide Research and Prevention, Sogsvannsveien 21, Building 12, N-0372 Oslo, Norway, Tel. +47 22 923715, E-mail admin@iasp.info, www.iasp.info).

Production

Juliane Munson, Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany Tel. +49 551 99950-0, Fax +49 551 99950-111, E-mail production@hogrefe.com

Subscriptions

Hogrefe Publishing, Herbert-Quandt-Str. 4, 37081 Göttingen, Germany Tel. +49 551 99950-900, Fax +49 551 99950-998

Advertising/Inserts

Marketing, Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany, Tel. +49 551 99950-0 Fax +49 551 99950-111, E-mail marketing@hogrefe.com

ISSN

ISSN-L 0227-5910, ISSN-Print 0227-5910, ISSN-Online 2151-2396

© 2019 Hogrefe Publishing. This journal as well as the individual contributions and illustrations contained within it are protected under international copyright law. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without prior written permission from the publisher. All rights, including translation rights, reserved.

K. Kõlves, Brisbane, Australia A. A. Leenaars, Windsor, Canada J. L. McIntosh, South Bend, IN, USA J. Mann, New York, NY, USA M. Marttunen, Helsinki, Finland L. Mehlum, Oslo, Norway B. L. Mishara, Montreal, Québec, Canada T. Niederkrotenthaler, Vienna, Austria S. Occhipinti, Nathan, Australia (Statistical Advisor) R. O’Connor, Glasgow, UK M. Phillips, Shanghai, China M. Pompili, Rome, Italy A. Preti, Cagliari, Italy M. Silverman, Chicago, IL, USA S. Stack, Detroit, MI, USA L. Vijayakumar, Chennai, India M. Voracek, Vienna, Austria D. Wasserman, Stockholm, Sweden E. J. de Wilde, Rotterdam, Netherlands P. Yip, Hong Kong SAR, China J. Zhang, Buffalo, NY, USA

Hogrefe Publishing GmbH, Merkelstr. 3, 37085 Göttingen, Germany Tel. +49 551 99950-0, Fax +49 551 99950-111, E-mail publishing@hogrefe.com,Web http://www.hogrefe.com North America: Hogrefe Publishing, 7 Bulfinch Place, Suite 202, Boston, MA 02114, USA, Tel. (866) 823-4726, Fax (617) 354-6875, E-mail customerservice@hogrefe.com, Web http://www.hogrefe.com

Published in 6 issues per annual volume.

Calendar year subscriptions only. Rates for 2019: Institutions – from US $360.00/€277.00 (print only; pricing Subscription prices for online access can be found in the journals catalog at hgf.io/journals2019); Individuals – US $179.00/€128.00 (print & online). Postage and handling – US $24.00/€18.00. Single copies – US $63.00/€49.50 + postage and handling. Payment

Payment may be made by check, international money order, or credit card, to Hogrefe Publishing, Merkelstr. 3, D-37085 Göttingen, Germany, or, for North American customers, to Hogrefe Publishing Corp., Journals Department, 7 Bulfinch Place, 2nd floor, Boston, MA 02114, USA.

Electronic Full Text

The full text of Crisis is available online at www.econtent.hogrefe.com

Abstracting Services

The journal is abstracted/indexed in Index Medicus/Medline, Current Contents/Social and Behavioral Sciences (CC/S&BS), Social Sciences Citation Index (SSCI), Social SciSearch, PsycINFO, PSYNDEX, Scopus, and EMCare. 2017 Impact Factor 1.583, 5-Year Impact Factor 2.119, Journal Citation Reports (Clarivate Analytics, 2018)

Crisis (2019), 40(1)

Contents Editorial

Sentinel Surveillance for Self-Harm: Existing Challenges and Opportunities for the Future Katrina Witt and Jo Robinson

Research Trends

Associations Between Suicide Risk Factors and Favorite Songs: Content Analysis and Cross-Sectional Study Benedikt Till, Michael Fraissler, Martin Voracek, Ulrich S. Tran, and Thomas Niederkrotenthaler

Suicide Prevention Training – Improving the Attitudes and Confidence of Rural Australian Health and Human Service Professionals Monika Ferguson, James Dollman, Martin Jones, Kathryn Cronin, Lynne James, Lee Martinez, and Nicholas Procter

A Comparison of Attitudes Toward Suicide Among Individuals With and Without Suicidal Thoughts and Suicide Attempts in South Korea Sang-Uk Lee, Mina Jeon, and Jong-Ik Park

Suicide on Instagram – Content Analysis of a German Suicide-Related Hashtag Florian Arendt

A 17-Year National Study of Prison Suicides in Belgium Louis Favril, Ciska Wittouck, Kurt Audenaert, and Freya Vander Laenen

Self-Harm and Suicide Coverage in Sri Lankan Newspapers: An Analysis of the Compliance With Recommended Guidelines Jane Brandt Sørensen, Melissa Pearson, Martin Wolf Andersen, Manjula Weerasinghe, Manjula Rathnaweera, D. G. Chathumini Rathnapala, Michael Eddleston, and Flemming Konradsen

Short Report

Is Suicidal Behavior in Mood Disorders Altered by Comorbid PTSD? René M. Lento, Amanda Carson-Wong, Jonathan D. Green, Christopher G. AhnAllen, and Phillip M. Kleespies

News, Announcements, and IASP

John Terry Maltsberger – Suicidologist Extraordinaire Mark J. Goldblatt, Elsa Ronningstam, Benjamin Herbstman, and Mark Schechter

Crisis (2019), 40(1)

Editorial Sentinel Surveillance for Self-Harm Existing Challenges and Opportunities for the Future Katrina Witt1,2 and Jo Robinson1 Orygen, the National Centre of Excellence in Youth Mental Health, Centre for Youth Mental Health, University of Melbourne, VIC, Australia 2 Turning Point, Eastern Health Clinical School, Monash University, Melbourne, VIC, Australia 1

Suicide is a leading cause of preventable death worldwide. Each year over 800,000 people take their own lives (World Health Organization, 2014). Self-harm behavior (which includes suicide attempt, nonsuicidal self-injury, and both suicidal and nonsuicidal self-poisoning) is significantly more common than suicide and is probably the greatest predictor of future suicide (Hawton, Zahl, & Weatherall, 2003). For these reasons, improving the monitoring of both suicide and self-harm is a key recommendation from the World Health Organization (World Health Organization, 2010). Although rates of suicide are generally well recorded, at least in high-income countries, via systems like the Office for National Statistics in England and Wales (Office of National Statistics, 2018), and the Australian Bureau of Statistics in Australia (Australian Bureau of Statistics, 2018), these are not without their limitations. In particular, they face issues in terms of their capacity to produce real-time data. However, data on emergency department presentations for self-harm are generally poorly recorded. As such, the establishment and ongoing maintenance of surveillance systems to monitor rates of self-harm represents a cornerstone of most countries’ suicide prevention strategies (Zalsman et al., 2017).

What Can Surveillance Systems Offer? Surveillance systems (also known as monitoring or sentinel systems) for self-harm can play an important role in suicide prevention by providing important epidemiological data on rates of, and risk factors for, self-harm presentations. But, when functioning well, they also have the potential to © 2019 Hogrefe Publishing

play a greater role by enabling data on all self-harm presentations to emergency departments within a specified geographic region, city, or country to be amalgamated in as close to real time as possible. In this way, surveillance systems for self-harm have the potential to operate similarly to other systems established to estimate the rate and spread of both chronic diseases and acute infectious diseases, at the population level. At present only a few such systems exist internationally. Arguably the four longest running dedicated surveillance systems for self-harm are the Multicentre Study of Self-Harm in England (Multicentre Study of Self-Harm in England, 2018), and the Hunter Area Toxicology Service in Australia (Whyte, Dawson, Buckley, Carter, & Levey, 1997), both of which have a regional focus, the Bristol Self-Harm Surveillance Register (Williams, 2015), which collects data city-wide, and the Irish National Self-Harm Registry (Perry et al., 2012), which collects data for the whole of the Republic of Ireland. Given that the emergency department is frequently the first point of contact with health-care services for people in crisis, data on presentations for self-harm and mental health more broadly in this setting are often used as a key performance indicator for mental health care and suicide prevention (Gill et al., 2017). To this end, all existing surveillance systems for self-harm primarily operate within emergency department settings to gather episode-level data on emergency department presentations for self-harm. Existing surveillance systems have had a number of positive impacts. First, data from these systems have led to the identification of clinically relevant subgroups of patients who have engaged in self-harm, enabling prevention initiatives to be better targeted toward these groups (Barrett, Griffin, Corcoran, O’Mahony, & Arensman, 2018; Cooper et al., 2010). These systems have also allowed rates of self-harm presentations to be tracked over time (Geulayov Crisis (2019), 40(1), 1–6 https://doi.org/10.1027/0227-5910/a000583

et al., 2016; Perry et al., 2012). This, in turn, has provided the data necessary to evaluate the impact of differences in the clinical management of self-harm between systems (Carroll et al., 2016), and following changes to clinical practice guidelines (Kapur et al., 2008; Kapur et al., 2013; Whyte, Francis, & Dawson, 2007), on rates of self-harm repetition. Geographic data from these systems can also inform decisions about clinical service provision and allocation of resources to areas of highest need (Corcoran, Arensman, & Perry, 2007). Additionally, these systems can have the benefit of providing the data necessary to identify emergent methods of self-harm in a timely manner (Cairns, Schaffer, Ryan, Pearson, & Buckley, 2018; Daly et al., 2018), such that appropriate preventive measures can be implemented (Hawton et al., 2011; Hawton, Fagg, Simkin, Bale, & Bond, 1997). Data from these surveillance systems may also facilitate the identification of clusters of self-harm at the population level (Hawton, Lascelles, & Ferrey, 2015).

What Are the Key Challenges in Such Systems? Despite their benefits, challenges exist with such systems. Developing a real-time, robust surveillance system for something as complex as self-harm is inevitably going to present issues for many countries. To assist stakeholders with this, the World Health Organization released a Practice Manual for Establishing and Maintaining Surveillance Systems of Suicide Attempts and Self-Harm (World Health Organization, 2016). To better understand these challenges, and how they have been overcome in practice, we surveyed the lead investigators of the four surveillance systems cited earlier, and asked them about the challenges that they have experienced during the establishment and ongoing delivery of these systems. Their responses centered on the following themes: developing robust case ascertainment protocols; stakeholder engagement strategies; supporting research innovation; and securing ongoing funding.

Developing and Refining Robust Case Ascertainment Protocols Developing consistent and reliable approaches to case ascertainment is fundamental to the robustness of any surveillance system for self-harm. Yet this is often hampered by a number of factors, not least of which is the continued lack of standard, internationally accepted, definitions of Crisis (2019), 40(1), 1–6

Editorial

what behavior(s) constitute self-harm (Silverman & De Leo, 2016). The World Health Organization Practice Manual for Establishing and Maintaining Surveillance Systems of Suicide Attempts and Self-Harm recommends all cases of self-harm leading to hospital presentation, irrespective of the level of suicidal intent and/or method used, should be eligible for inclusion in these systems (World Health Organization, 2016). However, while the Bristol Self-Harm Surveillance Register, Irish National Self-Harm Registry, and the Multicentre Study of Self-Harm in England include all cases of self-harm irrespective of method used, the Hunter Area Toxicology Service is embedded within a hospital-based toxicology service. Therefore, only cases of self-poisoning are captured by this latter system. Additionally, all existing systems currently include cases presenting to emergency departments, not only those leading to hospital admission. The World Health Organization further recommends the use of International Classification of Diseases (ICD) codes to achieve consistency and uniformity in the identification of self-harm cases (World Health Organization, 2016). However, administrative datasets, which often form the backbone of case ascertainment protocols within these systems, vary in completeness in ICD coding, particularly for external cause injury codes, which are generally not required for billing and reimbursement purposes (Hedegaard et al., 2018). As external cause codes are essential for identifying self-harm cases, systems that rely exclusively on ICD coding are likely to significantly underestimate the true number of self-harm cases (Hedegaard et al., 2018). For this reason, most existing surveillance systems combine data from multiple sources to ascertain relevant cases for inclusion. However, the data sources used varies widely between systems. In the Bristol Self-Harm Surveillance Register and the Irish National Self-Harm Registry, for example, potential cases are identified principally through electronic searches of emergency department records. In the Multicentre Study of Self-Harm in England, on the other hand, cases are largely identified through psychosocial assessment and psychiatric records, supplemented by electronic searches of emergency department records for non-assessed patients. The data collection process for the Multicentre Study of Self-Harm in England is therefore quite labor intensive, which, in turn, can lead to longer lags between data collection and dissemination of outcomes, meaning that the system is less able to provide data in real time. Machine-learning approaches may have a role here. Pilot studies that have employed machine-learning techniques in order to identify suicide attempt cases from electronic emergency department records have found that they significantly improve the timeliness and quality of manualized case ascertainment approaches (Faverjon & Berezowski, 2018; Metzeger et al., 2016). © 2019 Hogrefe Publishing

Editorial

Additionally, only one of these four systems, the Bristol Self-Harm Surveillance Register, has reported data on the accuracy of case ascertainment. This system successfully identifies between 98.0% and 99.0% of all cases of selfharm resulting in presentation to emergency departments in its catchment area (Knipe & Gunnell, 2016). However, efforts to improve one aspect of data collection, for example, timeliness, may impact on the ability of the surveillance system to accurately identify relevant cases, as might changes to emergency department data management systems and/or coding practices. Therefore, periodic audits to investigate the completeness of case ascertainment protocols within these services are required. Routine reporting of sensitivity, specificity, positive predictive values, and negative predictive values is also recommended (European Centre for Disease Prevention and Control, 2014). Ensuring case ascertainment protocols are consistent across time is therefore crucial to the robustness of the system. To this end, most systems have specific protocols around auditing of case ascertainment protocols to ensure these systems reliably and consistently identify self-harm cases over the longer term. However, the frequency of these audits differs widely. Within the Hunter Area Toxicology Service, trained staff conduct bi-annual audits to check the accuracy of case identification within the system. Within the Irish National Self-Harm Registry, staff periodically cross-code data in order to check accuracy of case ascertainment and to provide information on the inter-rater reliability of the case ascertainment protocol. The Bristol Self-Harm Surveillance Register is audited for case ascertainment every 2–3 years by comparing identified cases against electronic searching of emergency department records. However, systems generally leverage staff other than project personnel, such as junior psychiatry trainees and/or research assistants, to lead on auditing of case ascertainment protocols as there is rarely specific funding to support these so-called ancillary activities.

on suicide prevention, which, in turn, helps to increase awareness of the research and other outputs generated by this system. Demonstrated impacts on suicide prevention policy and service provision, including informing the recommendations contained in the World Health Organization’s Practice Manual for Establishing and Maintaining Surveillance Systems of Suicide Attempts and Self-Harm, have also contributed to the success of the Irish National Self-Harm Registry. To further engage stakeholders, the World Health Organization recommends all self-harm surveillance systems publish annual reports (World Health Organization, 2016). All existing systems with the exception of the Hunter Area Toxicology Service currently produce publicly available annual reports, with the latter system moving to annualized reporting in 2019. In addition to these annual reports, both the Irish National Self-Harm Registry and Multicentre Study of Self-Harm in England provide more frequent progress updates, usually at 6-month intervals, while the Bristol Self-Harm Surveillance Register prepares reports on an ad hoc basis as requested.

Stakeholder Engagement Strategies

This, in turn, can make it more difficult for these systems to open up new areas of research and/or to pursue new funding opportunities as they arise. This tension may stem from a mismatch in expectations between lead investigators and other stakeholders, particularly funders and policy-makers, as to the overall purpose and function of the self-harm surveillance system:

The involvement of stakeholders from the outset is important in both establishing and maintaining these systems. One mechanism for this is regular reporting on the outputs from the system to stakeholders such as funders and policy- makers and to the services themselves as part of ongoing service improvement efforts. Indeed, the provision of information relevant to service provision has been key to the ongoing success of the Multicentre Study of Self-Harm in England. For example, data from the Multicentre Study of Self-Harm in England are presented each year to the National Suicide Prevention Strategy Group, which helps to keep a focus on self-harm prevention within the broader discourse © 2019 Hogrefe Publishing

Supporting Research Innovation Balancing reporting requirements with research innovation can also be a challenge. All the aforementioned systems regularly publish data in peer-reviewed academic journals to further support the dissemination of outcomes from their respective system. However, there is often tension in the relative weight assigned to reports versus peer-reviewed academic papers, with research outputs not necessarily being one of the performance indicators prioritized by funding agencies: Reports are preferred by the funders instead of papers published in peer-reviewed journals. However, external funding agencies prioritize peer-reviewed papers over reports. (surveillance system lead investigator, Nov 1, 2018)

The value and outcomes of [the system] are often understated [with key] stakeholders… primarily [focused] on counting numbers as opposed to understanding the wider benefits … (surveillance system lead investigator, Nov 1, 2018)

Despite this, innovation in research has often been integral to the ongoing success of these surveillance systems: Crisis (2019), 40(1), 1–6

Editorial

Awareness of the research and other outputs from the work helps to maintain support for it. (surveillance system lead investigator, Oct 21, 2018)

Tensions between supporting research innovation alongside stakeholder reporting requirements can also arise depending on whether data are maintained by the project team or by the hospital and/or other service. Data custodians can be wary of sharing data other than for approved purposes owing to fears that patient privacy could be impacted. Clearly communicating expectations during the establishment phase of a surveillance system, for example, by developing an overarching Memorandum of Understanding, may help researchers and stakeholders to strike an appropriate balance between dissemination of process-oriented outcomes while ensuring research innovation is also fostered. This would clearly outline which party, or parties, retains ultimate responsibility over the data, the uses for which access to the data are approved, and who the approved users of the data are.

Securing Ongoing Funding While the World Health Organization recommends that, where possible, systems should receive funding from government sources (World Health Organization, 2016), most existing systems typically rely on multiple sources of limited-term funding. The Multicentre Study of Self-Harm in England has managed to achieve financial support from the Department of Health, with additional funding received from university sources on a year-on-year basis, but other systems rely on year-on-year funding or ad hoc funding from a variety of sources, including: government funding (e.g., the Irish National Self-Harm Registry), competitive research funding (e.g., the Bristol Self-Harm Surveillance Register and the Hunter Area Toxicology Service), ad hoc university funding (e.g., the Bristol Self-Harm Surveillance Register and the Hunter Area Toxicology Service), and by charging organizations for access to the data derived from the system (e.g., the Bristol Self-Harm Surveillance Register). A primary challenge threatening the ongoing continuation of all existing surveillance systems for self-harm is therefore a lack of funding security year to year: Year-on-year funding is challenging for a national surveillance system … this makes it difficult to plan registry related work longer than a year. (surveillance system lead investigator, Nov 1, 2018)

Related to funding insecurities is the issue of consistent staffing. The World Health Organization recommends that, at a minimum, any successful self-harm surveillance system requires: (1) a member of senior staff to act as the Crisis (2019), 40(1), 1–6

overall system manager; (2) a data collector (either a member of hospital staff or, alternatively, a trained research staff member with experience working with health information systems); and (3) several research staff, including a statistician, to prepare reports and to undertake analyses (World Health Organization, 2016). However, funding of academic staff in particular is not always prioritized by stakeholders: … academic support for [the] register (i.e., for analysis, report writing) is entirely done out of goodwill. (surveillance system lead investigator, Oct 21, 2018)

With regard to data collection, while a number of existing systems have leveraged hospital-based clinical staff to predominately undertake data collection work (Caldera, Herrera, Renberg, & Kullgren, 2004; Griffin et al., 2015; Hawton et al., 2006), these staff often experience difficulty in balancing clinical duties against data collection duties (Hawton et al., 2006). Hiring additional support personnel may therefore be desirable to assist with data collection. In the Multicentre Study of Self-Harm in England, for example, support staff are available to ensure data collection is completed by clinical staff and to provide clinical staff with assistance and advice to ensure data collection is undertaken in a timely manner. It is notable that in one study of a pilot self-harm surveillance system in India, a lack of administrative staff to assist with data collection was identified as the single largest barrier to the successful full implementation of the system (Rajendra et al., 2015). Staffing for database management is also important to ensure the ongoing success of these systems. Both the Bristol Self-Harm Surveillance Register (Knipe & Gunnell, 2016) and Hunter Area Toxicology Service (Pillans, Page, Ilango, Kashchuk, & Isbister, 2017), for example, manage data within customized and fully relational Microsoft Access databases. Expertise in database management and computer programming is therefore required to ensure the successful maintenance of these two systems in particular. A lack of ongoing funding also makes it difficult to plan staffing requirements into the future: Funding has been renewed on an annual basis, which makes for uncertainty, especially for staff. (surveillance system lead investigator, Oct 21, 2018)

Staff retention, however, is often key to the ongoing success of these systems. In the Multicentre Study of Self-Harm in England staff retention to key positions has been excellent; in the 42 years of the system’s existence, only four people have occupied the role of primary data collector with the incumbent occupying this role for over 15 years. Staff retention, particularly in senior leadership roles, has also been excellent within the Hunter Area Toxicology Service. © 2019 Hogrefe Publishing

Editorial

Future Directions

References

Notwithstanding these challenges, data from these four surveillance systems have had a number of positive impacts. These have included describing the sociodemographic, clinical, and treatment characteristics of people who present to hospital with self-harm, monitoring changes over time, and importantly, informing the development, delivery, and evaluation of treatment and policy initiatives. These data can also act as a lead indicator of progress toward suicide prevention targets (Caine, Reed, Hindman, & Quinlan, 2017). Guidance from the World Health Organization also means that the value of surveillance systems for self-harm and suicide prevention is being increasingly recognized. However, these types of systems are still few and far between, and even the existing systems still experience challenges with regard to the provision of accurate real-time data (Arensman, 2017). As such, their full potential is as yet untapped. Our own country has not had a reliable nationwide system for collecting data on self-harm presentations. Fortunately, this is beginning to change. For example, there has been recent investment from the Department of Health and Human Services in establishing a network of self-harm surveillance system in a number of emergency departments across the state of Victoria. This system is in its infancy but will ultimately provide accurate information about rates of, and risk factors for, self-harm presentations to hospital, as well as allowing for the impact of clinical and policy change to be evaluated. Additionally, these systems currently collect only episode-level data on self-harm presentations. At present, therefore, the ability to track individuals within these datasets over time is limited, which, in turn, has implications for the types of research questions that can be answered using data derived from these systems. Despite record levels of investment in clinical-based treatment services over the past decade (Snowdon, 2016), suicide rates continue to rise in a number of counties. With limited further investment, routine administrative reporting systems can be upgraded to provide real-time information on rates of self-harm. This would allow for the provision of up-to-date epidemiological data, and for the timely identification of changes in the patterns of suicide and self-harm. However, if working optimally, the benefits of these systems would not stop here. They would also have the capacity to act as barometers of the success of national suicide prevention strategies (Hawton et al., 2006), and to inform real-time change and intervention at both the population and individual levels. We would argue this is where their preventative potential truly lies.

Arensman, E. (2017). Suicide prevention in an international context: Progress and challenges. Crisis, 38, 1–6. https://doi.org/ 10.1027/0227-5910/a000461 Australian Bureau of Statistics. (2018). Causes of death, Australia. Retrieved from http://www.abs.gov.au/ausstats/abs@.nsf/mf/ 3303.0 Barrett, P., Griffin, E., Corcoran, P., O’Mahony, M., & Arensman, E. (2018). Self-harm among the homeless population in Ireland: A national registry-based study of incidence and assocaited factors. Journal of Affective Disorders, 229, 523–531. https://doi. org/10.1016/j.jad.2017.12.040 Caine, E., Reed, J., Hindman, J., & Quinlan, K. (2017). Comprehensive, integrated approaches to suicide prevention: Practical guidance. Injury Prevention, 24, i38–45. https://doi.org/10.1136/ injuryprev-2017-042366 Cairns, R., Schaffer, A., Ryan, N., Pearson, S.-A., & Buckley, N. (2018). Rising pregabalin use and misuse in Australia: Trends in utilization and intentional poisonings. Addiction. Advance online publication. https://doi.org/10.1111/add.14412 Caldera, T., Herrera, A., Renberg, E., & Kullgren, G. (2004). Parasuicide in a low-income country: Results from three-year hosptial surveillance in Nicaragua. Scandinavian Journal of Public Health, 32, 349–355. https://doi.org/10.1080/14034940410029496 Carroll, R., Corcoran, P., Griffin, E., Perry, I., Arensman, E., Gunnell, D., & Metcalfe, C. (2016). Variation between hosptials in inpatient admission practices for self-harm patients and its impact on repeat presentation. Social Psychiatry and Psychiatric Epidemiology, 51, 1485–1493. https://doi.org/10.1007/s00127-016-1247-y Cooper, J., Murphy, E., Webb, R., Hawton, K., Bergen, H., Waters, K., & Kapur, N. (2010). Ethnic differences in self-harm, rates, characteristics and service provision: Three-city cohort study. British Journal of Psychiatry, 197, 212–218. https://doi.org/10.1192/ bjp.bp.109.072637 Corcoran, P., Arensman, E., & Perry, I. (2007). The area-level association between hosptial-treated deliberate self-harm, depreviation and social fragmentation in Ireland. Journal of Epidemiology and Community Health, 61, 1050–1055. https://doi.org/ 10.1136/jech.2006.055855 Daly, C., Griffin, E., Ashcroft, D., Webb, R., Perry, I., & Arensman, E. (2018). Intentional drug overdose involving pregabalin and gabapentin: Findings from the National Self-Harm Registry Ireland, 2007–2015. Clinical Drug Investigation, 38, 373–380. https://doi. org/10.1007/s40261-017-0616-y European Centre for Disease Prevention and Control. (2014). Data quality monitoring and surveillance system evaluation – a handbook of methods and applications. Stockholm, Sweden: Author. Faverjon, C., & Berezowski, J. (2018). Choosing the best algorithm for event detection based on the intended application: A conceptual framework for syndromic surveillance. Journal of Biomedical Informatics, 85, 126–135. https://doi.org/10.1016/j.jbi. 2018.08.001 Geulayov, G., Kapur, N., Turnbull, P., Clements, C., Waters, K., Ness, J., … Hawton, K. (2016). Epidemiology and trends in non-fatal self-harm in three centres in England, 2000–2012: Findings from the Multicentre Study of Self-Harm in England. BMJ Open, 6, e010538. Gill, P., Saunders, N., Gandhi, S., Gonzalez, A., Kurdyak, P., Vigod, S., & Guttmann, A. (2017). Emergency department as a first contact for mental health problems in children and youth. Journal of the American Academy of Child and Adolescent Psychiatry, 56, 475–482. https://doi.org/10.1016/j.jaac.2017.03.012 Griffin, E., Arensman, E., Corcoran, P., Dillon, C., Williamson, E., & Perry, I. (2015). National Self-Harm Registry Ireland, Annual Report, 2014. Cork, Republic of Ireland: National Self-Harm Registry.

Crisis (2019), 40(1), 1–6

Editorial

Hawton, K., Bale, E., Casey, D., Shepherd, A., Simkin, S., & Harris, L. (2006). Monitoring deliberate self-harm presentations to general hospitals. Crisis, 27, 157–163. https://doi.org/10.1027/02275910.27.4.157 Hawton, K., Bergen, H., Simkin, S., Arensman, E., Corcoran, P., Cooper, J., … Kapur, N. (2011). Impact of different pack sizes of paracetamol in the United Kingdom and Ireland on intentional overdoses: A comparative study. BMC Public Health, 11, 460. https://doi.org/10.1186/1471-2458-11-460 Hawton, K., Fagg, J., Simkin, S., Bale, E., & Bond, A. (1997). Trends in deliberate self-harm in Oxford, 1985-1995: Implications for clinical services and the prevention of suicide. British Journal of Psychiatry, 171, 556–560. Hawton, K., Lascelles, K., & Ferrey, A. (2015). Identifying and responding to suicide clusters and contagion: A practical resource. London, UK: Public Health England. Hawton, K., Zahl, D., & Weatherall, R. (2003). Suicide following deliberate self-harm: Long-term follow-up of patients who presented to a general hospital. British Journal of Psychiatry, 182, 537–542. Hedegaard, H., Schoenbaum, M., Claassen, C., Crosby, A., Holland, K., & Proescholdbell, S. (2018). Issues in developing a surveillance case definition for nonfatal suicide attempt and intentional self-harm using International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) coded data. National Health Statistics Report, 108, 1–19. Kapur, N., Murphy, E., Cooper, J., Bergen, H., Hawton, K., Simkin, S., … Owens, D. (2008). Psychosocial assessment following selfharm: Results from the Multi-Centre Monitoring of Self-Harm Project. Journal of Affective Disorders, 106, 285–293. https:// doi.org/10.1016/j.jad.2007.07.010 Kapur, N., Steeg, S., Webb, R., Haigh, M., Bergen, H., Hawton, K., … Cooper, J. (2013). Does clinical management improve outcomes following self-harm? Results from the Multicentre Study of SelfHarm in England. PLoS One, 8, e70434. https://doi.org/10.1371/ journal.pone.0070434 Knipe, D., & Gunnell, D. (2016). Annual Report, 2016: Bristol SelfHarm Surveillance Register. Bristol, UK: Bristol Self-Harm Surveillance Register. Metzeger, M.-H., Tvardik, N., Gicquel, Q., Bouvry, C., Poulet, E., & Potinet-Pagliaroli, V. (2016). Use of emergency department electronic medical records for automated epidemiological surveillance of suicide attempts: A French pilot study. International Journal of Methods in Psychiatr Research, 26, e1522. https:// doi.org/10.1002/mpr.1522 Multicentre Study of Self-Harm in England. (2018). Multicentre study of self-harm in England. Retrieved from http://cebmh. warne.ox.ac.uk/csr/mcm/index.html Office of National Statistics. (2018). Suicide occurrences, England and Wales. Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/suicideinenglandandwales Perry, I., Corcoran, P., Fitzgerald, A., Keeley, H., Reulbach, U., & Arensman, E. (2012). The incidence and repetition of hospital-treated deliberate self harm: Findings from the world’s first national registry. PLoS One, 7, e31663. https://doi.org/10.1371/ journal.pone.0031663 Pillans, P., Page, C., Ilango, S., Kashchuk, A., & Isbister, G. (2017). Self-poisoning by older Australians: A cohort study. MJA, 206, 164–169. Rajendra, R., Krishna, M., Majgi, S., Heggere, N., Robinson, C., & Poole, R. (2015). A feasibility study to establish a deliberate self-harm register in a state hosptial in southern India. British Journal of Medical Practitioners, 8, a807. Silverman, M., & De Leo, D. (2016). Why there is a need from an international nomenclature and classification system for suicide. Crisis, 37, 83–87. https://doi.org/10.1027/0227-5910/a000419 Crisis (2019), 40(1), 1–6

Snowdon, J. (2016). Why have Australian suicide rates decreased? The Australian and New Zealand Journal of Psychiatry, 50, 1315. https://doi.org/10.1177/0004867415590630 Whyte, I., Dawson, A., Buckley, N., Carter, G., & Levey, C. (1997). Health care. A model for the management of self-poisining. The Medical Journal of Australia, 167, 142–146. Whyte, I., Francis, B., & Dawson, A. (2007). Safety and efficacy of intravenous N-acetylcysteine for acetaminophen overdose: Analysis of the Hunter Area Toxicology Serivce (HATS) database. Current Medical Research and Opinion, 23, 2359–2368. https:// doi.org/10.1185/030079907X219715 Williams, S. (2015). Establishing a self-harm surveillance register to improve care in a general hospital. British Journal of Mental Health Nursing, 4, 20–25. https://doi.org/10.12968/bjmh. 2015.4.1.20 World Health Organization (2010). Suicide: Prevention and control. Geneva, Switzerland: Author. World Health Organization (2014). Suicide: A global imperative. Geneva, Switzerland: Author. World Health Organization (2016). Practice manual for establishing and maintaining surveillance systems for suicide attempts and self-harm. Geneva, Switzerland: Author. Zalsman, G., Hawton, K., Wasserman, D., van Heeringen, K., Arensman, E., Sarchiapone, M., … Zohar, J. (2017). Evidence-based national suicide prevention taskforce in Europe: A consensus position paper. European Neuropsychopharmacology, 27, 418– 421. https://doi.org/10.1016/j.euroneuro.2017.01.012 History Accepted November 5, 2018 Published online January 23,2019 Acknowledgments The authors would like to acknowledge Professors Keith Hawton, Ella Arensman, Greg Carter, and David Gunnell for their input into this editorial. KW is funded by a postdoctoral fellowship from the American Foundation for Suicide Prevention. JR is funded by a National Health and Medical Research Centre Career Development Fellowship. The development of the Victorian Surveillance System is being supported by the Victorian Department of Health and Human Services. Dr. Katrina Witt is an American Foundation for Suicide Prevention Post-Doctoral Research Fellow at Monash University and Orygen, the National Centre of Excellence in Youth Mental Health at The University of Melbourne, Australia. Her research focuses on evidence-based interventions for self-harm and suicide prevention, including improving ED treatment responses to self-harm. Dr. Jo Robinson is a Senior Research Fellow at Orygen the National Centre of Excellence in Youth Mental Health in Australia, where she leads the suicide prevention research unit. Her work focuses on the development, and testing, of novel interventions across settings, on evidence synthesis, and on the translation of research evidence into practice and policy.

Research Trends

Associations Between Suicide Risk Factors and Favorite Songs Content Analysis and Cross-Sectional Study Benedikt Till1,3, Michael Fraissler1,3, Martin Voracek2,3, Ulrich S. Tran2,3, and Thomas Niederkrotenthaler1,3 Unit Suicide Research & Mental Health Promotion, Department of Social and Preventive Medicine, Center for Public Health, Medical University of Vienna, Austria

Department of Basic Psychological Research and Research Methods, School of Psychology, University of Vienna, Austria Wiener Werkstaette for Suicide Research, Vienna, Austria

2 3

Abstract. Background: For several decades, the question of whether personal suicidality is reflected in individual music preferences has been the subject of debate in suicide research. Despite many studies investigating the relationship between music use and suicidal behavior, it is still unclear whether suicide risk is reflected in individual music preferences. Aims: The present study aimed to assess whether music preferences are reflected in suicide risk factors. Method: We assessed suicidal ideation, depression, and hopelessness among 943 participants in a cross-sectional online survey. Participants provided up to five examples of their favorite music. We conducted a content analysis and coded all reported songs as suicide-related, coping-related, or unrelated to suicide. Results: Multivariate analyses controlling for gender, age, education level, and amount of daily music use indicated associations of preferences for suicide-related songs with suicidal ideation and depression. Limitations: Limitations of the present study include the use of a convenience sample and a cross-sectional design, the small number of participants with preferences for coping-related songs, and the relatively small effect size of the associations found. Conclusion: Music preferences appear to reflect suicide risk factors, with individuals who prefer suicide-related songs scoring higher in terms of suicidal ideation and depression. Keywords: suicide risk, music preferences, online survey, content analysis, cross-sectional study, Austria

Suicide is considered the result of a combination of dif ferent risk factors (Mann et al., 2005). One of the factors reported to be linked to suicidal behavior is the content of media used by audiences (Niederkrotenthaler, Reiden berg, Till, & Gould, 2014; Niederkrotenthaler et al., 2009; Niederkrotenthaler et al., 2010; Pirkis & Blood, 2001a, 2001b). For example, sensationalist portrayals of suicide in both fictional and nonfictional media have the poten tial to increase suicidal ideation in audiences and trigger imitational suicides (Gould & Shaffer, 1986; Niederk rotenthaler et al., 2009; Niederkrotenthaler et al., 2010; Niederkrotenthaler et al., 2014; ; Phillips, 1974; Phillips & Carstensen, 1986; Stack 2005; Till, Strauss, Sonneck, & Niederkrotenthaler, 2015; World Health Organization, 2017). Scholars often refer to this phenomenon as the Werther effect (Niederkrotenthaler et al., 2009; Nieder krotenthaler et al., 2010; Niederkrotenthaler et al., 2014; Phillips, 1974; Pirkis & Blood, 2001a, 2001b). More re cent studies have also found associations between suicide risk and preference for films with suicidal or sad content (Stack, Kral, & Borowski, 2014; Till, Tran, Voracek, Son © 2018 Hogrefe Publishing

neck, & Niederkrotenthaler, 2014). By contrast, suicide portrayals focusing on coping with adverse circumstances in newspapers, fictional films, or educative websites were found to be associated with a decrease in suicidal ide ation and behavior, the so-called Papageno effect (Nieder krotenthaler et al., 2010; Till et al., 2015; Till, Tran, Vo racek, & Niederkrotenthaler, 2017). Over the past few decades, there has been a lively schol arly debate about whether music preferences are associat ed with suicide or not (Stack, Lester, & Rosenberg, 2012; Till, Tran, Voracek, & Niederkrotenthaler, 2016). Studies show that Americans on average spend more than 2.5 hr per day listening to music (Rentfrow, Goldberg, & Zilca, 2011). In Austria, 80% of survey respondents indicat ed that they listened to music on a daily or weekly basis (Huber, 2010). In spite of this vast amount of time indi viduals spend listening to music, there is very little knowl edge about whether suicide risk is reflected in individual music preferences, and, so far, the respective evidence is inconclusive. Furthermore, many authors conducted stud ies with small sample sizes, some of them with less than Crisis (2019), 40(1), 7–14 https://doi.org/10.1027/0227-5910/a000523

B. Till et al., Suicide Risk and Music

100 participants (e.g., Ballard & Coates, 1995; Burge, Gold blatt, & Lester, 2002; Lester & Whipple 1996). Also, studies using a naturalistic approach to assess associations between the content of favorite music and suicide risk are lacking. The approaches used in previous studies do not allow for direct conclusions to be drawn with respect to how much in dividuals actually prefer music with suicidal content. Several researcher have explored individuals’ music genre preferences and found associations between suicide or suicide risk factors and fanship of certain music subcul tures, including heavy metal (Burge et al., 2002; Martin, Clarke, & Pearce, 1993; Scheel & Wastefeld, 1999; Stack, 1998; Stack, Gundlach, & Reeves, 1994), gothic (Young, Sweeting, & West, 2006), punk (Young et al., 2006), al ternative rock (Pimentel, Gouveia, de Santana, Chaves, & Rodrigues, 2009), blues (Stack, 2000), opera (Stack, 2002), and country music (Stack & Gundlach, 1992). In several subsequent studies, however, these associations have not been replicated (Lacourse, Claes, & Villeneuve, 2001; Lester & Whipple 1996; Maguire & Snipes, 1994; Recours, Aussaguel, & Trujillo, 2009; Till et al., 2016). In a recent study, preferences for music genres typically featuring sad or depressing themes, such as blues or jazz music, were associated with high psychoticism scores, but not with suicidal ideation or depression (Till et al., 2016). Also, there are case studies that examined adherence to a music subculture among individuals who died by sui cide. For example, Definis-Gojanovic, Gugic, and Sutlovic (2009) found evidence of a link between suicide and Emo subculture. Other researchers have examined suicide rates with regard to specific songs and found increases of suicides linked to the country song “Whiskey Lullaby” (see Stack & Bowman, 2012) or the Hungarian suicide song “Gloomy Sunday” (Gulyás, 2008; Stack, Krysinska, & Lester, 2007). Lester and Gunn (2011a, 2011b) analyzed 18 European national anthems and reported that the greater the propor tion of low notes and sad words in anthems, and the more anthems were rated as gloomy and sad, the higher the na tional suicide rate. On the basis of the limitations of studies using aggre gate data (Mauk, Taylor, White, & Allen, 1994), some re searchers exposed participants in a laboratory experiment to songs or music videos with suicidal content, but did not find any immediate impact on suicide risk factors (Ballard & Coates, 1995; Rustad, Small, Jobes, Safer, & Peterson, 2003). Till et al. (2016) used the Beach method (Stack et al., 2014; Till et al., 2014), that is, asking participants to indicate which songs they had listened to in the past from a predefined list of selected music of 50 popular songs, in cluding 25 songs related to suicide, in order to rate prefer ence for suicidal music. On the basis of these ratings, Till et al. (2016) assessed associations between cumulative ex Crisis (2019), 40(1), 7–14

posure to suicidal music and suicide risk factors, but no sig nificant associations were found. This approach, however, is hampered by the fact that predefined music lists are not representative of individual music preferences and may or may not represent music the individual typically chooses to listen to. The present study aimed to investigate associations between suicide risk factors and suicidal content in indi viduals’ favorite music as reported by the participants, using a large sample of respondents. Based on the notion that media portrayals of suicide reflect suicide risk, we hypothesized that participants’ preference for music with suicide-related content will be associated with a higher load of suicide risk factors (i.e., suicidal ideation, depres sion, and hopelessness). Conversely, we hypothesized that preference for music with coping-related content might be associated with lower suicide risk.

Method Participants A total of 2,221 participants were recruited for a Ger man-language online survey between December 2011 and August 2012 with posters, flyers, e-mails, and public announcements at facilities of the University of Vienna and the Medical University of Vienna, Austria (Till et al., 2014; Till et al., 2016).

Ethics Statement The study was approved by the Research Ethics Board of the Medical University of Vienna and the Vienna General Hospital AKH (study protocol 1290/2012).

Measures Suicide Probability Scale Suicidal ideation was assessed with the Suicide Probability Scale (Cull & Gill, 1988), which uses 36 self-report items (e.g., “I feel people would be better off if I were dead”) rat ed on a 4-point scale ranging from none or a little of the time to most or all of the time. Responses receive weighted scores from 0 to 5 points (see Cull & Gill, 1988). Brown (2000) reported a high internal reliability of α = .93 of this scale in a sample of 579 participants, which included suicide attempters, psychiatric inpatients, and controls, and test– retest reliability over a 3-week period among 80 individ uals of the general population was high with r = .92 (Cull © 2018 Hogrefe Publishing

B. Till et al., Suicide Risk and Music

& Gill, 1988). The concurrent validity of this measure has also been established with the scale’s total score correlat ing significantly with several related measures, including depression (r = .44–.73) and suicide threats (r = .67–.71; Brown, 2000; Cull & Gill, 1988). Furthermore, the scale has successfully differentiated between 168 suicide at tempters, 130 psychiatric inpatients, and 281 controls in a previous study using a double cross-validation design, with rates of correct classification ranging from 85.5% to 87.4% (Cull & Gill, 1988). Erlanger Depression Scale The Erlanger Depression Scale (Lehrl & Gallwitz, 1983) is a self-report measure of depression symptoms, with eight items (e.g., “I want to cry”) rated on a scale from 0 (completely wrong) to 4 (exactly right). Lehrl and Gallwitz (1983) reported a test–retest reliability of r = .78 among 60 psy chiatric patients with depressive symptoms and significant correlations with various relevant or related measures, in cluding distress (r = .55–.59) and doctor’s diagnosis for de pression (r = .57), in several studies, indicating satisfactory properties of the scale. Beck Hopelessness Scale The German version (Krampen, 1994) of the Hopelessness Scale by Beck and Steer (1988) is a self-report measure with 20 items assessing hopelessness. Items (e.g., “My fu ture seems dark to me”) are rated on a 6-point scale from 1 (very false) to 6 (very true). This scale is considered to be a reliable predictor of suicidal behavior (Beck, Brown, Ber chick, Stewart, & Steer, 1990; McMillan, Gilbody, Beres ford, & Neilly, 2007). In several studies with clinical and nonclinical samples, the scale had good internal consis tency (r = .81–.94) and test–retest reliability (r = .81–.90; Krampen, 1994). The concurrent validity of this measure has been established with the scale’s total score correlat ing significantly with several related measures, including depression (r = .31–.73) and suicidal ideation (r = .78–.82), in numerous studies (Krampen, 1994). Favorite Music We asked participants to provide up to five examples (i.e., song titles) of their favorite music using an open-ended question. Sociodemographics We asked participants to report their gender (male = 1, female = 2), age (using an open-ended question), and highest completed school level (compulsory education, appren ticeship training, intermediate technical and vocational school, secondary/high school graduation, university/ college degree), with compulsory education ranked as low est and university/college degree as highest school level. © 2018 Hogrefe Publishing

Furthermore, we asked participants to rate their average amount of daily music use on a 10-point scale from 0 (never) to 9 (the entire day).

Data Analysis In order to assess whether or not the participants had a preference for music with suicidal content, we conducted a content analysis of the lyrics of the songs the participants provided as examples of their favorite music. In the case of operas or musicals, we analyzed the synopsis in the litera ture. In cases of ambiguity about the song titles provided, we selected songs that were consistent with the genres of the other song titles provided by the respective participant. If this step did not resolve the issue, the song was exclud ed from the analysis (n = 299). Songs in languages other than German, English, or French were also excluded from the analysis, unless a translation of the lyrics was available (n = 180). In total, the participants provided 3,343 music examples. Of these 3,343 music examples, 435 songs ap peared more than once in the results list. In total, 2,494 songs were included in the content analysis. On the basis of the content analysis, we categorized all songs in one of three categories: (1) suicide-related songs, (2) coping-related songs, and (3) songs unrelated to suicide. A song was coded as suicide-related, when title or lyrics (or synopsis): (1) described a suicide (e.g., “He took a suite of rooms and hung himself ”); (2) made an explicit reference to a suicide (e.g., “Today I’m gonna kill myself ”), poten tial suicide (e.g., “Maybe I should kill myself ”), or a suicide method (e.g., “I’m gonna jump in the blue”); (3) portrayed suicide in a positive way (e.g., “Trust in my self-righteous suicide”); (4) described a very hopeless situation with no way out other than suicide (e.g., “No more tomorrow,” “no more options left”); (5) voiced the wish to die (e.g., “I wish I was dead”); or (6) individual acts of suicide were attribut ed to the song (e.g., “Gloomy Sunday,” see Gulyás, 2008; Stack et al., 2007). If songs had any of the aforementioned characteristics 1–5, but expressed a successful coping of the suicidal crisis or a constructive solution to the prob lems, then they were coded as coping-related. All other songs were coded as unrelated to suicide. Pieces of music without lyrics (e.g., classical music) were coded as unrelat ed to suicide, unless they are commonly linked to suicide or death (e.g., Requiem Mass in D minor by Wolfgang Am adeus Mozart). The songs were screened and analyzed by two coders (M.F., B.T.). Intercoder reliability was measured using 125 (5%) randomly selected songs. The percentage of agree ment between the two coders was greater than 97% for all codes, and Krippendorff ’s coefficient α ranged from .72 to 1.00, which indicates a high level of agreement (Krip Crisis (2019), 40(1), 7–14

B. Till et al., Suicide Risk and Music

pendorff, 2004). In cases of disagreement, the two coders jointly revisited the lyrics of the song and discussed the rationale for their coding until agreement was reached. Using this coding system, we calculated two scores for each participant: One score that indicated whether or not participants had one or more suicide-related songs among the examples they provided of their favorite music (yes = 1, no = 0), and a second score that indicated whether or not participants had one or more coping-related songs among their favorite songs (yes = 1, no = 0). These two scores were independent of each other (i.e., one or more suicide-relat ed songs vs. no suicide-related song among a respondent’s favorite songs; one or more coping-related songs vs. no coping-related song among a respondent’s favorite songs). One respondent (0.1%) was coded with “1” for both scores, because he listed both a suicide-related song and a coping-related song among his favorite songs. We conducted multiple linear regression analysis (entry method) for each of three outcome variables – suicidal ide ation, depression, and hopelessness – to estimate associa tions between suicide risk factors and music preferences. The two music preference scores were used as explanatory variables. All regression analyses controlled for the partic ipants’ gender, age, education, and amount of daily music use.

Results Demographic Characteristics of the Sample Of the 2,221 participants recruited for the survey, a total of 943 individuals (42.4%), with mean age of 29.7 years (SD = 9.7), including 631 women (66.9%) and 312 men (33.1%), completed the entire survey and were included in the study. In terms of highest completed school level, 0.8% (n = 8) of the participants had compulsory educa tion, 1.7% (n = 16) completed apprenticeship training, 1.9% (n = 18) completed intermediate technical and vo cational school, 56.0% (n = 528) were secondary/high school graduates, and 39.6% (n = 373) completed college or university.

Association Between Music Preferences and Suicide Risk Factors The participants provided an average number of 3.79 examples per person (Mdn = 4.00, IQR = 2.00, Min. = 0, Max. = 5) of their favourite music. In total, 62 songs Crisis (2019), 40(1), 7–14

Table 1. Descriptive statistics for suicide risk factors and music preferences of study participants (n = 943) α

Min.

Max.

Suicidal ideation

.89

48.52

14.12

121

Depression

.81

6.76

5.29

Hopelessness

.88

49.33

14.32

111

Suicide risk factors

Music preferences

Preference for suicide-related songs

8.5

Preference for coping-related songs

1.4

Note. Means (M), standard deviations (SD), and Cronbach alpha values (α), lowest (Min.) and highest (Max.) values of the outcome variables (suicidal ideation, depression, hopelessness), as well as frequencies (n) and percentages (%) of a preference for suicide-related songs or for coping-related songs.

were identified as suicide-related and nine as coping-re lated. The most frequently mentioned suicide-related songs were “Chop Suey!” by System of a Down (n = 7, 0.21%), “Sail” by Awolnation (n = 4, 0.12%), and “Alles aus Liebe” [All Out of Love] by Die Toten Hosen (n = 4, 0.12%). The most prevalent coping-related songs/pieces of music were The Magic Flute by Wolfgang Amadeus Mo zart (n = 4, 0.12%) and “Wonderful Life” by Hurts (n = 2, 0.06%). Lists of all suicide- and coping-related songs are made available as supplemental files (see tables in Elec tronic Supplementary Material 1 and Electronic Supple mentary Material 2). Of the 943 respondents, 80 partic ipants (8.5%) had at least one suicide-related song and 13 participants (0.4%) at least one coping-related song among the examples of their favorite music. See Table 1 for descriptive statistics for music preferences and all sui cide risk factors. Preference for suicide-related songs was a significant predictor of suicidal ideation and depression. Participants listing one or more songs with suicidal content as exam ples of their favorite music tended to have higher scores on suicidal ideation and depression. Hopelessness was not associated with preference for suicide-related songs, and preference for coping-related songs did not significantly predict any of the three outcome variables, which may be attributed to the low prevalence of coping-related songs among the provided music examples and the resulting lack of statistical power. While 62 songs were identified as suicide-related, only nine songs were coded as coping-re lated, and only 13 participants (1.4%) listed one or more coping-related songs among their favorite songs. Overall, female gender, young age, and low education level tended to correlate significantly with higher scores on most sui cide risk factors. See Table 2 for the standardized regres sion coefficients and standard errors of the predictors and control variables. © 2018 Hogrefe Publishing

B. Till et al., Suicide Risk and Music

Table 2. Results of multiple linear regression analyses to predict suicide risk factors in study participants (n = 943) Risk factor Suicidal ideation

Depression

Hope lessness

β (SE)

Preference for suicide-related songs

.08* (.03)

.10** (.03)

.04 (.03)

Preference for coping-related songs

−.02 (.03)

Predictors

−.01

(.03)

−.02 (.03)

−.01 (.03)

.07* (.03)

.00 (.03)

Age

−.07* (.04)

−.10** (.04)

−.01 (.04)

Education

−.08* (.03)

−.05

(.04)

−.08* (.04)

.01 (.03)

−.01

(.04)

−.02 (.04)

Covariates Gendera

Music consumption Model indices R2

.01

.02

.00

2.71*

3.86**

1.14

Note. Values are standardized regression coefficients (β), standard errors of the beta coefficients (SE), as well as adjusted R2 and F values. * p < .05. **p < .01. ***p < .001 (two-tailed). a Reference group: male.

to use media content related to suicide. In fact, a labora tory experiment conducted in Austria showed that among viewers of a film concluding with the protagonist’s suicide, suicidal ideation was greater the more they used the film for their own problem-solving (Till, Vitouch, Herberth, Sonneck, & Niederkrotenthaler, 2013). Suicidal individ uals perhaps also prefer music that reflects their current psychological state of mind. In contrast to these findings, preference for coping-related songs was not associated with suicide risk factors, suggesting that the association of music preferences and suicide risk might be restricted to music with suicidal content. A strength of the present study is the large sample size. Most previous research relied on relatively small sample sizes (e.g., Ballard & Coates, 1995; Burge et al., 2002; Lester & Whipple, 1996). Furthermore, this study is the first to examine individuals’ music preferences via content analysis, which provides a more reliable estimate of the music content that individuals are actually exposed to in everyday life than methodological approaches in previous studies that assessed preferences for specific music genres or knowledge of suicide-related songs from a predefined list of selected music.

Discussion

Limitations

In this study, preference for suicide-related songs predict ed higher scores on suicidal ideation and depression. This held true when gender, age, education, and amount of daily music use were controlled for. These findings are consist ent with prior related evidence suggesting a link between suicide rates and the amount of airtime devoted to specific suicide-related songs in the general public (Gulyás, 2008; Stack & Bowman, 2012; Stack et al., 2007). The findings are also in line with studies demonstrating associations be tween suicidal behavior and preferences for films with sui cidal content (Stack et al., 2014). Our findings could be ex plained by the notion that music is affecting its audience, or via the assumption that music preferences mirror or re flect the psychological or emotional state of the audience (see Stack & Bowman, 2011). According to the Werther ef fect (Phillips, 1974), it is plausible to assume that cumula tive exposure to suicidal content in songs has the potential to increase suicidal ideation and depression. Vice versa, individuals who are currently facing adverse life circum stances and are already thinking about suicide may seek music content related to suicide in order to act out or cope with their suicidal feelings (Etzersdorfer, 2008; Etzersdor fer & Sonneck, 1998). According to the concept of suicidal constriction (Shneidman, 1995), suicidal individuals tend to focus overly on the topic of suicide and generally seem

The current study also has some limitations. First, the lack of association between preference for coping-related songs and suicide risk factors may be based on the rela tively small number of coping-related songs identified in the content analysis. The low frequency of preferences for coping-related songs limits analytic power in the regres sion analyses. Second, we used a convenience sample to collect data, resulting in an overrepresentation of wom en, young individuals, and participants who had gradu ated from high school or college in our sample compared with the general population in Austria (Statistik Austria, 2011). Furthermore, music preferences accounted only for a small proportion of variance in suicide risk factors, indicating a small effect size. It is, however, important to note that this is consistent with typical effect sizes in psy chological science (Hemphill, 2003; Meyer et al., 2001; Richard, Bond Jr., & Stokes-Zoota, 2003). Another limi tation of the present study is that the findings cannot be generalized to other cultures because of cultural differenc es. Favorite suicide songs may follow national or regional cultural patterns with limited overlap between lists of fa vorite suicide songs between different countries. Future studies may want to explore the global spread of specific suicide songs. It is also still unclear how recipients process suicidal content during music reception and how such con

Crisis (2019), 40(1), 7–14

tent impacts suicide cognitions and emotions. The effects of the same music content may vary among different re cipients (Till et al., 2016). Future studies may benefit from exploring underlying emotional and cognitive processes of this phenomenon. Finally, based on the cross-sectional design of the study, causality concerning the associations between music preferences and suicide risk factors cannot be assessed. Longitudinal studies may help to gain further insight into the impact of suicide-related music on individ uals’ suicidal behavior.

Conclusion Considering that the relationship between suicidality and music consumption is still unclear (Stack et al., 2012; Till et al., 2016), the present study adds to the literature that suicide risk is, to some extent, reflected in individual pref erences for suicide-related music. This finding extends previous research based on preselected lists of music to test individual preferences, which, however, may not re flect songs the individual typically chooses to listen to. Participants who listed one or more suicide-related songs among their favorite music had slightly higher scores on suicidal ideation and depression.

Electronic Supplementary Material The electronic supplementary material is available with the online version of the article at https:/doi.org/ 10.1027/0227-5910/a000523 ESM 1. Table Title and performer of all suicide-related songs provided by study participants. ESM 2. Table Title and performer of all coping-related songs provided by study participants.

References Ballard, M. E., & Coates, S. (1995). The immediate effects of homicidal, suicidal, and nonviolent heavy metal and rap songs on the moods of college students. Youth & Society, 27, 148–168. https://doi.org/10.1177/0044118X95027002003 Beck, A. T., Brown, G., Berchick, R. J., Stewart, B. L., & Steer, R. A. (1990). Relationship between hopelessness and ultimate suicide: A replication with psychiatric outpatients. American Crisis (2019), 40(1), 7–14

B. Till et al., Suicide Risk and Music

Journal of Psychiatry, 147, 190–195. https://doi.org/10.1176/ ajp.147.2.190 Beck, A. T., & Steer, R. A. (1988). Manual for the Beck Hopelessness Scale. San Antonio, TX: Psychological Corporation. Brown, G. K. (2000). A review of suicide assessment measures for intervention research with adults and older adults. Philadelphia, PA: University of Pennsylvania. Burge, M., Goldblat, C., & Lester, D. (2002). Music preferences and suicidality: A comment on Stack. Death Studies, 26, 501–504. https://doi.org/10.1080/074811802760139021 Cull, J. G., & Gill, W. S. (1988). Suicide Probability Scale (SPS). Los Angeles, CA: WPS. Definis-Gojanovic, M., Gugic, D., & Sutlovic, D. (2009). Suicide and Emo youth subculture – a case analysis. Collegium Antropologicum, 33, 173–175. Etzersdorfer, E. (2008). Psychonanalytische Überlegungen zum Phänomen der Imitationssuizide [Psychoanalytic considerations on the phenomenon of imitational suicides]. In A. Herberth, T. Niederkrotenthaler, & B. Till (Eds.), Suizidalität in den Medien: Interdisziplinäre Betrachtungen [Suicidality in the media: Interdisciplinary contributions] (pp. 33–44). Münster, Germany: LIT. Etzersdorfer, E., & Sonneck, G. (1998). Preventing suicide by influencing mass-media reporting. The Viennese experience 1980-1996. Archives of Suicide Research, 4, 67–74. https://doi. org/10.1080/13811119808258290 Gould, M. S., & Shaffer, D. (1986). The impact of suicide in television movies: Evidence of imitation. New England Journal of Medicine, 315, 690–694. https://doi.org/10.1056/NEJM198609113151107 Gulyás, R. (2008). Todeswege und Engel. Darstellungen des Suizids und der Suizidalität im ungarischen Spielfilm von 1970 bis 2000 [Paths of death and angels. Portrayal of suicide and suicidal tendencies in Hungarian cinema between 1970 and 2000]. In A. Herberth, T. Niederkotenthaler & B. Till (Eds.), Suizidalität in den Medien: Interdisziplinäre Betrachtungen [Suicidality in the media: Interdisciplinary contributions] (pp. 107–119). Münster, Germany: LIT. Hemphill, J. F. (2003). Interpreting the magnitudes of correlation coefficients. American Psychologist, 58, 78–80. https://doi.org/ 10.1037/0003-066X.58.1.78 Huber, M. (2010). Wozu Musik? Musikalische Verhaltensweisen, Vorlieben und Einstellungen der Österreicher/innen [Why music? Musical behavior, preferences, and attitudes of Austrians]. Vienna, Austria: Institute for Music Sociology. Krampen, G. (1994). Skalen zur Erfassung von Hoffnungslosigkeit (H-Skalen). Deutsche Bearbeitung und Weiterentwicklung der H-Skalen von Aaron T. Beck [Scales for assessment of hopelessness (H-scales). German adaption and enhancement of the h-scales by Aaron T. Beck]. Göttingen, Germany: Hogrefe. Krippendorff, K. (2004). Content analysis: An introduction to its methodology. Thousand Oaks, CA: Sage. Lacourse, E., Claes, M., & Villeneuve, M. (2001). Heavy metal music and adolescent suicide risk. Journal of Youth and Adolescence, 30, 321–332. Lehrl, S., & Gallwitz, A. (1983). Erlanger Depressions-Skala EDS [Erlangen Depression Scale EDS]. Vaterstetten, Germany: Vless. Lester, D., & Gunn, J. F. (2011a). Lyrics of national anthems and suicide rates. Psychological Reports, 109, 137–138. Lester, D., & Gunn, J. F. (2011b). National anthems and suicide rates. Psychological Reports, 108, 43–44. Lester, D., & Whipple, M. (1996). Music preference, depression, suicidal preoccupation, and personality. Suicide and Life-Threatening Behavior, 26, 68–70. Maguire, E., & Snipes, J. (1994). Reassessing the link between country music and suicide. Social Forces, 72, 1239–1243. https://doi.org/10.2307/2580301 © 2018 Hogrefe Publishing

B. Till et al., Suicide Risk and Music

Mann, J. J., Apter, A., Bertolote, J., Beautrais, A., Currie, D., Haas, A., … Hendin, H. (2005). Suicide prevention strategies – a systematic review. JAMA, 294, 2064–2074. https://doi.org/10.1001/ jama.294.16.2064 Martin, G., Clarke, M., & Pearce, C. (1993). Adolescent suicide: Music preference as an indicator of vulnerability. Journal of the American Academy of Child and Adolescent Psychiatry, 32, 530– 535. https://doi.org/10.1097/00004583-199305000-00007 Mauk, G. W., Taylor, M. J., White, K. R., & Allen, T. S. (1994). Comments on Stack and Gundlach’s “The effect of country music on suicide”: An “achy breaky heart” may not kill you. Social Forces, 72, 1249–1255. https://doi.org/10.2307/2580303 McMillan, D., Gilbody, S., Beresford, E., & Neilly, L. (2007). Can we predict suicide and non-fatal self-harm with the Beck Hopelessness Scale? A meta-analysis. Psychological Medicine, 37, 769–778. https://doi.org/10.1017/S0033291706009664 Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. R., … Reed, G. M. (2001). Psychological testing and psychological assessment: A review of evidence and issues. American Psychologist, 56, 128–165. https://doi.org/10.1037/0003066x.56.2.128 Niederkrotenthaler, T., Reidenberg, D., Till, B., & Gould, M. (2014). Increasing help-seeking and referrals for individuals at risk for suicide by decreasing stigma: The role of mass media. American Journal of Preventive Medicine, 47, 235–243. https://doi. org/10.1016/j.amepre.2014.06.010 Niederkrotenthaler, T., Till, B., Kapusta, N. D., Voracek, M., Dervic, K., & Sonneck, G. (2009). Copycat effects after media reports on suicide: A population-based ecologic study. Social Science & Medicine, 69, 1085–1090. https://doi.org/10.1016/j. socscimed.2009.07.041 Niederkrotenthaler, T., Voracek, M., Herberth, A., Till, B., Strauss, M., Etzersdorfer, E., … Sonneck, G. (2010). The role of media reports in completed and prevented suicide: Werther versus Papageno effects. British Journal of Psychiatry, 197, 234–243. https://doi.org/10.1192/bjp.bp.109.074633 Phillips, D. P. (1974). The influence of suggestion on suicide: Substantive and theoretical implications of the Werther effect. American Sociological Review, 39, 340–354. https://doi. org/10.2307/2094294 Phillips, D. P., & Carstensen, L. L. (1986). Clustering of teenage suicides after television news stories about suicide. New England Journal of Medicine, 315, 685–689. https://doi.org/10.1056/ NEJM198609113151106 Pimentel, C. E., Gouveia, V. V., de Santana, N. L., Chaves, W. A., & Rodrigues, C. A. (2009). Preferencia musical e risco de suicidio entre jovens [Music preference and suicide risk among youths]. Jornal Brasileiro de Psiquiatria, 58, 26–33. https://doi. org/10.1590/S0047-20852009000100004 Pirkis, J., & Blood, W. (2001a). Suicide and the media. Part I: Reportage in nonfictional media. Crisis, 22, 146–154. https://doi. org/10.1027//0227-5910.22.4.146 Pirkis, J., & Blood, W. (2001b). Suicide and the media. Part II: Portrayal in fictional media. Crisis, 22, 155–162. https://doi. org/10.1027//0227-5910.22.4.155 Recours, R., Aussaguel, F., & Trujillo, N. (2009). Metal music and mental health in France. Culture, Medicine and Psychiatry, 33, 473–488. https://doi.org/10.1007/s11013-009-9138–2 Rentfrow, P. J., Goldberg, L. R., & Zilca, R. (2011). Listening, watching, and reading: The structure and correlates of entertainment preferences. Journal of Personality, 79, 223–257. https://doi. org/10.1111/j.1467-6494.2010.00662.x Richard, F. D., Bond, C. F., Jr., & Stokes-Zoota, J. J. (2003). One hundred years of social psychology quantitatively described. Review of General Psychology, 7, 331–363. https://doi. org/10.1037/1089-2680.7.4.331 © 2018 Hogrefe Publishing

Rustad, R., Small, J. E., Jobes, D. A., Safer, M. A., & Peterson, R. J. (2003). The impact of rock videos and music with suicidal content on thoughts and attitudes toward suicide. Suicide and Life-Threatening Behavior, 33, 120–131. Scheel, K. R., & Westfield, J. (1999). Heavy metal music and adolescent suicidality: An empirical investigation. Adolescence, 34, 253–273. Shneidman, E. S. (1995). Suicide as psychache: A clinical approach to self-destructive behavior. Northvale, NJ: Jason Aronson. Stack, S. (1998). Heavy metal, religiosity and suicide acceptability. Suicide and Life-Threatening Behavior, 28, 388–394. Stack, S. (2000). Blues fans and suicide acceptability. Death Studies, 24, 223–231. https://doi.org/10.1080/074811800200559 Stack, S. (2002). Opera subculture and suicide for honor. Death Studies, 26, 431–437. https://doi.org/10.1080/07481180290086763 Stack, S. (2005). Suicide in the media: A quantitative review of studies based on nonfictional stories. Suicide and Life-Threatening Behavior, 35, 121–133. https://doi.org/10.1521/suli.35.2. 121.62877 Stack, S., & Bowman, B. (2011). Durkheim at the movies: A century of suicide in film. Crisis, 32, 175–177. https://doi. org/10.1027/0227-5910/a000121 Stack, S., & Bowman, B. (2012). Suicide movies: Social patterns 1900-2009. Cambridge, MA: Hogrefe Publishing. Stack, S., & Gundlach, J. (1992). The effect of country music on suicide. Social Forces, 71, 211–218. https://doi.org/10.2307/ 2579974 Stack, S., Gundlach, J., & Reeves, J. (1994). The heavy metal subculture and suicide. Suicide and Life-Threatening Behavior, 24, 15–23. Stack, S., Kral, M., & Borowski, T. (2014). Exposure to suicide movies and suicide attempts. Sociological Focus, 47, 61–70. https:// doi.org/10.1080/00380237.2014.856707 Stack, S., Krysinska, K., & Lester, D. (2007). Gloomy Sunday: Did the “Hungarian suicide song” really create a suicide epidemic? Omega: Journal of Death and Dying, 56, 349–358. https://doi. org/10.2190/OM.56.4.c Stack, S., Lester, D., & Rosenberg, J. S. (2012). Music and suicidality: A quantitative review and extension. Suicide and Life-Threatening Behavior, 42, 654–671. https://doi.org/10.1111/j.1943278X.2012.00120.x Statistik Austria. (2011). Bildungsstand der Bevölkerung im Alter von 25 bis 64 Jahren, 1971 bis 2009 [Educational background of the population between 25 and 64 years of age,1971–2009].Retrieved from http://www.statistik.at/web_de/statistiken/bildung_und_ kultur/bildungsstand_der_bevoelkerung/020912.html Till, B., Strauss, M., Sonneck, G., & Niederkrotenthaler, T. (2015). Determining the effects of films with suicidal content: A laboratory experiment. British Journal of Psychiatry, 207, 72–78. https://doi.org/10.1192/bjp.bp.114.152827 Till, B., Tran, U. S., Voracek, M., & Niederkrotenthaler, T. (2016). Music and suicidality: A study on associations between music preferences and risk factors of suicide. Omega: Journal of Death and Dying, 72, 340–356. https://doi.org/10.1177/ 0030222815575284 Till, B., Tran, U. S., Voracek, M., & Niederkrotenthaler, T. (2017). Papageno vs. Werther effect online: Randomized controlled trial of beneficial and harmful impacts of educative suicide prevention websites. British Journal of Psychiatry, 211, 109–115. https:// doi.org/10.1192/bjp.bp.115.177394 Till, B., Tran, U. S., Voracek, M., Sonneck, G., & Niederkrotenthaler, T. (2014). Associations between film preferences and risk factors for suicide: An online survey. PLOS ONE, 9, e102293. https://doi. org/10.1371/journal.pone.0102293 Till, B., Vitouch, P., Herberth, A., Sonneck, G., & Niederkrotenthaler, T. (2013). Personal suicidality in the reception of and identificaCrisis (2019), 40(1), 7–14

tion with suicidal film characters. Death Studies, 37, 383–392. https://doi.org/10.1080/07481187.2012.673531 World Health Organization. (2017). Preventing suicide. A resource for media professionals: Update 2017. Geneva, Switzerland: Author. Young, R., Sweeting, H., & West, P. (2006). Prevalence of deliberate self-harm and attempted suicide within contemporary Goth youth subculture: Longitudinal cohort study. British Medical Journal, 332, 1058–1061. https://doi.org/10.1136/bmj.38790. 495544.7C

Received June 19, 2017 Revision received November 29, 2017 Accepted December 14, 2017 Published online May 31, 2018 Benedikt Till, DSc, is a psychologist and Assistant Professor at the Unit Suicide Research & Mental Health Promotion, Department of Social and Preventive Medicine, Center for Public Health, Medical University of Vienna. He is board member of the Wiener Werkstaette for Suicide Research and works in the field of media psychology and suicide research. Michael Fraissler, MD, is a medical doctor in training. He completed his diploma thesis at the Unit Suicide Research & Mental Health Promotion, Department of Social and Preventive Medicine, Center for Public Health, Medical University of Vienna, Austria, and is a member of the Wiener Werkstaette for Suicide Research.

Crisis (2019), 40(1), 7–14

B. Till et al., Suicide Risk and Music

Martin Voracek, DSc, DMSc, PhD, is Professor of Psychological Research Methods/Research Synthesis, Head (Research Methods Unit) and Deputy Head (Department of Basic Psychological Research and Research Methods) at the School of Psychology, University of Vienna, Austria, and Deputy Chair of the Wiener Werkstaette for Suicide Research. Ulrich S. Tran, DSc, is a clinical psychologist and Senior Lecturer of Methods in Psychology at the Department of Basic Psychological Research and Research Methods, School of Psychology, University of Vienna, Austria. He is a board member of the Wiener Werkstaette for Suicide Research. Thomas Niederkrotenthaler, MD, PhD, MMS, is Associate Professor and head of the Suicide Research & Mental Health Promotion Unit, Department of Social and Preventive Medicine, Vienna, Austria. He is co-chair of the IASP’s Media & Suicide Special Interest Group, and chairman of the Wiener Werkstaette for Suicide Research (http://www.suizidforschung.at).

Benedikt Till Unit Suicide Research & Mental Health Promotion Department of Social and Preventive Medicine Center for Public Health Medical University of Vienna Kinderspitalgasse 15 A-1090 Vienna Austria benedikt.till@meduniwien.ac.at

Research Trends

Suicide Prevention Training – Improving the Attitudes and Confidence of Rural Australian Health and Human Service Professionals Monika Ferguson1, James Dollman2, Martin Jones2, Kathryn Cronin3, Lynne James1, Lee Martinez2, and Nicholas Procter1 School of Nursing and Midwifery, University of South Australia, Adelaide, SA, Australia University Department of Rural Health, University of South Australia, Adelaide, SA, Australia 3 Country Health South Australia Local Health Network, Port Lincoln, SA, Australia 1 2

Abstract. Background: Suicide rates are higher in rural and remote areas of Australia compared with major cities. Aim: To evaluate the impact of a brief, community-based suicide prevention educational intervention on the attitudes and confidence of rural South Australian health and human service professionals. Method: Participants attended a 1-day suicide prevention education program, and completed a survey at four time points: baseline (T1), immediately pretraining (T2), immediately posttraining (T3), and 4-month follow-up (T4). Main outcome measures were self-reported attitudes and confidence when working with people vulnerable to suicide. Results: A total of 248 people attended the training, with 213 participants completing the survey at T1, 236 at T2, 215 at T3, and 172 at T4. There were significant improvements in 11 of the 14 attitude items between T2 and T3 (immediate change), and between T1 and T4 (maintained change). Further, there were significant improvements in all four confidence items between T2 and T3, and T1 and T4. Limitations: Despite the repeated-measures design, findings are limited by the lack of a control group. Conclusion: Findings extend the international evidence by indicating the value of brief suicide prevention education for improving health and human service professionals’ attitudes and confidence in rural Australia. Keywords: suicide prevention education, attitudes, confidence, rural Australia

Suicide is a major public health concern. Globally, it is estimated that over 800,000 people die by suicide each year, making it the second leading cause of death for people aged 15–29 years (World Health Organization [WHO], 2014). In Australia, 2,866 suicide deaths were registered in 2016 (Australian Bureau of Statistics [ABS], 2017a). Further, national data indicate that over 2.1 million Australian adults experience suicidal ideation in their lifetime and over 500,000 make an attempt to end their life (Johnston, Pirkis, & Burgess, 2009). This has profound and enduring impacts on individuals, their family (Cerel, Jordan, & Duberstein, 2008), and friends (Clark & Goldney, 2000), as well as health professionals (Seguin, Bordeleau, Drouin, Castelli-Dransart, & Giasson, 2014). The economic costs are also substantial. One recent estimate calculated the economic impact of suicide and nonfatal suicidal behavior on the Australian workforce at $6.73 billion per year, at© 2018 Hogrefe Publishing

tributed to various costs, including: lost productivity due to absenteeism and premature mortality, medical and rehabilitation services, and postvention services (Kinchin & Doran, 2017). Similar to global trends, certain vulnerable groups in Australia exhibit higher rates of suicide than others, such as those living in rural and remote regions. National data indicate that despite accounting for less than 30% of the Australian population (ABS, 2017b), those living outside of major cities are 66% more likely to die by suicide than those living in major cities (ABS, 2011). Possible reasons for this difference include adverse socioeconomic conditions, increased access to lethal means, attitudes that discourage help-seeking, and reduced access to services, such as health care (Stark & Riordan, 2011). Further, additional factors are associated with the higher rates of suicide outside of major cities, such as working in the farming industry Crisis (2019), 40(1), 15–26 https://doi.org/10.1027/0227-5910/a000524

(Kennedy, Maple, Mckay, & Brumby, 2014) and male gender (ABS, 2011). It is becoming increasingly recognized that addressing suicide requires a multifaceted approach, with prevention efforts across all levels of society (Department of Health and Ageing, 2008; WHO, 2014). In particular, educational initiatives seeking to raise awareness, reduce stigma, and encourage help-seeking and supportive responses are a common prevention approach. Numerous systematic reviews highlight the value of suicide prevention education programs across a range of professions and sectors of society (Isaac et al., 2009), as well as for certain groups, such as nurses (Ferguson et al., 2017), and those working with specific at-risk populations, such as youth (Robinson et al., 2013) and Indigenous peoples (Clifford, Doran, & Tsey, 2013). These reviews indicate the importance of education programs for improving participant knowledge, attitudes, skills, and confidence when responding to signs of suicidal ideation and behavior, particularly immediately after training. For health and human service professionals specifically, these programs are considered vital, given the role that these individuals play in preventing suicide though their clinical actions, such as assessing risk (Brunero, Smith, Bates, & Fairbrother, 2008). The rationale behind such programs is that most people who die by suicide have some contact with a health or human service professional prior to the event (De Leo, Draper, Snowdon, & Kolves, 2013), and therefore how these professionals respond can influence the outcomes for the individual (e.g., referral to appropriate services; Brunero et al., 2008). It can also have implications for the individual’s perceptions of the support they have received (Cutcliffe, McKenna, Keeney, Stevenson, & Jordan, 2013; Cutcliffe, Stevenson, Jackson, & Smith, 2006). In turn, this might influence their future help-seeking behavior (e.g., willingness to return to a service where they previously received a supportive response). While evaluations of suicide prevention education programs for health professionals outside of Australia have indicated the promise of this approach (e.g., Appleby et al., 2000; Gask, Dixon, Morriss, Appleby, & Green, 2006; Gask, Lever-Green, & Hays, 2008; Morriss, Gask, Battersby, Francheschini, & Robson, 1999), there is a dearth of literature exploring its effectiveness for health and human service professionals in rural and/or remote Australia. In their paper describing the Community Response to Eliminating Suicide (CORES) program in rural Tasmania, Jones, Walker, Miles, De Silva, & Zimitat (2015) highlight the importance of community-based gatekeeper training for building and empowering rural communities to prevent suicide, through improved interpersonal skills and suicide awareness, as well as for building peer support and awareness of support services within the community. A Crisis (2019), 40(1), 15–26

M. Ferguson et al., Suicide Prevention Training in Rural Australia

brief self-report evaluation with 120 participants who attended a 1-day CORES session showed that training can improve comfort with talking to others about suicide (including family, friends, colleagues, and strangers), as well as improve confidence in participants’ ability to provide appropriate assistance. While the findings are promising, participants were members of the general public, not specifically health and human service professionals. Given the need to address the high rate of suicide in rural Australia, the potential for training of health and human service professionals to be an effective suicide prevention activity warrants further exploration.

Aims The primary aim of this research was to investigate the impact of participating in a 1-day suicide prevention education program on rural South Australian health and human service professionals’ attitudes toward suicide prevention and their confidence in suicide management. The study also sought to determine whether any changes in attitudes and confidence were maintained beyond the training program.

Method Ethics approval was granted from the following: University of South Australia Human Research Ethics Committee; SA Health Human Research Ethics Committee; Aboriginal Health Research Ethics Committee; and Department for Education and Child Development.

Participants and Recruitment The target population in this study was all health and human service professionals working in and/or around eight South Australian rural towns. With a combined population of approximately 100,000 people, this included six outer regional towns (Whyalla, Mount Gambier, Port Pirie, Port Augusta, Peterborough, and Renmark); one remote town (Port Lincoln); and one very remote town (Coober Pedy). Participation was open to a range of health and human service professionals (e.g., nurses and allied health staff, social workers, teachers and other school staff, case managers, and youth workers), across public, private, and nongovernment organizations. Students undertaking tertiary education in any of these professions in the training towns were also eligible to attend the training; however, only © 2018 Hogrefe Publishing

M. Ferguson et al., Suicide Prevention Training in Rural Australia

those studying at the University of South Australia were eligible to participate in the research. A multifaceted recruitment strategy was implemented: E-mails were sent to potential participants from team leaders within services; project posters were displayed at relevant offices/services; and interviews with members of the research team were featured in local newspapers and radio. Potential participants were invited to contact the research team to receive further information and to register their interest in participating.

Intervention Description All participants attended a 1-day (6.5 hrs) suicide prevention education program. The training program was developed as a key priority of the University of South Australia Department of Rural Health. University Departments of Rural Health were established in 1996 across regional Australia in response to workforce shortages in regional and remote Australia. Their remit is to focus on the development of multidisciplinary training experiences for students and the regional health workforce in rural Australia. Further, they work with regional communities to increase health service accessibility. A key focus of the University of South Australia Department of Rural Health at the time

of this study was to work with regional communities to increase capacity in the area of suicide prevention. In partnership with the University of South Australia’s School of Nursing and Midwifery, and SA Health, Government of South Australia, the current training program was designed to help regional South Australian communities better recognize and respond to suicide. The education session was facilitated by a multidisciplinary project team, comprising mental health academics and clinicians, a lived experience health worker, and an Aboriginal health project officer. The project team developed the content, drawing on peer-reviewed evidence embedded within the recently produced “Engaging With the Suicidal Person: A Resource From Shared Learning in Clinical Practice” clinical resource, jointly produced by the University of South Australia and SA Health. Further, synthesizing global and national evidence in suicide prevention (Department of Health and Ageing, 2008; WHO, 2014), training was delivered through a series of lecture-style and small breakout sessions, using interactive learning activities (see Figure 1 for sample program). A unique aspect of the training session was the lived experience reflections segments, involving a mental health consumer from rural South Australia describing experiences of suicidal ideation, behavior, and help-seeking. The mental health consumer outlined aspects of care they viewed as helpful and unhelpful, Figure 1. Sample training day program.

Formatiert: Schriftart: Fett, 40(1), 15–26 CrisisNicht (2019), Kursiv Formatiert: Schriftart: Nicht Kursiv Formatiert: Schriftart: Nicht Kursiv

including the importance of health professionals enquiring about suicidal intent, and conveyed the message that suicide is preventable. To support ongoing learning, all participants received a take-home resource pack comprising training slides, relevant suicide prevention resources (e.g., peer-reviewed literature), support service contact details, and a further reading list. Participants were provided with morning tea and lunch, and a certificate of attendance was issued electronically following the event. There was no cost to participants for attending the training.

Outcome Measures Participants completed a brief self-report questionnaire at four time points: baseline (T1; at registration, ranging from 14 weeks to 1 day before training); immediately pretraining (T2); immediately posttraining (T3); and 4-month follow-up (T4). T1 and T4 surveys were completed online via SurveyMonkey, while T2 and T3 surveys were completed in hard copy. The main outcome measures were attitudes toward suicide and confidence in suicide management. Attitudes to Suicide Prevention Scale This 14-item scale assesses health professionals’ attitudes toward the assessment and management of suicide risk (Herron, Ticehurst, Appleby, Perry, & Cordingley, 2001). Each item is rated on a 5-point scale (1 = strongly disagree to 5 = strongly agree), with a lower score indicating a more positive attitude. Four items (4, 7, 9 and 14) are negatively worded, and therefore were reverse coded prior to data analysis. Minor wording changes were made to some items, according to current suicidology nomenclature and the Australian context (e.g., replacing “suicidal patients” with “individuals at risk of suicide,” and “committing suicide” with “ending their life by suicide”). This scale has demonstrated strong test–retest reliability among health professionals (r = .85), and satisfactory internal consistency (Cronbach’s α = .77; Herron et al., 2001). Confidence in Suicide Management Confidence was measured with Morriss and colleagues’ (1999) scale, which assesses confidence when working with individuals experiencing suicidal states. As per later adaptations of the scale (Appleby et al., 2000; Gask et al., 2006, 2008), one of the original items (“dealing with suicidal clients can be improved by attending a training course”) was removed, resulting in a 4-item scale. The original measure included a 10-cm visual analog scale (Morriss et al., 1999); however, an 11-point scale was adopted in the current study for compatibility with online survey formatting (with options ranging from 0 = not at Crisis (2019), 40(1), 15–26

M. Ferguson et al., Suicide Prevention Training in Rural Australia

all conﬁdent to 10 = extremely conﬁdent). A higher score on each item indicates greater self-reported confidence. Although no psychometric properties have been reported, this scale has been widely used to assess changes in health professionals’ confidence associated with suicide prevention education programs (e.g., Appleby et al., 2000; Gask et al., 2006, 2008; Morriss et al., 1999), allowing comparability with the international literature. Minor wording changes were also made to this scale. Participant Demographics Demographic and profession-related questions were completed at both T1 and T2 (to account for those who attended the training without completing the T1 survey). Questions included: gender, age, current occupation, years of experience in current occupation, workplace setting (e.g., hospital vs. community), mental health setting, highest academic qualification, prior attendance at suicide prevention training, and previous experiences with suicide, both personally and professionally. Program Evaluation Program evaluation questions were completed at T3 to understand participants’ experiences with the training. This included four rating scale items, marked on a 5-point scale (ranging from strongly agree to strongly disagree): “The training was relevant to my role,” “the training has made a positive contribution to my professional development,” “the training held my interest,” and “the training was of a good standard.” Additionally, four open-ended questions assessed the most and least helpful elements of the training, whether participants would recommend the training to their colleagues, and any suggestions for improvement. Follow-Up Interview A subset of participants were invited to participate in a follow-up semistructured interview, to explore their experiences of using the training skills in practice. Details related to the recruitment and results of this phase of the research are reported elsewhere (Jones et al., 2018).

Data Analysis Random intercept models (RIM) were used to explore changes in the main outcome measures between time points. Consistent with previous research (Gask et al., 2006; Morriss et al., 1999), differences in attitude and confidence scores were analyzed at the item level, rather than overall score, to allow for discussion of individual item changes. As attitude scores were highly skewed, medians and interquartile ranges (25–75%) are presented. Given that the distribution of confidence scores was nor© 2018 Hogrefe Publishing

M. Ferguson et al., Suicide Prevention Training in Rural Australia

mal, means with standard deviations are used for descriptive purposes. Initially, full RIM models tested whether any change had occurred in attitudes and confidence across the four time points, controlling for participants’ age, years of service, and previous suicide prevention training, based on past research indicating the influence of these variables on attitude scores (Botega et al., 2007; Brunero et al., 2008; Herron et al., 2001). While some studies of health professionals have found an influence of profession on attitudes (Appleby et al., 2000; Brunero et al., 2008; Herron et al., 2001), the diversity of the current sample meant that there was insufficient power to control for this variable. Similarly, there is mixed evidence for the impact of previous experience with suicide (personal or professional) on attitudes (Botega et al., 2007; Brunero et al., 2008; Samuelsson & Åsberg, 2002); consequently, this was not controlled for here. As training was conducted in distinct rural regions, there may have been a degree of dependence in the training responses by location; accordingly robust standard errors were calculated for all analyses. Where change was detected, post hoc analyses were conducted, using RIM, to compare variable scores between all paired combinations of measurement points, again controlling for age, years of service, and previous training. Significance for all analyses was set at p < .05. Demographic details were analyzed descriptively, based on information provided by participants at either T1 or T2 (for those who did not complete the baseline survey). Program evaluation questions were also analyzed descriptively.

n = 345 potential participants registered their interest in attending training

Analyses were performed using Stata (version 14; StataCorp LP, TX).

Results Response Rate and Retention Initially, 345 participants expressed an interest in attending the training. Of these, 248 attended, with 213 participants completing the survey at T1, 236 at T2, 215 at T3, and 172 at T4 (see Figure 2 for completion rates and dropouts). It should be noted that not all participants completed all four assessments, with multiple participants withdrawing after T1 (often this was due to competing demands making them unavailable to attend the training), and a number of participants arriving at the training without registering and/or completing the T1 assessment. Further, some participants left the training session early, and therefore did not complete T3. The eight separate training sessions were conducted between August and November 2015. Session sizes varied, from 13 to 57 participants (M = 31 participants).

Participant Demographics As outlined in Table 1, participants were primarily female (82%), with a mean age of 43 years. Participants were of diverse professional backgrounds, with nursing (general-

n = 65 withdrew pre-T1: n = 32 decided not to continue (e.g., could no longer attend training) n = 30 did not respond after expressing interest n = 3 ineligible to participate in research

Figure 2. Flow of participants through the research process.

n = 213 completed T1 survey n = 30 withdrew post-T1 (did not attend training) n = 236 completed T2 survey

n = 215 completed T3 survey

n = 172 completed T4 survey

Crisis (2019), 40(1), 15–26

ist and mental health = 26%) and social work (17%) being most common. On average, participants had 8 years of work experience in their current field. The majority of participants either held an undergraduate (37%) or a non-university qualification (such as TAFE certificate; 31%). For health professionals specifically, current work environments varied, with 24% of participants identifying working in the community and 20% in a hospital. Further, 24% reported working in a specific mental health setting (e.g., mental health inpatient unit). Less than half (40%) of participants reported attending some form of suicide prevention education in the past. Previous experience with suicide was high: 77% of participants reporting at least one occurrence of the death or attempt of a family member, friend, or someone in their care, or their own suicide attempt.

Attitude Change As outlined in Table 2, there were no changes in 11 of the 14 attitude items between T1 and T2, indicating the stability of this measure. There were significant decreases in attitude scores on 11 items between T2 and T3, indicating a positive shift in attitudes on most items between pre- and immediately posttraining. Similarly, there were significant decreases in scores on the same 11 items between T1 and T4, again indicating a maintained positive shift in attitudes on most items between baseline and the 4-month follow-up. To further illustrate these changes, Figure 3 shows the percentage of participants who disagreed with each item at each measurement point. Of note, while there were minimal changes between measurement points on higher scoring items (e.g., Items 1, 2, 7, 8, 10, and 12) there were more notable changes in other items, such as an increase of approximately 30% of participants disagreeing with Item 11 – “I don’t feel comfortable assessing for suicide risk” – between T2 and T3, and an increase of approximately 20% of participants disagreeing with Item 4 – “Working with individuals at risk of suicide is rewarding” – from T2 to T3. Figure 3 also illustrates stronger disagreement with certain statements compared with others; while over 90% of participants disagreed with Items 1, 2, 10, and 12 at most measurement points, only a maximum of 30% of participants disagreed with Item 9 – “People have the right to take their own lives” – at any measurement point.

M. Ferguson et al., Suicide Prevention Training in Rural Australia

Table 1. Demographic details of participants who completed the T1 and T2 survey (n = 242) Characteristics Gender Male, n (%) Female, n (%) Age, M (SD)

As shown in Table 3, there were no changes in confidence scores between T1 and T2, again indicating stability of Crisis (2019), 40(1), 15–26

43 (17.8%) 199 (82.2%) 42.7 years (12.27 years)

Occupationa Aboriginal support/mental(health)/ case worker Counsellor

18 (7.4%) 5 (2.07%)

Mental health nurse

19 (7.9%)

Nurse

43 (17.8%)

Occupational therapist

8 (3.31%)

Psychologist

3 (1.2%)

School counsellor/behavioral support

4 (1.7%)

Social worker

40 (16.5%)

Teacher

11 (5.4%)

Youth worker/officer Case worker/manager Other Years in profession, M (SD)

7 (2.9%) 4 (1.7%) 77 (31.8%) 8.23 years (10.17 years)

Highest academic qualification Non-university qualification

76 (31.4%)

Undergraduate

89 (36.8%)

Postgraduate

29 (12.0%)

Other

48 (19.8%)

Work environment Hospital

48 (19.8%)

Community

59 (24.4%)

GP/primary care

8 (3.3%)

Other

30 (12.4%)

Non-health professional

96 (29.7%)

Works in mental health setting

57 (23.6%)

Had previous suicide prevention education

96 (39.7%)

Previous experience with suicideb Had a person in their care who died by suicide

18 (7.4%)

Had a person in their care who attempted suicide

84 (34.7%)

Had a family/friend who died by suicide Had a family/friend who attempted suicide

Confidence Change

n (%)

109 (45.0%) 88 (36.4%)

Personally attempted suicide

23 (9.5%)

No previous experience with suicide

55 (22.7%)

Note. aOccupation clusters include tertiary students. bParticipants could select multiple responses.

M. Ferguson et al., Suicide Prevention Training in Rural Australia

Table 2. Change in participant attitudes from T1 to T4 Mdn (IQR25-75) at each measurement point

Full model statistics

n = 213

n =236

n =215

n =172

1 (1–2)

1 (1–2)#

1 (1–2)*

2. Suicide prevention is not my responsibility

1 (1–2)

(1–2)#

1 (1–2)

1 (1–2)*^

−3.21

.001

3. Making more funds available to the appropriate health service would make no difference to the suicide rate

2 (1–2)

1 (1–2)

−1.21

.227

4. Working with individuals at risk of suicide is rewarding

2 (2–3)

2 (2–3)#

2 (1–2)*

2 (2–3)*^

−3.57

< .0001

5. If people are serious about ending their life they don’t tell anyone

2 (2–3)

2 (2–3)*#

2 (1–2)*

2 (1–2)*^

−11.32

< .0001

6. I feel defensive when other people offer advice about suicide prevention

2 (2–3)

2 (2–2)#

2 (1–2)*

−4.38

< .0001

7. It is easy for people not involved in clinical practice to make judgments about suicide prevention

2 (2–3)

2 (2–4)

2 (2–3)

−0.68

.494

8. If a person survives a suicide attempt, then this was a ploy for attention

2 (1–2)

2 (1–2)*#

1 (1–2)

1 (1–2)*#

−2.73

.006

9. People have the right to take their own lives

3 (3–4)

3 (2–4)

3 (3–4)

−0.36

.716

1 (1–2)*

2 (1–2)*^

−4.55

< .0001

1. I resent being asked to do more about suicide

(1–2)*#

1 (1–2)*

−3.03

.002

10. Since sociodemographic factors, such as employment and poverty, contribute to suicide, there is little an individual can do to prevent it

2 (1–2)

11. I don’t feel comfortable assessing for suicide risk

2 (2–3)

2 (2–3)#

2 (2–2)*

2 (1–2)*^

−6.64

< .0001

12. Suicide prevention measures are a drain on resources that would be more useful elsewhere

1 (1–2)

1 (1–2)#

1 (1–2)*

1 (1–2)*^

−3.40

.001

13. There is no way of knowing who is going to end their life by suicide

3 (2–4)

3 (2–4)#

2 (2–4)*

2 (2–3)*#^

−3.23

.001

14. What proportion of suicides do you consider preventable?

2 (2–4)

2 (2–4)#

2 (2–2)*

2 (2–3)*^

−5.26

< .0001

Note. T1 = Baseline. T2 = Pretraining. T3 = Posttraining. T4 = 4-Month follow-up. *Statistically different from T1. ^Statistically different from T2. #Statistically different from T3. Bold = post hoc comparisons not conducted.

1 - I resent being asked to do more about suicide 2 - Suicide prevention is not my responsibility 3 - Making more funds available to the appropriate health service would make no difference to the suicide rate 4 - Working with individual's at risk of suicide is rewarding 5 - If people are serious about ending their life they don't tell anyone

Attitude items

6 - I feel defensive when people offer advice about suicide prevention 7 - It is easy for people not involved in clinical practice to make judgements about suicide prevention

T1 T2

8 - If a person survives a suicide attempt, then this was a ploy for attention

T3 9 - People have the right to take their own lives

10 - Since socio-demographic factors, such as employment and poverty, contribute to suicide, there is little an individual can do to prevent it 11 - I don't feel comfortable assessing for suicide risk 12 - Suicide prevention measures are a drain on resources which would be more useful elsewhere 13 - There is no way of knowing who is going to end their life by suicide 14 - What proportion of suicides do you consider preventable? 0

100

% of participants

Figure 3. Percentage of participants who disagreed with each Attitude to Suicide Prevention Scale item from T1 to T4 (strongly disagree and dis

Figure 3. responses Percentage combined). of participants who disagreed with each Attitude to Suicide Prevention Scale item from T1 to T4 (strongly disagree and agree disagree responses combined). © 2018 Hogrefe Publishing

[Frau Cichos: In Punkt 14 wäre es möglich am Ende des Satzes ein Fragezeichen einzufügen?]

Crisis (2019), 40(1), 15–26

M. Ferguson et al., Suicide Prevention Training in Rural Australia

Table 3. Change in participant confidence from T1 to T4 M (SD) at each measurement point T1

Full model statistics

n = 213

n =236

n =215

n =172

I am confident that I have the skills to use my time well with individuals at risk of suicide

4.93 (2.49)#

4.88 (2.38)#

6.75 (1.81)*

6.86 (1.83)*

12.45

< .0001

After working with an individual once I would be confident that I could recognize potential suicide risk

4.15 (2.36)#

4.40 (2.24)#

6.01 (2.11)*

6.14 (1.90)*

12.45

< .0001

I am confident that I could differentiate mild depression from suicide risk

4.49 (2.41)#

4.69 (2.29)#

6.02 (2.05)*

6.44 (1.98)*#

11.32

< .0001

I am confident in dealing with the needs of individuals at risk of suicide

4.47 (2.46)#

4.55 (2.38)#

6.37 (1.96)*

6.57 (1.89)*

19.23

< .0001

Note. T1 = Baseline. T2 = Pretraining. T3 = Posttraining. T4 = 4-Month follow-up. *Statistically different from T1. #Statistically different from T3.

this measure. There were significant increases in confidence on all four items between T2 and T3, indicating improvements between pre- and immediately posttraining. Further, similar significant increases were seen on all items between T1 and T4, indicating maintained improvements in confidence between baseline and the 4-month follow-up.

Program Evaluation A total of 215 participants completed the T3 questionnaire (91% of T2 participants). Feedback regarding the program was positive: 97% (n = 208) of participants agreed (either agreed or strongly agreed) that the program was relevant to their role, made a positive contribution to their professional development, and was of a good standard; 96% (n = 206) agreed that the training held their interest; and 91% (n = 195) of participants stated that they would recommend the training to a colleague. The majority of participants (n = 177; 82%) made a comment about the most helpful element/s of the day. Many participants reported multiple aspects of the training as being helpful, particularly the lived experience reflections. By contrast, only 41 participants (19%) made a comment about the least helpful element/s of the day. There was notable variation in these comments, with the majority relating to specific elements with which participants already felt sufficiently knowledgeable. Other comments related to the training venues, food, and noise levels.

Discussion The present study developed and evaluated a community- based suicide prevention education program for health Crisis (2019), 40(1), 15–26

and human service professionals in rural and remote South Australia. Positive improvements in participant-reported attitudes and confidence associated with the training were found, both immediately after training and at short-term follow-up. Combined, these findings indicate that a brief, multidisciplinary training program can support rural and remote communities to acquire attitudes and confidence to better enable them to support people who are vulnerable to suicide.

Changes in Attitudes There was an improvement in 11 of the 14 attitude items immediately posttraining, and this was maintained at the 4-month follow-up. Similar improvements in attitudes have been found using the Attitude to Suicide Prevention scale in studies of various health professionals, both in Australia (Brunero et al., 2008) and in the UK (Appleby et al., 2000; Gask et al., 2006, 2008; Herron et al., 2001), as well as in studies using other measures of short-term attitude change (e.g., the Suicide Behavior Attitude Questionnaire, Botega et al., 2007; Santos, Simões, de Azevedo Erse, Façanha, & Marques, 2014; the Suicide Opinion Questionnaire, Chan, Chien, & Tso, 2009; and the Understanding Suicidal Patients Scale, Kishi et al., 2014). It should be noted, however, that there was no significant change in three of the attitude items: Item 3 – “Making more funds available to the appropriate health service would make no difference to the suicide rate,” Item 7 – “It is easy for people not involved in clinical practice to make judgments about suicide prevention,” and Item 9 – “People have the right to take their own lives.” While attitudes were relatively positive for the first statement (median of 2/5 at T1 and T2, and 1 at T3 and T4), indicating a potential ceiling effect, the lack of change in the other items may be attributed to these being heavily ingrained attitudes © 2018 Hogrefe Publishing

M. Ferguson et al., Suicide Prevention Training in Rural Australia

that require more than a 1-day training session to shift. In particular, the lack of change for the third item is consistent with past literature. In Herron and colleagues’ (2001) validation study of the Attitudes Towards Suicide Prevention scale, four groups of UK health professionals (GPs, community psychiatric nurses, accident and emergency nursing staff, and junior psychiatrists; n = 218) all scored a mean of 3.03–3.61 (out of 5) on this particular item. Further, in two studies by Gask and colleagues, which primarily involved nurses (n = 458, Gask et al., 2006; n = 203, Gask et al., 2008), there was no change in this item immediately posttraining, or at the 4-month follow-up, with mean scores ranging between 3.01 and 3.42 at any one measurement point. In the present study, the median score remained at 3 for all four time points, indicating ambivalence. This may be attributed to the interpretation of this item. Some participants who were health professionals informally discussed this item with the researchers at the training sessions, and noted that they were unsure whether it related to suicide specifically or to euthanasia, and expressed the view that as health professionals it is their role to save lives. While it is important for health professionals to be nonjudgmental in their approach, these views may help to explain the lack of change in this item.

Changes in Confidence Self-reported confidence in working with people vulnerable to suicide increased for all four items in the present study (approximate median score of 4.5/11 at T1 to 6.5/11 at T3 and T4). Unlike some of the attitude items, increases in all confidence items were sustained at follow-up, further indicating the value of brief educational interventions for impacting this domain. These improvements are consistent with past literature employing the same measure with health professionals, both immediately before and after brief training in the UK (Appleby et al., 2000; Morriss et al., 1999), and at the 4–6-month follow-up (Gask et al., 2006, 2008). Similar improvements have also been found using other measures of confidence among non-health professional samples. Jones and colleagues’ (2015) evaluation of the CORES training in rural Tasmania, Australia, found increased community members’ self-reported comfort when talking to others about suicide and confidence in ability to provide appropriate assistance after brief training. Improvements in confidence may be attributed to a range of factors. In particular, numerous participants noted the value of the lived experience component of the program, both in the formal evaluation and in informal discussions at the training sessions, as well as in follow-up, semistructured interviews (Jones et al., 2018). This ele© 2018 Hogrefe Publishing

ment involved a local community member discussing his experiences with suicidal behavior, and the value of small gestures in saving his life, such as an individual simply taking the time to ask him if he was OK. Therefore, in addition to the formal knowledge and skills developed during the training, perhaps some of the complexities of suicide were demystified in the present training, leading to an improvement in attendee confidence. Further, a number of participants informally mentioned that the training was a timely opportunity to reinforce what they are already doing well, which might also contribute to increased confidence.

Program Evaluation Evaluation of the training at T3 provided valuable information about participants’ experiences with the training program. There was strong agreement that the training was relevant to participants’ roles, made a positive contribution to their professional development, held their interest, and was of a good standard. In support, almost 90% of participants said they would recommend the training to a colleague. Participants also shed light on which elements of the training were most beneficial, with the lived experience reflections being noted most frequently. This provides important information for future training programs, complementing the attitude and confidence measures to indicate that a brief training program is perceived as beneficial, and that efforts should be made to incorporate lived experience components.

Strengths and Limitations A major strength of this study is that, to the best of the authors’ knowledge, it is the first to evaluate the impact of a community-based, suicide prevention education program for health and human service professionals in rural and remote Australia through a rigorous, repeated-measures survey. This extends the evidence base for effective suicide prevention strategies for regional Australia, where suicide rates exceed those in major cities, and which therefore require unique attention. Another strength is the noteworthy follow-up response rate of approximately 70% at T4. This is large in comparison with similar studies. Gask and colleagues (2006, 2008) achieved a response rate of approximately 30% at follow-up in their studies following brief suicide prevention education programs for health professionals, perhaps attributed to the use of postal surveys, rather than the Web-based approach used here. The high response rate in the present study reduces the possibility of positive response bias, allows for greater confidence in the stability of attitude and confidence change associated Crisis (2019), 40(1), 15–26

with the training program, and permits generalizability to the initial sample. Limitations of the study should also be acknowledged. Firstly, despite repeated-measures design, with stable responses to attitude and confidence outcome measures between T1 and T2, this study did not include a control group, and therefore conclusions regarding the impact of the program should be viewed cautiously. However, it should be noted that although considered the gold standard, a randomized controlled trial was deemed impractical, given the community-based nature of this training and the ethical imperative of ensuring that it reached as many people as possible (the project budget only allowed for the included number of training sessions and locations, making a waitlist control group not feasible). Further, while the study examined changes in the self-reported constructs of attitudes and confidence, which are believed to influence behavior (Aldrich, Harrington, & Cerel, 2014), the present findings cannot be used to make inferences about the impacts of training in practice. To address this, a series of semistructured interviews were conducted with a subset of participants. Applicability of outcome measures should also be considered. While the study relied on outcome measures commonly used with health professional populations, these may not apply to all participants in the current sample. However, given the absence of validated suicide prevention outcome measures for all professional groups included in the study, those which were applicable to the majority (i.e., health professionals) were deemed most appropriate. Further, the wording of some items was adjusted, where necessary, to reflect the broad sample and the various places within which they work. However, the small magnitude of these adjustments (i.e., minor wording changes) is unlikely to have impacted the psychometric properties of the scales. Lastly, the heterogeneity of the current sample means that results cannot be generalized to specific, individual professions.

Future Directions Findings support the value of a brief, community-based suicide prevention training program for health and human service professionals in rural Australia. This adds to the growing body of evidence and policy supporting the role of targeted education as an important suicide prevention strategy in these regions. In light of the low self-reported attendance at suicide prevention training in the past, and comments made by participants regarding the desire for more training, this study highlights the need for financial support for training in these areas. It also suggests that training should be prioritized in undergraduate education, to enable emerging professionals to enter the workforce Crisis (2019), 40(1), 15–26

M. Ferguson et al., Suicide Prevention Training in Rural Australia

with a baseline set of skills in suicide prevention. The financial burden of this form of suicide prevention is low, particularly with regard to the potential savings. Although somewhat dated, Appleby et al. (2000) predicted that if their UK-based brief training program, which was similar to the current program, could reduce the suicide rate by 2.5%, this would equate to a saving of in excess of £99,000 per suicide prevented, plus a saving of over £3,300 per life year gained. While the findings of improvements in attitudes and confidence contribute to a growing body of literature in favor of suicide prevention education programs (Ferguson et al., 2017), future research needs to extend this by considering the impact that such training has both in clinical practice (e.g., case note analysis to assess use of training skills) and on those individuals experiencing suicidal states who have encountered a health or human service professional who has undergone advanced training. Further, it will be important to determine which elements of training programs are linked to positive participant outcomes. In the present study, some insight was gained through open-ended questions, such as the value of lived-experience reflections. This knowledge will help to refine the content of future training programs, to ensure that it is targeting the real-world needs of professionals and the people they work with. Finally, strong support for improvements in confidence was found in this study. In addition to the need to explore how this might translate to changes in practice, specifically at the point of direct care, another important future direction is the role that this change may play in trainee well-being. Seguin and coworkers’ (2014) review describes the profound impact suicide can have on health professionals, and therefore by improving confidence in their ability to work with individuals at risk of suicide, training programs may decrease the impact of suicide and improve well-being.

Conclusion The findings from the present study indicate that a brief educational intervention can contribute to improvements in health and human service professionals’ attitudes and confidence when working with people vulnerable to suicide in rural and remote South Australian communities, both immediately and in the short-term. Future studies should investigate longer-term changes in practice, including associated outcomes for people experiencing suicide-related distress. Acknowledgments This project was funded by the University of South Australia Department of Rural Health. © 2018 Hogrefe Publishing

M. Ferguson et al., Suicide Prevention Training in Rural Australia

The authors would like to thank Mr. Michael Marsh for contributing to the training team and sharing his lived experience; Ms. Rebecca Shammas for contributing to the training development and delivery; Ms. Bronwyn Ryan for contributing to planning the training program; and Ms. Hayley Colyer and Ms. Melissa Gibson for assisting with training administration.

References Aldrich, R. S., Harrington, N. G., & Cerel, J. (2014). The Willingness to Intervene Against Suicide Questionnaire. Death Studies, 38, 100–108. https://doi.org/10.1080/07481187.2012.738763 Appleby, L., Morriss, R., Gask, L., Roland, M., Lewis, B., Perry, A., … Faragher, B. (2000). An educational intervention for front-line health professionals in the assessment and management of suicidal patients (The STORM Project). Psychological Medicine, 30, 805–812. Australian Bureau of Statistics. (2011). Australian social trends, March 2011, Health outside major cities, cat. no. 4102.0. Canberra, Australia: Commonwealth of Australia. Australian Bureau of Statistics. (2017a). Causes of death, Australia, 2016, cat. no. 3303.0. Canberra, Australia: Commonwealth of Australia. Australian Bureau of Statistics. (2017b). Regional population growth, Australia, 2014-2015, cat. no. 3218.0. Canberra, Australia: Commonwealth of Australia. Botega, N. J., Silva, S. V., Reginato, D. G., Rapeli, C. B., Cais, C. F., Mauro, M. L., … Cecconi, J. P. (2007). Maintained attitudinal changes in nursing personnel after a brief training on suicide prevention. Suicide and Life-Threatening Behavior, 37, 145–153. https://doi.org/10.1521/suli.2007.37.2.145 Brunero, S., Smith, J., Bates, E., & Fairbrother, G. (2008). Health professionals’ attitudes towards suicide prevention initiatives. Journal of Psychiatric and Mental Health Nursing, 15, 588–594. https://doi.org/10.1111/j.1365-2850.2008.01278.x Cerel, J., Jordan, J. R., & Duberstein, P. R. (2008). The impact of suicide on the family. Crisis, 29, 38–44. https://doi. org/10.1027/0227-5910.29.1.38 Chan, S. W., Chien, W. T., & Tso, S. (2009). Evaluating nurses’ knowledge, attitude and competency after an education programme on suicide prevention. Nurse Education Today, 29, 763–769. https://doi.org/10.1016/j.nedt.2009.03.013 Clark, S. E., & Goldney, R. D. (2000). The impact of suicide on relatives and friends. In K. Hawton, & K. van Heeringen (Eds.), The international handbook of suicide and attempted suicide (pp. 467–484). West Sussex, UK: John Wiley & Sons. Clifford, A. C., Doran, C. M., & Tsey, K. (2013). A systematic review of suicide prevention interventions targeting indigenous peoples in Australia, United States, Canada and New Zealand. BMC Public Health, 13. Retrieved from https://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-13-463 Cutcliffe, J. R., Mckenna, H., Keeney, S., Stevenson, C., & Jordan, J. (2013). ‘Straight from the horse’s mouth’: Rethinking and reconfiguring services in Northern Ireland in response to suicidal young men. Journal of Psychiatric and Mental Health Nursing, 20, 466–472. https://doi.org/10.1111/jpm.12012 Cutcliffe, J. R., Stevenson, C., Jackson, S., & Smith, P. (2006). A modified grounded theory study of how psychiatric nurses work with suicidal people. International Journal of Nursing Studies, 43, 791–802. https://doi.org/10.1016/j.ijnurstu.2005.09.001 © 2018 Hogrefe Publishing

De Leo, D., Draper, B. M., Snowdon, J., & Kolves, K. (2013). Contacts with health professionals before suicide: Missed opportunities for prevention? Comprehensive Psychiatry, 54, 1117–1123. https://doi.org/10.1016/j.comppsych.2013.05.007 Department of Health and Ageing. (2008). Living Is For Everyone (LIFE) Framework (2007). Canberra, Australia: Commonwealth of Australia. Ferguson, M., Reis, J., Rabbetts, L., Ashby, H. J., Bayes, M., McCracken, T., Ross, C., & Procter, N. (2017). The effectiveness of suicide prevention programs for nurses: A systematic review. Crisis. https://doi.org/10.1027/0227-5910/a000479 Gask, L., Dixon, C., Morriss, R., Appleby, L., & Green, G. (2006). Evaluating STORM skills training for managing people at risk of suicide. Journal of Advanced Nursing, 54, 739–750. https://doi. org/10.1111/j.1365-2648.2006.03875.x Gask, L., Lever-Green, G., & Hays, R. (2008). Dissemination and implementation of suicide prevention training in one Scottish region. BMC Health Services Research, 8. Retrieved from https:// bmchealthservres.biomedcentral.com/articles/10.1186/ 1472-6963-8-246 Herron, J., Ticehurst, H., Appleby, L., Perry, A., & Cordingley, L. (2001). Attitudes towards suicide prevention in front-line health staff. Suicide and Life-Threatening Behavior, 31, 342–347. Isaac, M., Elias, B., Katz, L. Y., Belik, S. -L., Deane, F. P., Enns, M. W., ... The Swampy Cree Suicide Prevention Team (12 members). (2009). Gatekeeper training as a preventative intervention for suicide: A systematic review. The Canadian Journal of Psychiatry, 54, 260–268. https://doi.org/10.1177/070674370905400407 Johnston, A. K., Pirkis, J. E., & Burgess, P. M. (2009). Suicidal thoughts and behaviours among Australian adults: Findings from the 2007 National Survey of Mental Health and Wellbeing. Australian and New Zealand Journal of Psychiatry, 43, 635–643. https://doi.org/10.1080/00048670902970874 Jones, M., Ferguson, M., Walsh, S., Martinez, L., Cronin, C., & Procter, N. (2018). Perspectives of rural health and human service practitioners following suicide prevention training program in Australia: A thematic analysis. Health and Social Care in the Community, 26(3). https://doi.org/10.1111/hsc.12532 Jones, S., Walker, C., Miles, A. C. J., De Silva, E., & Zimitat, C. (2015). A rural, community-based suicide awareness and intervention program. Rural and Remote Health, 15, 2972. Retrieved from https://www.rrh.org.au/journal/article/2972 Kennedy, A., Maple, M. J., Mckay, K., & Brumby, S. A. (2014). Suicide and accidental death in Australia’s farming communities: A review of the literature. Rural and Remote Health, 14, 2517. Retrieved from https://www.rrh.org.au/journal/article/2517 Kinchin, I., & Doran, C. M. (2017). The economic cost of suicide and non-fatal suicide behavior in the Australian workforce and the potential impact of a workplace suicide prevention strategy. International Journal of Environmental Research and Public Health, 14. Retrieved from http://www.mdpi.com/16604601/14/4/347 Kishi, Y., Otsuka, K., Akiyama, K., Yamada, T., Sakamoto, Y., Yanagisawa, Y., … Thurber, S. (2014). Effects of a training workshop on suicide prevention among emergency room nurses. Crisis, 35, 357–361. https://doi.org/10.1027/0227-5910/a000268 Morriss, R., Gask, L., Battersby, L., Francheschini, A., & Robson, M. (1999). Teaching front-line health and voluntary workers to assess and manage suicidal patients. Journal of Affective Disorders, 52, 77–83. Robinson, J., Cox, G., Malone, A., Williamson, M., Baldwin, G., Fletcher, K., & O’Brien, M. (2013). A systematic review of school-based interventions aimed at preventing, treating, and responding to suicide-related behavior in young people. Crisis, 34, 164–182. https://doi.org/10.1027/0227-5910/a000168 Crisis (2019), 40(1), 15–26

Samuelsson, M., & Åsberg, M. (2002). Training program in suicide prevention for psychiatric nursing personnel enhance attitudes to attempted suicide patients. International Journal of Nursing Studies, 39, 115–121. Santos, J. C., Simões, R. M. P., De Azevedo Erse, M. P. Q., Façanha, J. D. N., & Marques, L. A. F. A. (2014). Impact of “+contigo” training on the knowledge and attitudes of health care professionals about suicide. Revista Latino-Americana de Enfermagem, 22, 679–684. Seguin, M., Bordeleau, V., Drouin, M. -S., Castelli-Dransart, D. A., & Giasson, F. (2014). Professionals’ reactions following a patient’s suicide: Review and future investigation. Archives of Suicide Research, 18, 340–362. https://doi.org/10.1080/13811118.2013.8 33151 Stark, C., & O’Riordan, V. (2011). Rurality and suicide. In C. O’Connor, S. Platt, & J. Gordon (Eds.), International handbook of suicide prevention: Research, policy and practice (1st ed., pp. 253–273). London, UK: John Wiley & Sons. World Health Organization. (2014). Preventing suicide: A global imperative. Geneva, Switzerland: Author.

Received October 6, 2017 Revision received December 14, 2017 Accepted December 15, 2017 Published online June 21, 2018 Dr. Monika Ferguson is a lecturer in mental health and a member of the Mental Health and Suicide Prevention Research Group at the University of South Australia. Her current program of research focuses on the evaluation of suicide prevention strategies for health professionals and the wider community. Associate Professor Jim Dollman is an exercise scientist who has lectured and researched at the University of South Australia since 1993. He has expertise in physical activity promotion, and is currently focused on improving support-seeking for health in rural adults who are challenged by relatively low access to appropriate support.

Crisis (2019), 40(1), 15–26

M. Ferguson et al., Suicide Prevention Training in Rural Australia

Associate Professor Martin Jones is Director, University of South Australia, Department of Rural Health. He previously worked for the NHS for 28 years as a mental health nurse, manager, and executive. His research interests include increasing workforce efficacy in adherence therapy and supporting people living with comorbidity, and supporting nurses to assume prescribing responsibilities. Kathryn Cronin has worked as a mental health nurse in a variety of Australian settings and is now Senior Mental Health Nurse for Country Health SA. Kathryn has a strong commitment to people accessing consumer-focused services that are provided by a skilled, connected, and sustainable workforce. Lynne James is Principal Project Officer, Suicide Prevention, Office of the Chief Psychiatrist, SA Health, and Adjunct Lecturer at the University of South Australia. Lynne plays a key role in suicide prevention policy development, and also works closely with communities to establish Suicide Prevention Networks across South Australia. Lee Martinez is Mental Health Academic, University of South Australia, Department of Rural Health. Lee works with community to improve the health and well-being of rural people, focusing on increasing the rural mental health workforce. Lee’s current research explores the voice of mental health lived experience in service delivery and education. Professor Nicholas Procter is the University of South Australia’s Inaugural Chair in Mental Health Nursing, and convener of the Mental Health and Suicide Prevention Research Group, located within the Sansom Institute for Health Research. His research interests span a range of mental health topics as well as suicide prevention.

Monika Ferguson School of Nursing and Midwifery GPO Box 2471 Adelaide, South Australia 5001 Australia monika.ferguson@unisa.edu.au

Research Trends

A Comparison of Attitudes Toward Suicide Among Individuals With and Without Suicidal Thoughts and Suicide Attempts in South Korea Sang-Uk Lee1,2, Mina Jeon3, and Jong-Ik Park4 Department of Mental Health Research, National Center for Mental Health, Seoul, South Korea Department of Preventive Medicine, School of Medicine, Kyung Hee University, Seoul, South Korea 3 Department of Psychology and Human Development, UCL Institute of Education, London, UK 4 Department of Psychiatry, Kangwon National University School of Medicine, Chunchon, South Korea 1 2

Abstract. Background: A suicidal person can go through different stages that include suicidal thoughts and suicide attempts. For a few individuals, these can end up with suicide. However, there have been no studies investigating any differences in attitudes toward suicides among individuals with no suicidal thoughts, those with suicidal thoughts, and those with suicide attempts. Aims: This study was carried out to compare attitudes toward suicide among three different groups: individuals with a history of no suicidal thoughts, those with a history of suicidal thoughts, and those with a history of suicide attempts. Method: To examine Koreans’ attitudes toward suicide, we analyzed the data from the 2013 National Suicide Survey involving 1,500 participants aged between 19 and 75 years. Results: Different attitudes toward suicide were found among the three groups. Persons reporting that they had made a suicide attempt in their life showed the most permissive attitudes toward suicide. Limitations: Since this research is based on cross-sectional data, it is difficult to eliminate the possibility of changes in attitude toward suicide completely after having a suicidal thought and suicide attempt. Conclusion: These results can be a useful source for constructing effective messages for suicide prevention campaigns and can ultimately contribute to an improvement in the public’s perceptions of suicide in the future. Keywords: attitudes toward suicide, suicidal thoughts, suicide attempts

The World Health Organization (WHO) reported that approximately 800,000 people die by suicide annually (WHO, 2014). Therefore, suicide is recognized as a global issue and the WHO announced the Mental Health Action Plan that aimed to reduce suicide rates by 10% within 7 years from 2013 to 2020 (WHO, 2013). In particular, South Korea has maintained the highest suicide rate for the past 10 years among Organisation for Economic Co-operation and Development countries (OECD, 2014). In 2013, the suicide rate was at 28.5 per 100,000 people, which was approximately two- to fivefold higher than other OECD countries (OECD, 2014), making suicide prevention even more urgent in South Korea. Those who are suicidal seem to go through a set of stages, which includes suicidal thoughts, suicide attempts, and, sometimes, suicide (Sadock, 2009). Although suicidal ideation does not necessarily result in suicide, it possibly © 2018 Hogrefe Publishing

increases the risk of suicide attempts or suicide (Galfalvy, Oquendo, & Mann, 2008; Harriss & Hawton, 2005; Harriss, Hawton, & Zahl, 2005; Kim et al., 2011; Stefansson, Nordström, & Jokinen, 2012). Those who have attempted suicide are also considered to be in a high-risk group (Beck & Robert, 1989; Chun, Park, & Lee, 2007; Ghio et al., 2011) since a history of suicide attempts is reported to be the most important predictor of suicide (Goldstein, Black, Nasrallah, & Winokur, 1991; Sinclair, Harriss, Baldwin, & King, 2005; Suominen et al., 2004). Recent studies have found that individuals’ attitude toward suicide is another important factor associated with the risk of suicide. For instance, holding approving or permissive attitudes toward suicide was significantly associated with suicidal ideation, suicide attempts, and increased mortality from suicide (Gil, 2005; Kim & Yoon, 2011; McAuliffe, Corcoran, Keeley, & Perry, 2003; Stein, Brom, Elizur, & WitCrisis (2019), 40(1), 27–35 https://doi.org/10.1027/0227-5910/a000528

ztum, 1998; Zemaitiene, & Zaborskis, 2005). Although previous studies have explored attitudes toward suicide in people with a history of suicidal thoughts and suicide attempts, none of these studies has compared different attitudes toward suicides among individuals with no suicidal thoughts, those with suicidal thoughts, and those with suicide attempts. However, identification of specific attitudes toward suicide that are more associated with the risk of suicide thoughts or suicide attempts can be useful for suicide prevention as suicidal ideation and suicide attempts are not the same phenomenon (Chung et al., 2010; Kim & Lee, 2014) and simply having suicidal thoughts does not always result in suicide attempts (Jeong & Seo, 2014). Moreover, the results of previous studies on attitudes toward suicide have not been effectively applied to suicide prevention programs or campaigns since the findings from most of these studies were presented as a number of key factors that can summarize various items in attitudes toward suicide questionnaires, for example, acceptability, preventability, and tabooing (Aldrich & Cerel, 2009; Kim & Yoon, 2011; Ji, Hong, & Lee, 2016). This method of presenting the results does not tell us which items should be considered first to reduce such factors compared with other variables. For this reason, it is difficult to develop effective messages for suicide prevention. Therefore, this study compared the attitudes toward suicide among three different groups and investigated particular attitudes associated with the experience of suicidal thoughts or suicide attempts.

Method Ethics Statement The current study was approved by the Kangwon University Hospital’s Institutional Review Board (Approval Number: KNUH-2013-06-007-001).

Procedure Data from the 2013 National Suicide Survey were used in the present study (Kim et al., 2015). In order to obtain a sample population that best represents the entire population of South Korea, the study employed a stratified regional sample of the population in 2010 based on the housing census report. The initial round of stratification produced 16 provinces and metro-cities across the nation. The second round involved the classification of the population using the housing census report and enumeration of districts, ensuring that they were prorated to the population of the Crisis (2019), 40(1), 27–35

S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

enumerated districts. Through this method, 150 sample districts were extracted. From each of the sample districts, a sample of 10 households was randomly selected on the basis of house addresses, followed by the selection of one member of each household for the study. Ultimately, 1,500 men and women aged 19–75 years, residing in 16 different metropolises, were randomly selected based on home addresses. Interviewers visited the selected houses and conducted one-on-one structured interviews using a standardized questionnaire. Interviewers were professionals who received training from a survey institution. Selected participants were replaced by other participants when they refused to take part in the study or when interviewers could not meet the participants despite visiting them four times. Interviewers visited 4,399 houses and 1,500 participants were interviewed. The response rate was 34.1%.

Measurements To examine participants’ attitudes toward suicide, the present study used Renberg’s Attitude Towards Suicide (ATTS) questionnaire (Renberg, 2011). Although there are various assessment tools for attitudes toward suicide, the ATTS questionnaire was chosen for this study as it has been used in many different countries (Kodaka, Poštuvan, Inagaki, & Yamada, 2011). When the ATTS was first implemented in 1996, the questionnaire consisted of 20 items designed to evaluate general attitudes toward suicide. Later, the questionnaire was modified to include 34 items by adding personal items regarding suicide (Renberg & Jacobsson, 2003). The present study utilized the ATTS questionnaire developed in 2011, which comprises 40 items including three optional items, to survey community attitudes toward suicide (Renberg, 2011). We excluded three optional items and only utilized 37 items of the ATTS. The questionnaire, which was received via e-mail from the author of the ATTS, was translated into Korean by one of the researchers of the present study. Subsequently, it was back-translated by a Korean–American psychology student to minimize any deviations from the original questionnaire. Finally, the questionnaire underwent a final inspection by a psychiatrist and suicide prevention program practitioners. It used a 5-point Likert scale that ranged from 1 (strongly disagree) to 5 (strongly agree).

Data Analysis The participants were divided into three groups depending on experiences of suicidal ideation and suicide attempts ever in life. Individuals who had thought about suicide seriously were classified as individuals with suicidal thoughts, © 2018 Hogrefe Publishing

S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

while those who had attempted suicide were considered as people with suicide attempts. Lastly, individuals who had never thought about or attempted suicide were classified as individuals with no suicidal thoughts. Descriptive statistics were calculated using the mean, standard deviation for continuous variables, and percentages for categorical variables. Analysis of variance (ANOVA) was performed to identify the differences in attitudes toward suicide among the three groups. Post hoc comparisons were also conducted to examine significant differences between the three groups in each ATTS item. As per convention, the statistical significance level for ANOVA was two-sided at p < .01. Multinomial logistic regression analyses were performed to identify the association between each ATTS item and the experience of suicidal thoughts or suicide attempts ever in life. Using the group of no suicidal ideation as the comparison group, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Additional logistic regression was also performed to identify the odds ratios by comparing individuals with suicidal thoughts and those with suicide attempts in terms of ATTS scores. The statistical significance level for logistic regression was two-sided at p < .05. All analyses were performed with SPSS version 21 (IBM SPSS Inc., Chicago, IL).

Results The majority of the 1,500 participants did not experience suicidal thoughts (N = 1,165, 77.7%), but 301 (20.0%) had a history of suicidal thoughts and 34 (2.3%) had a history of suicide attempts. Sociodemographic characteristics of participants such as age, gender, education levels, marital status, and income are presented in Table 1. The ANOVA results showed significant differences between the three groups regarding responses on 17 out of 37 attitude items (Table 2). The multinomial logistic regression analysis confirmed that individuals with suicidal thoughts and suicide attempts were more likely to have permissive attitudes toward suicide than were individuals with no such experiences, even after controlling for several variables (Table 3). Moreover, three patterns of differences were revealed from 17 items. First, both participants with a history of suicide ideations and those with suicide attempts scored significantly higher than those without suicidal thoughts: no suicidal thoughts (a) < suicidal thoughts (b), suicide attempts (c). This is shown in 10 items, including: (1) “Suicide is an acceptable means to terminate an incurable disease”; (3) “When a person commits suicide, it is something that he/she has considered for a long time”; (4) “Loneliness could for me be a reason to take my life”; (5) “Almost everyone has at one time or another thought about

Table 1. Demographic characteristics of the three groups

Age

Gender

No suicidal thoughts

Suicidal thoughts

Suicide attempts

(N = 1,165)

(N = 301)

(N = 34)

45.9

±15.4

46.1

±15.2

43.6

±11.5

(N)

(%)

(N)

(%)

(N)

(%)

Male

550

47.2

133

44.2

35.8

Female

615

52.8

168

55.8

64.7

Geographical location

Metropolitan city

540

46.4

122

40.5

47.1

Province

625

53.6

179

59.5

52.9

Education background

Below middle school

224

20.9

23.6

20.6

Civil status

Income (million Won)

High school

470

40.3

133

44.2

44.1

Above university

451

38.7

32.2

35.3

Single

237

20.3

21.9

17.6

Married

845

72.5

202

67.1

67.6

Divorced/bereaved

7.1

11.0

14.7

Below 249

457

39.3

141

46.8

47.1

250–399

416

35.8

27.2

29.4

Above 400

290

24.9

25.9

23.5

Note. aExchange rate in October 2013, 1 Won = US $0.001.

Crisis (2019), 40(1), 27–35

S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

Table 2. Results of ANOVA on attitude toward suicide ATTS

Categories

15. People who talk about suicide do not necessarily commit suicide.

No suicidal thoughts

3.73

0.80

Suicidal thoughts

3.94

0.70

Suicide attempts

3.62

1.04

No suicidal thoughts

2.61

0.98

Suicidal thoughts

2.74

1.02

Suicide attempts Figure 1. The results of logistic regression analysis of attitude toward suicide between those

3.26

1.40

16. Most suicide attempts are caused by conflicts with a close person.

with suicidal thoughts andhelp those with suicide attempts. Question number of ATTS matches 17. I would like to get to commit suicide if I were to suffer from a No suicidal thoughts 2.90 1.02 the item number in Tabledisease. 3. ATTS = Attitude Towards Suicide questionnaire. OR = odds severe, incurable Suicidal thoughts 3.16 1.01 ratio. Reference = individuals with a history of suicidal thoughts [author: "individuals with Suicide attempts 3.65 0.98 no history…"?]. Adjusted for age, gender, geographical location, education levels, marital status,Note. and income. ATTS = Attitude Towards Suicide questionnaire. a = No suicidal thoughts. b = Suicidal thoughts. c = Suicide attempts.

Post hoc

9.29**

a<b

8.70**

a<c

15.30**

a<b<c

a Reverse scoring question. *p < .01. **p < .001.

Figure 1. The results of logistic regression analysis of attitude toward suicide between those with suicidal thoughts and those with suicide attempts.

Note: Question number of ATTS matches the item number in Table 3. ATTS = Attitude Towards Suicide questionnaire. OR = odds ratio. Reference = individuals with a history of suicidal thoughts. Adjusted for age, gender, geographical location, education levels, marital status, and income.

A d ju s t e d O R

0 1

Q u e s t io n n u m b e r o f A T T S

suicide”; (7) “Suicide can sometimes be a relief for those involved”; (8) “I would consider the possibility of taking my life if I were to suffer from a severe, incurable disease”; (10) “On the whole, I do not understand how people can take their lives (reverse coding item)”; (13) “Anybody can commit suicide”; (14) “I can understand that people suffering from a severe, incurable disease commit suicide”; and (17) “I would like to get help to commit suicide if I were to suffer from a severe incurable disease.” Second, only suicide attempters scored significantly higher than those without suicidal thoughts (a < c) in five items, for example: (2) “Many suicide attempts are made because of revenge or to punish someone else”; (6) “I could say that I would take my life without actually meaning it”; (9) “Once a person has suicidal thoughts, he/she will never let them © 2018 Hogrefe Publishing

go”; (12) “A person suffering from a severe, incurable disease expressing wishes to die should get help to do so”; and (16) “Most suicide attempts are caused by conflicts with a close person.” Lastly, participants with suicidal thoughts scored significantly higher than those without suicidal thoughts (a < b) in two items: (11) “Relatives usually have no idea about what is going on when a person is thinking about suicide” and (15) “People who talk about suicide do not necessarily commit suicide.” As shown in Model 1, the highest increase in the experience of suicidal thoughts ever in life was associated with Item 5 – “Almost everyone has at one time or another thought about suicide” (OR = 2.03, 95% CI = 1.74–2.36) – while greater experience of suicide attempts ever in life was associated with Item 14 – “I can understand that peoCrisis (2019), 40(1), 27–35

Crisis (2019), 40(1), 27â&#x20AC;&#x201C;35

Suicide can sometimes be a relief for those involved.

I would consider the possibility of taking my life if I were to suffer from a severe, incurable disease.

Once a person has suicidal thoughts, he/she will never let them go.

On the whole, I do not understand how people can take their lives.c

Relatives usually have no idea about what is going on when a person is thinking about suicide.

A person suffering from a severe, incurable disease expressing wishes to die should get help to do so.

Anybody can commit suicide.

I can understand that people suffering from a severe, incurable disease commit suicide.

People who talk about suicide do not necessarily commit suicide.

Most suicide attempts are caused by conflicts with a close person.

I would like to get help to commit suicide if I were to suffer from a severe, incurable disease.

(1.13-1.46)

(1.01-1.30)

(1.22-1.73)

(1.24-1.66)

2.21

1.91

0.86

3.30

1.77

2.37

1.14

1.65

2.29

1.66

1.51

1.64

(1.50-3.25)

(1.35-2.69)

(0.58-1.27)

(1.95-5.58)

(1.22-2.58)

(1.67-3.37)

(0.75-1.71)

(1.20-2.27)

(1.54-3.39)

(1.18-2.33)

(1.09-2.09)

(1.20-2.25)

1.29

1.14

1.45

1.43

1.61

1.10

1.33

1.41

0.97

1.67

1.20

1.12

2.03

(1.13-1.46)

(1.00-1.30)

(1.22-1.73)

(1.24-1.66)

(1.41-1.84)

(0.98-1.23)

(1.14-1.56)

(1.25-1.59)

(0.85-1.11)

(1.47-1.90)

(1.06-1.36)

(1.00-1.26)

(1.74-2.37)

2.25

1.94

0.84

3.37

1.77

2.42

1.13

1.64

2.30

1.67

1.54

1.65

2.09

(1.52-3.31)

(1.37-2.75)

(0.56-1.26)

(1.98-5.73)

(1.21-2.59)

(1.70-3.45)

(0.74-1.71)

(1.18-2.26)

(1.55-3.42)

(1.18-2.35)

(1.11-2.14)

(1.20-2.25)

(1.37-3.18)

(1.42-3.07)

(1.27-2.70)

Note. ATTS = Attitude Towards Suicide questionnaire. Reference = No suicidal thoughts (a). a Adjusted for gender and age. bAdjusted for gender and age, geographical location, education background, civil state, and income. cReverse scoring question.

1.28

1.14

1.45

1.44

(1.40-1.82)

(0.98-1.22)

1.09

1.60

(1.14-1.57)

(1.24-1.58)

1.34

1.40

(0.86-1.11)

(1.47-1.90)

1.67

0.98

(1.06-1.36)

(1.02-1.26)

1.20

1.12

(1.39-3.20)

(1.14-1.43)

1.85

1.28

1.13

1.46

1.41

1.60

1.07

1.32

1.42

0.98

1.67

1.19

1.11

2.03

1.28

1.20

1.07

(1.13-1.46)

(1.00-1.29)

(1.22-1.74)

(1.22-1.64)

(1.4-1.83)

(0.96-1.20)

(1.13-1.55)

(1.25-1.60)

(0.86-1.12)

(1.46-1.89)

(1.05-1.35)

(0.99-1.25)

(1.74-2.37)

(1.13-1.44)

(1.06-1.36)

(0.94-1.22)

2.24

1.90

0.86

3.45

1.77

2.43

1.11

1.66

2.26

1.67

1.51

1.63

2.10

2.06

1.83

1.82

(1.52-3.31)

(1.34-2.69)

(0.57-1.28)

(2-5.95)

(1.21-2.59)

(1.69-3.49)

(0.73-1.69)

(1.2-2.29)

(1.53-3.36)

(1.18-2.36)

(1.09-2.10)

(1.19-2.24)

(1.37-3.21)

(1.4-3.04)

(1.26-2.67)

(1.28-2.60)

(1.17-2.15)

I could say that I would take my life without actually meaning it.

2.11

1.28

(1.07-1.38)

(1.29-2.62)

1.58

(1.74-2.36)

2.03

(1.42-3.06)

1.22

1.84

(1.19-1.50)

Almost everyone has at one time or another thought about suicide.

2.08

(1.25-2.61)

(0.95-1.23)

1.33

(1.14-1.43)

1.28

1.81

1.08

(1.17-2.16)

Loneliness could for me be a reason to take my life.

(1.07-1.38)

(1.28-2.57)

1.59

1.22

1.81

(1.19-1.50)

When a person commits suicide, it is something that he/she has considered for a long time.

(0.95-1.23)

1.34

1.08

(1.16-2.12)

Many suicide attempts are made because of revenge or to punish someone else.

1.57

(1.19-1.50)

1.33

Suicide is an acceptable means to terminate an incurable disease.

Model 3b

Suicidal thoughts (b) Suicide attempts (c) Suicidal thoughts (b) Suicide attempts (c) Suicidal thoughts (b) Suicide attempts (c)

Model 2a

OR (95% CI)

N0. ATTS

Model 1

Table 3. Result of multinominal logistic regression analysis on attitude toward suicide

32 S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

ÂŠ 2018 Hogrefe Publishing

S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

ple suffering from a severe, incurable, disease commit suicide” (OR = 3.30, 95% CI = 1.95–5.58). Specifically, a greater than twofold excess in the experience of suicide attempts ever in life was associated with a number of items, that is, Items 4, 5, 9, 12, 14, and 17. These associations remained significant after adjustment for several variables as shown in Model 2 and Model 3. The logistic regression analysis confirmed different odds ratios between individuals with suicidal thoughts and suicide attempts in terms of ATTS scores (Figure 1). The statistically significant odds ratios were found in 10 items including Items 2, 3, 4, 6, 9, 12, 14, 15, 16, and 17 after adjustment for age, gender, geographical location, education level, marital status, and income, and especially large differences were shown in Items 9, 12, and 14 (OR > 2.0).

Discussion The purpose of this study was to investigate Koreans’ attitudes toward suicide by utilizing a representative sample in Korea, and to assess the association with suicidal thoughts and suicide attempts. To achieve this objective, the study analyzed data from the 2013 National Suicide Survey involving 1,500 participants aged between 19 and 75 years. Overall, the present study supports previous findings of a significant association between attitudes toward suicide and the experience of suicidal thoughts or suicide attempts (Arnautovska & Grad, 2010; McAuliffe et al., 2003; Stein et al., 1998; Zemaitiene, & Zaborskis, 2005). Individuals with permissive attitudes toward suicide were more likely to have experienced suicidal ideation and suicide attempts. Unlike most previous studies that mainly focused on investigating the association between attitudes toward suicide and suicide risk, the current study attempted to contribute new insights into the field of suicide research by comparing the attitudes toward suicide among individuals with no history of suicidal thoughts, those with a history of suicidal thoughts, and those with a history of suicide attempts. Significant differences in attitudes toward suicide among the three groups were found in 17 out of 37 items in the ATTS. Furthermore, the results of multinomial logistic regression analysis indicated that people with suicidal thoughts and suicide attempts were more likely to have permissive attitudes toward suicide than people with no suicidal thoughts. More accepting attitudes toward suicides were also found in persons reporting on suicide attempts. Suicide attempters further showed more than a twofold increased risk of having the experience of a suicide attempt in three items of the ATTS compared with people with suicidal thoughts. Therefore, this suggests that people © 2018 Hogrefe Publishing

with a higher risk of suicide are more likely to have permissive attitudes toward suicide. The relationship between attitudes toward suicide and experiences of suicidal thoughts and suicide attempts should be taken into account when implementing educational programs and public awareness campaigns to enhance people’s perception of suicide, so as to eventually prevent suicide (Kim & Jung, 2011). Aldrich and Cerel (2009) criticized the poor application of research results to suicide prevention programs including educational resources, advertising campaigns, and media content. Kim and Yoon (2011) suggested that this is potentially due to most existing studies focusing on factors that are difficult to be included in suicide prevention programs or campaigns. The factors involved in previous studies were usually sociodemographic factors, socioeconomic variables, and history of medical treatment. Therefore, it is difficult to create effective campaign messages based on these factors (Kim & Yoon, 2011). The present study thus investigated each item of the ATTS individually. The results showed that although people with a history of suicidal thoughts and suicide attempts were more likely to have permissive attitudes toward suicide than were individuals with no suicidal thoughts, each group showed permissive attitudes toward different items of the ATTS. The suicidal thoughts group was more likely to have permissive attitudes toward the statement, “People who talk about suicide do not necessarily commit suicide,” while the suicide attempter group was more likely to have permissive attitudes toward the statement, “I can understand that people suffering from a severe, incurable, disease commit suicide.” This finding suggests that people who have experienced suicidal thoughts and suicide attempts should be approached through different strategies by considering their different attitudes toward suicide. Furthermore, regarding the method of initiating public awareness education and campaigns for suicide, Kim and Jung (2011) stated that resistant messages against permissive attitudes toward suicide will be more effective in decreasing the risk of suicide. However, most suicide prevention institutions in South Korea are currently spreading messages such as “Life is precious” and “You can keep precious life” for suicide prevention campaigns and advertisements (Song et al., 2014). For this reason, suicide prevention campaigns and education in South Korea should develop and implement resistant messages against permissive attitudes toward suicide in the future. Thus, the items that were highly associated with suicide attempters in this study need to be considered first when making an effective resistant message for public campaigns and education content for suicide prevention; for instance: “Once a person has suicidal thoughts, he/she will never let them Crisis (2019), 40(1), 27–35

go”; “A person suffering from a severe, incurable, disease expressing wishes to die should get help to do so”; “I can understand that people suffering from a severe, incurable, disease commit suicide.” In the future, the changes in attitudes toward suicide should be investigated after implementing education/campaigns using the results of the present study.

Limitations We must cautiously interpret the results of this study within the context of the following limitations. Because all material used in this analysis is from cross-sectional data, it is difficult to eliminate the possibility of changes in attitude toward suicide after having suicidal thoughts or making a suicide attempt. Therefore, the results of the current study cannot be used to conclude that there was a causal relationship between attitudes and suicidal behaviors. There is furthermore the possibility that individuals with a history of suicidal thoughts or attempts may gain a new attitude to justify their suicidal behaviors after such behaviors. Therefore, it is necessary to conduct further research investigating the causal relationship between attitudes and suicidal behaviors using longitudinal data.

Conclusion Despite these limitations, this study suggests there are different attitudes toward suicide among people with no suicidal thoughts, those with suicidal thoughts, and those with suicide attempts. This finding implies the importance of approaching people using different strategies based on their experience of suicidal thoughts and attempts. Moreover, some of the ATTS items that were found to be related to suicidal thoughts and suicidal attempts should be applied to future suicide prevention campaigns and education programs. Acknowledgments This research was supported by the Ministry of Health and Welfare.

References Aldrich, R. S., & Cerel, J. (2009). The development of effective message content for suicide intervention. Crisis, 30(4), 174–179. https://doi.org/10.1027/0227-5910.30.4.174 Crisis (2019), 40(1), 27–35

S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

Arnautovska, U., & Grad, O. T. (2010). Attitudes toward suicide in the adolescent population. Crisis, 31, 22–29. https://doi. org/10.1027/0227-5910/a000009 Beck, A. T., & Robert, A. S. (1989). Clinical predictors of eventual suicide: A 5-to 10-year prospective study of suicide attempters. Journal of Affective Disorders, 17, 203–209. https://doi. org/10.1016/0165-0327(89)90001-3 Chun, S. K., Park, B. G., & Lee, H. B. (2007). A study of risk factors related to mental illness persons’ suicidal ideation. Journal of Mental Health and Social Work, 26, 348–376. Chung, I. J., Park, J. Y., & Kim, E. Y. (2010). Factors predicting suicidal ideation and suicidal attempts of school youth and out-ofschool youth. Journal of Mental Health and Social Work, 34(4), 222–251. Galfalvy, H. C., Oquendo, M. A., & Mann, J. J. (2008). Evaluation of clinical prognostic models for suicide attempts after a major depressive episode. Acta Psychiatrica Scandinavica, 117, 244–252. https://doi.org/10.1111/j.1600-0447.2008.01162.x Ghio, L., Zanelli, E., Gotelli, S., Rossi, P., Natta, W., & Gabrielli, F. (2011). Involving patients who attempt suicide in suicide prevention: A focus group study. Journal of Psychiatric and Mental Health Nursing, 18, 510–518. https://doi.org/10.1111/j.13652850.2011.01697.x Gil, S. (2005). Suicide attempters vs. ideators: Are there differences in personality profiles? Archives of Suicide Research, 9, 153– 161. https://doi.org/10.1080/13811110590904007 Goldstein, R. B., Black, D. W., Nasrallah, A., & Winokur, G. (1991). The prediction of suicide. sensitivity, specificity, and predictive value of a multivariate model applied to suicide among 1906 patients with affective disorders. Archives of General Psychiatry, 48, 418–422. Harriss, L., & Hawton, K. (2005). Suicidal intent in deliberate selfharm and the risk of suicide: The predictive power of the suicide intent scale. Journal of Affective Disorders, 86, 225–233. https:// doi.org/10.1016/j.jad.2005.02.009 Harriss, L., Hawton, K., & Zahl, D. (2005). Value of measuring suicidal intent in the assessment of people attending hospital following self-poisoning or self-injury. British Journal of Psychiatry, 186, 60–66. https://doi.org/10.1192/bjp.186.1.60 Jeong, M. S., & Seo, S. Y. (2014). Predictors of attempts at suicide: A focus on the attune adolescent facing a suicidal crisis. Studies on Korean Youth, 25(2), 145–171. Ji, N. J., Hong, Y. P., & Lee, W. Y. (2016). Comprehensive psychometric examination of the attitudes towards suicide (ATTS) in South Korea. International Journal of Mental Health Systems, 10(1), 2. https://doi.org/10.1186/s13033-016-0035-0 Kim, B. R., Jung, K. J., Lee, S. U., Sea, J. H., Kim, E. U., Kim, S. H., … Ahn, Y. M. (2015). The Korea National Suicide Survey (KNSS): Rationale and design. Korean Journal of Biological Psychiatry, 22, 1–6. Kim, J. H., & Jung, Y. I. (2011). Risk factors associated with suicidal attempts in Korea: Exploring the links with the views on suicide, death, and life. Korean Journal of Health Education and Promotion, 28(2), 109–123. Kim, J. H., & Yoon, Y. M. (2011). Effects of suicidal attempt on subsequent suicidal risk: Testing mediation model of attitudes toward suicide and searching for inoculation message strategies. Journal of Health and Social Science, 30, 49–75. Kim, J. S., & Lee, S. J. (2014). Factors affecting suicide attempts of the elderly with suicide ideation: Secondary data analysis of KNHANES. Journal of the Korea Gerontological Society, 34(1), 73–85. Kim, S. W., Stewart, R., Kim, J. M., Shin, I. S., Yoon, J. S., Jung, S. W., … Jun, T. Y. (2011). Relationship between a history of suicide attempt and treatment outcomes in patients with depression. Journal of Clinical Psychopharmacology, 31, 449–456. © 2018 Hogrefe Publishing

S.-U. Lee et al., A Comparison of Attitudes Toward Suicide

Kodaka, M., Poštuvan, V., Inagaki, M., & Yamada, M. (2011). A systematic review of scales that measure attitudes toward suicide. International Journal of Social Psychiatry, 57(4), 338–361. https://doi.org/10.1177/0020764009357399 McAuliffe, C., Corcoran, P., Keeley, H. S., & Perry, I. J. (2003). Risk of suicide ideation associated with problem-solving ability and attitudes toward suicidal behavior in university students. CRISIS-TORONTO, 24(4), 160–167. https://doi.org/10.1027//02275910.24.4.160 Organisation for Economic Co-operation and Development. (2014). OECD health data 2013. Retrieved from http://www.oecd-ilibrary. org/social-issues-migration-health/suicides_20758480- table10 Renberg, E. S. (2011). Attitudes Towards Suicide (ATTS). Umea, Sweden: Umea University, Department of Clinical Sciences Division of Psychiatry. Renberg, E. S., & Jacobsson, L. (2003). Development of a questionnaire on attitudes towards suicide (ATTS) and its application in a Swedish population. Suicide and Life-Threatening Behavior, 33, 52–64. https://doi.org/10.1521/suli.33.1.52.22784 Sadock, H. S. (2009). Suicide. In B. J. Sadock & V. A. Sadock (Eds.), Kaplan and Sadock’s comprehensive textbook of psychiatry (pp. 2717–2731). Philadelphia, PA: Lippincott Williams & Wilkins. Sinclair, J. M., Harriss, L., Baldwin, D. S., & King, E. A. (2005). Suicide in depressive disorders: A retrospective case-control study of 127 suicide. Journal of Affective Disorders, 87, 107–113. https:// doi.org/10.1016/j.jad.2005.03.001 Song, I. H., Kwon, S, W., Kim, J. S., You, J. W., Park, J. H., Kim, L. J., … An, S. M. (2014). Cross-national comparison of public awareness campaigns for suicide prevention: Analysis of campaign strategies and contents of New Zealand, USA, Ireland, Scotland, Australia, and Korea. Journal of The Korea Contents Association, 14(7), 253–270. Stefansson, J., Nordström, P., & Jokinen, J. (2012). Suicide Intent Scale in the prediction of suicide. Journal of Affective Disorders, 136, 167–171. https://doi.org/10.1016/j.jad.2010.11.016 Stein, D., Brom, D., Elizur, A., & Witztum, E. (1998). The association between attitudes toward suicide and suicidal ideation in adolescents. Acta Psychiatrica Scandinavica, 97(3), 195–201. https://doi.org/10.1111/j.1600-0447.1998.tb09987.x Suominen, K., Isometsa, E., Suokas, J., Haukka, J., Achte, K., & Lönnqvist, J. (2004). Completed suicide after a suicide attempt: A 37-year follow-up study. American Journal of Psychiatry, 161, 562–563. https://doi.org/10.1176/appi.ajp.162.3.633

World Health Organization. (2013). Mental health action plan 2013−2020. Retrieved from http://www.who.int/mental_health/ publications/action_plan/en/ World Health Organization. (2014). Questions and answers on suicide. Retrieved from http://www.who.int/features/qa/24/en/ Zemaitiene, N., & Zaborskis, A. (2005). Suicidal tendencies and attitude towards freedom to choose suicide among Lithuanian schoolchildren: Results from three cross-sectional studies in 1994, 1998, and 2002. BMC Public Health, 5(1), 83. https://doi. org/10.1186/1471-2458-5-83

Received February 4, 2017 Revision received December 19, 2017 Accepted December 22, 2017 Published online June 21, 2018 Jong-Ik Park, MD, PhD, is professor in the Department of Psychiatry, Kangwon National University School of Medicine. He is general director of Chuncheon National Hospital and advisor to the Korea Suicide Prevention Center. Dr. Park received his bachelor’s degree from Seoul National University School of Medicine and his master’s degree and PhD from the Department of Psychiatry, University of Ulsan College of Medicine. Sang-Uk Lee, PhD, is a researcher in the Department of Mental Health Research, National Center for Mental Health, Seoul, South Korea. Dr. Lee received his bachelor’s degree from the Department of Sport Medicine, Kyung-Hee University, and his master’s degree and PhD from the Department of Preventive Medicine, School of Medicine, Kyung Hee University, South Korea. Mina Jeon, MS, is a PhD student in the Department of Psychology and Human Development at University College London, UK. She received her master’s degree from the Department of Psychology and Language Sciences, University College London, UK. Jong-Ik Park Department of Psychiatry Kangwon National University School of Medicine 156 Baengnyeong-ro Chuncheon South Korea lugar@kangwon.ac.kr

Crisis (2019), 40(1), 27–35

Research Trends

Suicide on Instagram – Content Analysis of a German Suicide-Related Hashtag Florian Arendt Department of Communication Science and Media Research, University of Munich (LMU), Germany Abstract. Background: Suicide is the second leading cause of death among 15–29-year-olds globally. Unfortunately, the suicide-related content on Instagram, a popular social media platform for youth, has not received the scholarly attention it deserves. Method: The present study provides a content analysis of posts tagged as #selbstmord, a German suicide-related hashtag. These posts were created between July 5 and July 11, 2017. Results: Approximately half of all posts included words or visuals related to suicide. Cutting was by far the most prominent method. Although sadness was the dominant emotion, self-hate and loneliness also appeared regularly. Importantly, inconsistency – a gap between one’s inner mental state (e.g., sadness) and one’s overtly expressed behavior (e.g., smiling) – was also a recurring theme. Conversely, help-seeking, death wishes, and professional awareness–intervention material were very rare. An explorative analysis revealed that some videos relied on very fast cutting techniques. We provide tentative evidence that users may be exposed to purposefully inserted suicide-related subliminal messages (i.e., exposure to content without the user’s conscious awareness). Limitations: We only investigated the content of posts on one German hashtag, and the sample size was rather small. Conclusion: Suicide prevention organizations may consider posting more awareness– intervention materials. Future research should investigate suicide-related subliminal messages in social media video posts. Although tentative, this finding should raise a warning flag for suicide prevention scholars. Keywords: social media, youth, Instagram, cutting, subliminal messages

Suicide is the second leading cause of death among 15–29-year-olds globally (World Health Organization [WHO], 2017a). There is a web of multiple risk factors including emotional states (e.g., depressed mood), personality traits (e.g., low self-esteem), deficits in problem-solving abilities, and stressors such as specific life events (e.g., a breakdown of a close relationship; Orbach & Iohan-Barak, 2009). Among the myriad factors, the media are considered a key factor (Mann et al., 2005; Niederkrotenthaler & Stack, 2017; WHO, 2017b). Research has found that the media can have detrimental and beneficial effects. On the one hand, there is evidence that news reporting on suicide and depictions of suicide in entertainment programming can increase suicidal behavior – a phenomenon known as the Werther effect (Phillips, 1974; Stack, 2005). On the other hand, research has also found that the media can decrease suicidal behavior by responsible reporting on suicide, for example, by publishing stories on individuals who have overcome a suicidal crisis – a phenomenon known as the Papageno effect (Nieder krotenthaler et al., 2010). Therefore, media effects on suicidal behavior depend on the way suicide is depicted (see Niederkrotenthaler & Stack, 2017). Much of our knowledge about media effects on suicidal behavior stems from studies conducted in the context Crisis (2019), 40(1), 36–41 https://doi.org/10.1027/0227-5910/a000529

of traditional media environments, largely dominated by printed newspapers and television. Importantly, the media landscape has changed dramatically during recent years, with social media becoming a key player. Instagram – a prime example of a popular social media platform among youth (see next section) – has not received the scholarly attention it deserves. This social media platform distributes a plethora of suicide-related content that is publicly available every day. The present study aims to improve our understanding by providing a content analysis of a German suicide-related hashtag on Instagram.

Social Media and Suicidal Behavior Although the Internet has increasingly gained the attention of suicide researchers (Biddle, Donovan, Hawton, Kapur, & Gunnell, 2008), our knowledge of the role social media play in terms of imitational and preventive effects is limited. Nevertheless, research has shown that social media are linked to both benefits and risks in regard to the health of adolescents (Reid-Chassiakos, Radesky, Christakis, Moreno, & Cross, 2016). Whereas the benefits may include a reduction in social isolation, encouragement toward recovery, and a decrease in self-harm urges, risks © 2018 Hogrefe Publishing

F. Arendt, Suicide on Instagram

may include a disturbing effect on (especially young) users and a contagious triggering effect of self-harming behaviors (Baker & Lewis, 2013; Brown et al., 2018). Instagram is a prime example of a modern social media platform that is especially popular among youth (Pew Research Center, 2015). Instagram is a web-based photoand video-sharing application that allows users to watch and share pictures and short videos. Pictures or videos on a specific topic can be identified by a so-called hashtag – a word or phrase preceded by the symbol #. A hashtag allows users to easily access posts with a specific theme or content and it enables users to access and feel connected to a theme-related community (Moreno, Ton, Selkie, & Evans, 2016). Importantly for the present study, there are a number of suicide-related hashtags in different languages (e.g., #suicide, #suicidio), including the popular German hashtag that is the central focus of the present study (#selbstmord; English translation of the German word selbstmord = suicide). Unfortunately, the suicide-related content on Instagram has not received the scholarly attention it deserves. Scholars have only recently started investigating self-harm on Instagram. Importantly, previous studies focused on nonsuicidal self-injury (NSSI) – the deliberate destruction of body tissue in the absence of a conscious suicide intention (Nock & Favazza, 2009). For example, Brown and colleagues (2018) investigated posts on #cutting (and related hashtags) and found that pictures of NSSI are frequently posted. This finding is supported by a recent study (Miguel et al., 2017). Notably, recovery-oriented resources are virtually absent, as Miguel and colleagues (2017) have shown. Although Instagram has introduced its own content advisory warning message – a message that pops up explaining that you will possibly view emotionally disturbing graphic content and the message also provides helpful resources; individuals can nevertheless proceed viewing the content – a previous study has shown that only one third of relevant suicide-related hashtags trigger this message (Moreno et al., 2016). Key for the present study is that there is hardly any knowledge related to self-injury with a conscious suicide intention. Thus, the present study provides a content analysis of postings to #selbstmord. Due to this lack of previous studies, the present work purposefully employed a descriptive and explorative approach. We derived a set of variables from the suicide prevention literature (see Shneidman, 1987; Sonneck, Kapusta, Tomandl, & Voracek, 2012) and coded publicly accessible posts on #selbstmord: We placed special emphasis on the suicide method, bearing possible method-specific imitation effects in mind (Etzersdorfer & Sonneck, 1998; Schmidke & Häfner, 1988). We also coded for the presence of emotions (e.g., sadness, self-hate), awareness–intervention materials including © 2018 Hogrefe Publishing

recovery-oriented resources, overt expressions of death wishes, and help-seeking. The use of social media platforms such as Instagram has increased in recent years (Pew Research Center, 2015). Thus, it is important that scholars, mental health professionals, social media platform operators, parents, and teachers are aware of problematic aspects of young people’s social media use (Miguel et al., 2017). Suicide prevention research has a special responsibility in terms of monitoring for potential problems.

Method We conducted a content analysis of posts tagged with the hashtag #selbstmord. We downloaded the 250 most recent posts on July 11, 2017, and coded for the presence or absence of specific content elements. These posts were posted on Instagram and tagged with the hashtag #selbstmord between July 5 and July 11 (i.e., within 1 week). All variables were coded dichotomously (0 = absent, 1 = present) by one coder. For the reliability analysis, we randomly selected 10% of the material. The material was coded by the same coder a second time. The analysis generally indicated reliable measurements. Krippendorff ’s α values were generally acceptable (all values, α > .74). There were two exceptions: The measurement of the variables woman (α = .58) and man (α = .35) were not reliable. Although Krippendorff ’s α is very sensitive to inconsistent coding in small samples where one code dominates (i.e., 0 or “absent” in the present study), these low α values indicate that the sex-related variables could not be measured in a reliable way. This was unexpected because sex is typically one of the most reliable variables to code. However, in the present sample, the images and videos often represented young individuals. Sex-related cues were ambiguous and thus hard to evaluate. Furthermore, as we will outline, many videos used very fast cutting techniques. Because of these techniques, sex-related cues were difficult to code as well. Despite this low level of reliability, we decided to report on our planned analyses regarding sex. However, we emphasize that the sex-related results should be interpreted with caution. We can only provide tentative evidence on this aspect.

Variables Reference to Suicide This variable was coded as present when the post included an explicit reference to suicidal thoughts, plans, or behaviors. Crisis (2019), 40(1), 36–41

F. Arendt, Suicide on Instagram

Suicide Method We coded whether the post depicted a specific suicide method: cutting, jumping, firearm, hanging, drowning, poisoning, and railway. Each method was represented by its own variable. Demographics We coded whether the post visually (e.g., pictures, visuals) or verbally (e.g., words, phrases) represented a female or a male. We did not code for sex when the post represented drawings or animated characters. Furthermore, the age of the individual was coded. We coded whether the post included young individuals (i.e., youth: children, adolescents, young adults), older adults (including the elderly), or held no age-related information. Emotions The following emotional expressions were coded when they were visually or verbally present: sadness (e.g., tears, unambiguously sad faces), loneliness (e.g., the presence of the word lonely), and self-hate (e.g., verbal statements such as “I hate myself ”). Inconsistency We coded whether the post explicitly addressed a conflict experienced when internal cognitive and emotional content was inconsistent with overtly expressed behavior. Inconsistency was coded as present, for example, when the post presented a verbal statement such as “I’m fine” alongside a (visually or verbally represented) depressed mood. Death Wish

DE ON INSTAGRAM Furthermore, we coded whether explicit death wishes

n a H

w D

w il a R

y a

g in

e ir F

m u J

in p

in tt u C

P o s t s In c lu d in g M e t h o d [ % ]

(e.g., “I want to die”) had been verbally communicated.

M e th o d

Figure 1. Percentages of posts showing a specific suicide method. Error bars indicate bootstrapped 95% confidence intervals. Reading example: 22% of all posts (N = 250) depicted cutting.

Help-Seeking We coded this variable when statements such as “Help me,” “I need help,” or “Who can help me?” were present. Professional Awareness Material We coded whether the post included professional awareness material (e.g., a reference to a counseling service).

Data Analysis We calculated descriptive statistics (percentage values) and 95% confidence intervals based on bootstrapping techniques. The percentage values are significantly different when the confidence intervals do not overlap. The confidence interval lower and upper limits are reported after the percentage value.

Results There was an explicit reference to suicide in 45.6% (95% CI = 39.6–51.6) of all posts. At least one specific method was depicted in 34.4% (95% CI = 28.8–40.8) of all posts. As can be seen in Figure 1, cutting was by far the most frequently depicted method, followed by jumping, firearm, poisoning, railway, drowning, and hanging. Although both women and men (or boys and girls) regularly appeared visually in posts, women (and girls) appeared in significantly more posts (37.2%, 95% CI = 31.2–43.2), than men (or boys) did (16.8%, 95% CI = 12.4–21.6). Posts with pictures of individuals typically included youths (37.6%, 19 95% CI = 31.6–43.2). Conversely, older adults appeared in only 0.4% (95% CI = 0.0–1.2) of all posts. In the majority of the posts, however, age could not be coded because of an absence of age-related information (62.0%, 95% CI = 56.4–68.0). Concerning emotional states, we found that sadness was depicted in 26.8% (95% CI = 21.6–32.4) of all posts and thus represented the most frequently depicted emotion. Self-hate (95% CI = 9.2%, 5.6–12.8), and loneliness (95% CI = 5.2%, 2.8–8.4), were represented in significantly fewer posts, but also appeared regularly. Inconsistency was apparent in 12.4% (95% CI = 8.4–16.8) of posts – an unexpectedly high number of appearances. Only a few posts included death wishes (2.8%, 95% CI = 0.8–5.2) or help-seeking (2.4%, 95% CI = 0.8–4.4). Notably, there was only one post that included professional awareness material (0.4%, 95% CI = 0.0–1.2). Interestingly, this content was not posted by a suicide prevention organization, but included a screenshot made by a user showing Instagram’s advisory content warning message.

Figure 1. Percentages of posts showing a specific suicide method. Error bars indicate Crisis (2019), 40(1), 36–41

rapped 95% confidence intervals. Reading example: 22% of all posts (N = 250) depicted

F. Arendt, Suicide on Instagram

Explorative Analysis

rials including recovery-oriented resources should be posted by professional suicide prevention organizations more frequently. As noted by Brown and colleagues (2018), preventive work might benefit from advancements in machine learning that can help to identify vulnerable users and posts. When users are identified, access to counseling services should be provided.

Videos were included in 26.4% (95% CI = 21.2–32.0) of all posts. During the coding process, an in-depth analysis of the video content revealed an unexpected finding: Many videos relied on very fast cutting techniques (i.e., a film editing technique that relies on the use of several consecutive shots of very brief duration). In these videos, it was sometimes almost impossible to get the depicted content by viewing the videos at a regular speed. When analyzed in a frame-by-frame fashion, some videos presented suicide-related content (i.e., an explicit depiction of suicidal behavior) in these extremely fast segments. The suicide-related content was represented visually (e.g., a short clip of a girl slicing her throat) and by words. Regarding the latter, the words (e.g., dead, suicidal, depression) were presented as short flashes. To gain a general impression of some of these possible suicide-related subliminal messages, sample videos can be obtained from the author upon request. We return to this important explorative finding in the Discussion section.

Discussion Social media platforms such as Instagram have only recently begun to gain scholarly attention in the suicide prevention domain. As already noted, previous research on suicide-related content on Instagram has focused on NSSI. We extended this body of research by investigating posts with a German suicide hashtag. We found that cutting was the most prominent method. Importantly, many posts presented an explicit graphic depiction of the method. This is a significant finding when bearing possible method-specific imitation effects in mind (Etzersdorfer & Sonneck, 1998; Schmidke & Häfner, 1988). Sadness was the dominant emotion. Explicit death wishes and help-seeking were rare. Interestingly, a gap between the Instagrammer’s inner mental state (e.g., sadness, self-hate) and his or her overtly expressed behavior (e.g., smiling) – a phenomenon termed inconsistency in the present study – was a recurring theme. Instagram is an important social media channel, especially for youth. Thus, there is great potential for suicide prevention to reach young, vulnerable individuals via this platform. It is important to note that Instagram already uses warning messages when individuals want to expose themselves to specific hashtags. Unfortunately, a recent study found that only one third of problematic suicide-related hashtags generated Instagram’s own content advisory message (Moreno et al., 2016). Warning messages are important, but not enough. Awareness–intervention mate© 2018 Hogrefe Publishing

Subliminal Messages Explorative analyses revealed that many videos relied on very fast cutting techniques and presented suicide-related content within these fast-paced scenes. Most notably, some of this content is presented at subthreshold durations. Thus, it is very hard, sometimes impossible, to be consciously aware of this suicide-related content when these videos are viewed at a regular speed. The advertising literature has coined the term subliminal messages for similar advertising content. Subliminal messages describe “a technique of exposing consumers to product pictures, brand names, or other marketing stimuli without the consumers having conscious awareness” (Trappey, 1996, p. 517). Importantly, subliminal messages in advertising are forbidden in many countries owing to their deceptive and manipulative character. Although we cannot categorically prove whether the producers of these suicide-related videos had the persuasive intention of detrimentally changing other users’ behavior, it is nevertheless highly questionable from an ethical point of view. Although the effects of subliminally presented stimuli are complex, we know from previous psychological research that subliminally presented stimuli can elicit substantial effects on human information processing (Olsen & Fazio, 2002). A short exposure to these stimuli can even lead to automatic behavior changes (Bargh, Chen, & Burrows, 1996). The general idea is that subliminally presented stimuli can prime (i.e., activate) corresponding concepts in memory (even outside of conscious awareness), which in turn activate mentally associated behavioral schemata. Subliminal messages may thus increase the accessibility of suicide-related behavioral schemata (e.g., related to a specific method). Although the mere automatic (re-)activation of behavioral schemata in memory does not fully determine behavior, it increases the likelihood of its execution (see Bargh, 2017). The problem becomes even more apparent when considering the following fact: As revealed by Moreno and colleagues (2016), some suicide-related hashtag terms overlap with existing non-suicide terms such as #cat. Moreno and colleagues (2016) argued that children or adolescents who want to see pictures or videos of cute cats may be inadvertently exposed to posts with suicide-related content Crisis (2019), 40(1), 36–41

F. Arendt, Suicide on Instagram

because of the overlap. In the context of subliminal messages, they may be exposed to suicide-related content even without consciously being aware of it – an alarming phenomenon. We want to emphasize that the empirical evidence and our post hoc theorizing on this phenomenon are tentative. Despite its tentative nature, it should definitely raise a warning flag for both suicide researchers and mental health professionals. More research is undoubtedly needed on this possible alarming subliminal phenomenon and its effects.

Limitations This study has several limitations. First, we only investigated the content of posts on one German hashtag. This decreases the generalizability of our findings. However, many posts use English content. This is especially true for the video posts. Second, the study used a descriptive and explorative approach without a focus on a deductive test of hypotheses. Nevertheless, this approach enabled us to reveal an important phenomenon (subliminal suicide messages). It is up to future research to provide a more systematic assessment. Third, the sample of posts was rather small (compared with the number of posts that are available on that hashtag). Furthermore, these posts were posted on Instagram within a period of only 1 week. This decreases the generalizability of the findings. Fourth, the coding of sex-related content (i.e., the variables woman and man) was not reliable (see explanation in Method section). Future studies should provide replications of sex-related findings. Fifth, we did not investigate variables related to the source of the posts (i.e., who posted the content). Sixth, we investigated only three emotions (i.e., sadness, loneliness, and self-hate). Future research should definitely broaden the perspective.

Conclusion In the context of traditional media such as newspapers and television, media guidelines have been developed to increase responsible reporting (Bohanna & Wang, 2012; Pirkis, Blood, Beautrais, Burgess, & Skehan, 2006). Similar guidelines should be developed for social media platforms as well. If future studies can support our explorative finding, these guidelines should include a passage on subliminal messages. There is clearly a need to monitor suicide-related content on social media platforms. Suicide prevention scholars as well as mental health proCrisis (2019), 40(1), 36–41

fessionals should be aware of the possible consequences of social media platforms such as Instagram. Owing to its high number of younger users, this social media platform clearly deserves more attention. In a first step, research has to provide more empirical evidence: What kind of (subliminal) messages do users post? Who posts them? What is the content of the posts? How often are they posted? In a second step, possible policy interventions should be discussed by suicide experts. The goal is to increase awareness among health professionals, journalists, parents, teachers, and social media companies. Subliminal messages in advertising are forbidden in many countries because of their deceptive and manipulative character. Discussions are needed on whether a ban is also an appropriate solution in the social media domain. Furthermore, professional suicide prevention organizations may consider posting more awareness–intervention materials including help sources on Instagram, a prime example of modern social media platforms that is especially popular among youth. This allows professional suicide prevention organizations to provide preventive materials especially to this (young) segment of the population.

References Baker, T., & Lewis, S. (2013). Responses to online photographs of non-suicidal self-injury: A thematic analysis. Archives of Suicide Research, 17, 223–235. https://doi.org/10.1080/13811118.201 3.805642 Bargh, J. (2017). Before you know it: The unconscious reasons we do what we do. London, UK: William Heinemann. Bargh, J., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71, 230–244. https://doi.org/10.1037/0022-3514.71.2.230 Biddle, L., Donovan, J., Hawton, K., Kapur, N., & Gunnell, D. (2008). Suicide and the internet. BMJ, 336, 800–802. https://doi.org/ 10.1136/bmj.39525.442674.AD Bohanna, I., & Wang, X. (2012). Media guidelines for the responsible reporting of suicide. Crisis, 33, 190–198. https://doi. org/10.1027/0227-5910/a000137 Brown, R., Fischer, T., Goldwich, A., Keller, F., Young, R., & Plener, P. (2018). #cutting: Non-suicidal self-injury (NSSI) on Instagram. Psychological Medicine, 48, 337–346. https://doi.org/10.1017/ S0033291717001751 Etzersdorfer, E., & Sonneck, G. (1998). Preventing suicide by influencing mass-media reporting. The Viennese experience 1980–1996. Archives of Suicide Research, 4, 67–74. https://doi. org/10.1023/A:1009691903261 Mann, J., Apter, A., Bertolote, J., Beautrais, A., Currier, D., Haas, A., … Hendin, H. (2005). Suicide prevention strategies: A systematic review. JAMA, 294, 2064–2074. https://doi.org/10.1001/ jama.294.16.2064 Miguel, E., Chou, T., Golik, A., Cornacchio, D., Sanchez, A., DeSerisy, M., & Comer, J. (2017). Examining the scope and patterns of deliberate self-injurious cutting content in popular social media. Depression and Anxiety, 34, 786–793. https://doi.org/10.1002/ da.22668 © 2018 Hogrefe Publishing

F. Arendt, Suicide on Instagram

Moreno, M., Ton, A., Selkie, A., & Evans, Y. (2016). Secret society 123: Understanding the language of self-harm on Instagram. Journal of Adolescent Health, 58, 78–84. https://doi.org/10.1016/j. jadohealth.2015.09.015 Niederkrotenthaler, T., & Stack, S. (2017). Media and suicide. International perspectives on research, theory, and policy. New York, NY: Transaction Publishers. Niederkrotenthaler, T., Voracek, M., Herberth, A., Till, B., Strauss, M., Etzersdorfer, E., … Sonneck, G. (2010). The role of media reports in completed and prevented suicide: Werther versus Papageno effects. British Journal of Psychiatry, 197, 234–243. https://doi.org/10.1192/bjp.bp.109.074633 Nock, M., & Favazza, A. (2009). Understanding nonsuicidal self-injury: Origins, assessment and treatment. In M. Nock (Ed.), Nonsuicidal self-injury: Definition and classification (pp. 9–18). Washington, DC: American Psychological Association. Olsen, M. & Fazio, R. (2002). Implicit acquisition and manifestation of classically conditioned attitudes. Social Cognition, 20, 89–102. https://doi.org/10.1521/soco.20.2.89.20992 Orbach, I., & Iohan-Barak, M. (2009). Psychopathology and risk factors for suicide in the young: theoretical and empirical. In D. Wasserman & C. Wasserman (Eds.), Oxford textbook of suicidology and suicide prevention: A global perspective (pp. 633–642). Oxford, UK: Oxford University Press. Pew Research Center. (2015). Teens, social media and technology overview 2015. Retrieved from http://www.pewinternet.org/ 2015/04/09/teens-social-media-technology-2015/ Phillips, D. P. (1974). The influence of suggestion on suicide: Substantive and theoretical implications of the Werther effect. American Sociological Review, 39, 340–354. https://doi. org/10.2307/2094294 Pirkis, J. E., Blood, R. W., Beautrais, A., Burgess, P. M., & Skehan, J. (2006). Media guidelines on the reporting of suicide. Crisis, 27, 82–87. https://doi.org/10.1027/0227-5910.27.2.82 Reid-Chassiakos, Y., Radesky, J., Christakis, D., Moreno, M., & Cross, C. (2016). Children and adolescents and digital media. Pediatrics, 138, e30162593. https://doi.org/10.1542/peds.2016-2593 Schmidke, A., & Häfner, H. (1988). The Werther effect after television films: New evidence for an old hypothesis. Psychological Medicine, 18, 665–676. https://doi.org/10.1017/S00 33291700008345

Shneidman, E. (1987). A psychological approach to suicide. In G. Vandenbos & B. Bryant (Eds.), Cataclysms, crises, and catastrophes (pp. 151–183). Washington, DC: American Psychological Association. Sonneck, G., Kapusta, N., Tomandl, G., & Voracek, M. (2012). Kri senintervention und Suizidverhütung [Crisis intervention and suicide prevention]. Vienna, Austria: Facultas. Stack, S. (2005). Suicide in the media: A quantitative review of studies based on nonfictional stories. Suicide & Life-Threatening Behavior, 35, 121–133. https://doi.org/10.1521/suli.35.2. 121.62877 Trappey, C. (1996). A meta-analysis of consumer choice and subliminal advertising. Psychology & Marketing, 13, 517–530. https:// doi.org/10.1002/(SICI)1520-6793(199608)13:5<517::AIDMAR5>3.0.CO;2-C World Health Organization. (2017a). Suicide. Fact sheet. Retrieved from http://www.who.int/mediacentre/factsheets/fs398/en/ World Health Organization. (2017b). Preventing suicide: A resource for media professionals. Update 2017. Retrieved from http:// www.who.int/mental_health/suicide-prevention/resource_ booklet_2017/en/

Received July 27, 2017 Revision received December 21, 2017 Accepted December 22, 2017 Published online June 21, 2018 Florian Arendt, PhD, is a postdoctoral researcher at the Department of Communication Science and Media Research, LMU Munich, Germany. His primary research interests lie in the field of health communication.

Florian Arendt Department of Communication Science and Media Research University of Munich (LMU) Oettingenstraße 67 80538 Munich Germany florian.arendt@ifkw.lmu.de

Crisis (2019), 40(1), 36–41

Research Trends

A 17-Year National Study of Prison Suicides in Belgium Louis Favril1, Ciska Wittouck2, Kurt Audenaert2, and Freya Vander Laenen1 Institute for International Research on Criminal Policy (IRCP), Faculty of Law and Criminology, Ghent University, Belgium Department of Psychiatry and Medical Psychology, Faculty of Medicine and Health Sciences, Ghent University, Belgium

1 2

Abstract. Background: Suicide is a leading cause of mortality in prisoners worldwide, yet empirical data on this matter are lacking in Belgium. Aims: This study sought to describe characteristics associated with a consecutive series of suicides in Belgian prisons from 2000 to 2016 inclusive, in order to inform suicide prevention strategies. Method: All documented cases of suicide (N = 262) were reviewed using a standardized assessment checklist. Official records were abstracted for prisoners’ sociodemographic, criminological, and clinical information, as well as for suicide-related characteristics. Results: Over the 17-year study period, suicides accounted for one third of all deaths in Belgian prisons. The average annual suicide rate in Belgium from 2000 to 2016 was 156.2 per 100,000 prisoners. Examination of all cases highlights both individual (psychiatric disorders and a history of suicide attempt) and situational (the early period of incarceration, interfacility transfers, and placement in solitary confinement) factors common in many prison suicides; some of them amenable to (clinical) management, which presents several potential avenues for suicide prevention. Limitations: Given the absence of a matched control group, no conclusions could be ascertained regarding risk factors. Conclusion: Suicide is a common, preventable cause of death among prisoners in Belgium. The results underscore the timely need for national standards and guidelines for suicide prevention in Belgian prisons. Keywords: suicidal behavior, inmates, imprisonment, suicide prevention

Suicide is a leading cause of mortality in custodial settings worldwide (Konrad et al., 2007; Rabe, 2012). Based on data sampled across 24 high-income countries, suicide rates in prisoners are estimated to be at least three times higher than in the general population, reflecting rates in excess of 100 suicides per 100,000 inmates in many European countries (Fazel, Ramesh, & Hawton, 2017). Numerous factors may contribute to this disproportionally high incidence of suicide in prisons. First, prisoners represent a selection of vulnerable individuals who are already at a greater risk of suicide before imprisonment. High rates of psychosocial adversities, mental illness, substance use, impulsive-aggressive personality traits, and prior suicidal behavior have been consistently identified among prisoners (Enggist, Møller, Galea, & Udesen, 2014; Fazel & Baillargeon, 2011; Fazel, Hayes, Bartellas, Clerici, & Trestman, 2016) – all of which are well-established risk factors for suicide (O’Connor & Nock, 2014; Turecki & Brent, 2016). Second, in addition to these individual-level risk factors, characteristics unique to the prison environment and the correctional regime (e.g., the loss of freedom and autonomy, overcrowding, bullying, poor social support, and lack of purposeful activity) may negatively affect prisoners’ mental health (Goomany & Dickinson, 2015) and have been shown to increase suicide risk during incarceration (Blaauw, Winkel, & Kerkhof, 2001; Huey & McNulty, 2005; Leese, Thomas, & Snow, 2006; Liebling, 2006). Crisis (2019), 40(1), 42–53 https://doi.org/10.1027/0227-5910/a000531

Taken together, as outlined by the stress–diathesis model of suicidal behavior (van Heeringen, 2012), contemporary research emphasizes the exposure of vulnerable individuals to a stressful setting as a sound empirical explanation for prison suicide (Dye, 2010; Favril, Vander Laenen, Vandeviver, & Audenaert, 2017c; Liebling & Ludlow, 2016; Rivlin, Hawton, Marzano, & Fazel, 2013; Stoliker, 2018), in which suicidal behavior is purported to result from the dynamic interaction between both proximal and distal factors, including individual state- and trait-dependent characteristics, and prison-specific environmental influences (Marzano et al., 2016). Reducing the number of prison suicides has been highlighted as an international priority by the World Health Organization (WHO, 2007). To this end, several descriptive (Gauthier, Reisch, & Bartsch, 2015; Hayes, 2012; O’Driscoll, Samuels, & Zacka, 2007; Shaw, Baker, Hunt, Moloney, & Appleby, 2004) and case-control (Blaauw, Kerkhof, & Hayes, 2005; Fruehwald, Matschnig, Koenig, Bauer, & Frottier, 2004; Humber, Webb, Piper, Appleby, & Shaw, 2013) studies over the past years have sought to delineate relevant factors associated with prison suicide, in order to inform prevention strategies at individual, staff, and organizational level. According to a meta-analysis of 34 studies comprising around 5,000 cases of suicide, the most important risk factors associated with suicide in prisoners include single-cell occupation, remand status, © 2018 Hogrefe Publishing

L. Favril et al., Prison Suicides in Belgium

serving a life sentence, recent suicidal ideation, a history of attempted suicide, and having a psychiatric diagnosis or a history of alcohol misuse (Fazel, Cartwright, Norman-Nott, & Hawton, 2008). Almost a decade after publication of this oft-cited meta-analysis, the National Institute for Health and Care Excellence (NICE, 2017) recently recommended that it is (still) crucial to identify and understand the factors related to suicide when developing interventions to prevent such outcomes among prisoners. To date, however, not a single study has been noted for suicide mortality among prisoners in Belgium – a country that has, by international standards, a high suicide rate in the general population (Organisation for Economic Cooperation and Development, 2017). Given the clinical and public health significance of prison suicide, and the absence of an evidence-based suicide prevention policy in Belgian prisons, we aimed to conduct a national review of all suicide cases that occurred in the whole prison estate of Belgium from 2000 to 2016.

Method Setting With an incarceration rate of 105 per 100,000 individuals (Walmsley, 2016), the average daily prison population in Belgium was just over 11,000 in 2015 (Directory-General of Penitentiary Institutions [DG EPI], 2016). From 2000 to 2013, the inmate population at census grew from 8,464 to 11,645 prisoners, and slightly decreased to 10,619 prisoners in 2016. Nonetheless, despite this decline, overcrowding has been the norm in Belgian prisons during the 17-year period covered by this study (operating at 110% capacity in 2015). Males comprise approximately 95% of all inmates in Belgium. While in some countries individuals held in custody on criminal charges (i.e., pre-trial/ remand prisoners) are housed in separate facilities from sentenced prisoners, the majority of Belgian prisons house both types of inmates. Notably, in Belgium, 8.2% of the total prison population in 2015 (N = 904) were offenders who were deemed criminally irresponsible (ODCI). As formulated by De Smet and colleagues (2016), in Belgium, ODCI (also referred to as internees or mentally ill offenders) for their criminal actions because of mental illness or intellectual disability are subject to a specific safety measure with the dual objective of protecting society and providing court-mandated care to the offender. While Belgian law requires that ODCI should be in a hospital, clinic, or other appropriate institution outside of prison, in practice, about one third of all such offenders still reside in prison. © 2018 Hogrefe Publishing

Study Design and Data Sources The present study set out to conduct a retrospective analysis of a consecutive series of suicides that occurred in the whole Belgian prison estate between 2000 and 2016 inclusive. In Belgium, every death occurring during imprisonment is subject to a coronial inquest and an inquiry by the public prosecutor. Only deaths where suicide was the official cause of death by a coroner’s verdict were included in the present study. Suicides among offenders who were administratively registered in the prison database but who were not physically residing in a correctional facility, for example, in the case of electronic monitoring or those who were granted a penitentiary leave, were not included in the study. The DG EPI provided a list of all prisoners who died by suicide from January 1, 2000, to December 31, 2016. For each suicide case, we requested the inmate’s official records (i.e., general prison files, psychosocial evaluation reports, and suicide review reports). During this 17-year period, the total number of deaths that were classified as suicides by coroner’s verdict was 262, of which most personal records were available for scrutiny. Employing a standardized assessment checklist to ensure consistent data collection across the whole study population, all available records were abstracted by the two researchers (L.F. and C.W.) for sociodemographic, criminological, and clinical information, as well as for suicide-related characteristics. Questionable information was scored based on team consensus. Primary sources of data comprised inmates’ general prison files (containing sociodemographic and criminological information) and suicide review reports (a mandatory report prepared by the local prison governor following each suicide describing its details and circumstances). These two sources of information were available for all cases, except for 13 suicide review reports that could not be retrieved. Supplementary sources of information, if available, consisted of psychosocial evaluation reports and suicide notes.

Data Analysis The annual reports of the DG EPI were examined for official statistics of the average daily population (ADP) and the total number of deaths in Belgian prisons for 2000–2016. Combined with the number of suicides identified in this study, unadjusted annual suicide rates were calculated per 100,000 inmates using the ADP figures of the respective year as denominator. Since no matched control group was included in the current study, data analysis was essentially descriptive in nature. If an item of information was missing for a case, the case was removed from the analysis of Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

that item. The denominator in all estimates is therefore the number of valid cases for each item (Shaw et al., 2004).

Ethical Considerations Ethical approval for the study protocol was granted by the Ethics Committee of Ghent University, Faculty of Law and Criminology. The study was registered with the Belgian Commission for the Protection of Privacy (CPP). Permission to carry out the study was obtained from the DG EPI – the department of the Federal Public Service of Justice responsible for all prison affairs in Belgium.

Results Suicide Rates According to the data obtained from DG EPI, there were a total of 794 deaths in the Belgian prison system between 2000 and 2016. Of these deaths, 262 (33.0%) were classified as suicides – reflecting an average of 15.4 suicides per annum. Based on ADP figures, the annual suicide rate fluctuated widely; from 86.9 per 100,000 prisoners in 2004

to 270.7 per 100,000 in 2001 (see Table 1). During the 17-year period covered by this study, the average suicide rate in Belgium was estimated to be 156.2 per 100,000 prisoners per year.

Sociodemographics Table 2 shows the sociodemographic characteristics of all cases. Of the total 262 suicides included in this study, 248 were males (94.7%) and 14 were females (5.3%). At the time of suicide, the mean age of all cases was 37.0 years (SD = 10.9, range = 19–72), and prisoners’ modal age was 25–34 years (36.3%). Only five prisoners (1.9%) were aged 65 or older. The large majority of cases (77.5%) were Belgian nationals. About three prisoners in four were single at the time of death (which includes being divorced, separated, or widowed), whereas 24.8% were married.

Criminological Variables Criminological features of all prisoners included in the study are presented in Table 2. At the time of suicide, 109 inmates were on remand (41.6%), 47 were ODCI (17.9%),

Table 1. Total number of deaths and suicides in Belgian prisons, 2000–2016 Year

ADP

Number of deathsa

2000

8,464

Number of suicides (% of prison deaths) 19

(35.8)

Suicide rate per 100,000 224.5

2001

8,497

(44.2)

270.7

2002

8,769

(48.9)

250.9

2003

9,008

(25.0)

122.1

2004

9,201

(25.8)

86.9

2005

9,238

(38.2)

140.7

2006

9,579

(36.8)

146.2

2007

9,873

(28.6)

141.8

2008

9,891

(31.4)

161.8

2009

10,238

(30.2)

127.0

2010

10,536

(43.5)

189.8

2011

10,974

(24.5)

109.4

2012

11,330

(27.7)

114.7

2013

11,645

(26.8)

128.8

2014

11,578

(33.9)

172.7

2015

11,041

(36.4)

144.9

2016

10,619

(24.5)

122.4

10,028

46.7

15.4

(33.0)

156.2

Note. ADP = average daily population. aTotal number of deaths in prison, including suicides (source: Directory-General of Penitentiary Institutions, 2016).

Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

and 106 were sentenced (40.5%). For the latter group, modal sentence length was more than 3 years up to 5 years (26.4%). Irrespective of custodial status, one third of all cases (32.4%) was charged with, or convicted of, murder or manslaughter, followed by acquisitive crimes with or without violence (29%), and sexual offences (16%). Approximately half of all prisoners died by suicide during their first-ever episode of imprisonment; 47.5% had a prior history of incarceration. Next, the interval from the prisoners’ initial reception into prison to the time of suicide was examined for all cases (Table 3). Results show that 20 inmates (7.6%) died by suicide within the first 24 hr of confinement. One fifth of all suicides (21.8%) occurred during the first month of incarceration, with 15.6% of suicides occurring within the first week after initial reception in prison. Cumulatively, half of all suicides (50.8%) occurred within 6 months of reception. The median length of time between prisoners’ initial reception and their suicides was 171.5 days (corresponding to 5.6 months). When considering interfacility transfers, the majority of all cases (N = 162, 61.8%) died in the prison where they were initially incarcerated; the remaining 100 cases were transferred to another prison at least once during their period of detention. Taking the interval between the last transfer (i.e., reception into the prison where the suicide occurred; the initial confinement for individuals without a transfer) and the suicide into account (Table 3), 21.8% of all suicides took place within 7 days of the last transfer (30.9% within the first month). In this case, the median length of incarceration was 87 days (2.8 months).

Table 2. Demographic and criminological details of all suicide cases (N = 262)

Gender: male

248

94.7

Age, years 18–24

11.8

25–34

36.3

35–44

30.2

45–54

13.7

55–64

6.1

≥65 Nationality: Belgian Marital status: married First incarceration

1.9

203

77.5

24.8

165

52.5

Custodial status Remand

109

41.6

Sentenced

106

40.5

17.9

3.8

ODCIa Sentence lengthb ≤1 year >1–3 years

22.6

>3–5 years

26.4

>5–10 years

24.5

>10 years

22.6

Index/main offence Sexual offence

16.0

Murder/manslaughter

32.4

Acquisitive crimes

29.0

Drug-related offences

11.1

Other

11.5

Clinical Characteristics

Note. aOffenders deemed criminally irresponsible. bSentenced prisoners only (N = 106).

The clinical features of the suicide cases where information was available are presented in Table 4. In this study, 86.7% of prisoners who died by suicide were identified as

Table 3. Time interval from reception and last transfer to suicide (N = 262) Since initial incarcerationa Duration of incarceration

Since last transferb

N (%)

Cumulative %

N (%)

Cumulative %

20 (7.6)

7.6

22 (8.4)

8.4

>1 day – 1 week

21 (8.0)

15.6

35 (13.4)

21.8

>1 week – 1 month

16 (6.1)

21.8

24 (9.2)

30.9

>1–6 months

76 (29.0)

50.8

93 (35.5)

66.4

>6–12 months

39 (14.9)

65.6

42 (16.0)

82.4

>1–5 years

67 (25.6)

91.2

41 (15.6)

98.1

>5 years

23 (8.8)

100

5 (1.9)

100

≤24 hr

Note. aInitial reception into prison, not necessarily the facility where the suicide occurred. bTime interval from last transfer to the prison where the suicide occurred. For the subgroup of prisoners without an interfacility transfer (N = 162), last transfer refers to the initial incarceration.

Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

Circumstances of the Suicides

Table 4. Charted clinical features of suicide cases N (History of) mental disordera b

Valid %

156

86.7

50.4

Before imprisonment

17.7

While incarcerated

35.5

Prior suicide attempt

Note. Percentages calculated based on the number of valid cases. a Data available for 180 cases. bData available for 141 cases.

Table 5. Characteristics of suicide cases (N = 249) Characteristics

Valid %

Single cell

151

60.6

Shared cell

24.1

Solitary confinement

11.2

Outside cell (in prison)

3.6

Outside prison

0.4

00:01–03:00

12.9

03:01–06:00

14.9

06:01–09:00

6.4

09:01–12:00

9.6

12:01–15:00

8.8

15:01–18:00

16.9

18:01–21:00

15.3

21:01–24:00

15.3

224

90.0

Laceration

3.2

Jumping from height

2.8

Self-poisoning/overdose

3.6

Location

Timing

Method Hanging/self-strangulation

Other Increased monitoring measure

0.4

104

41.8

having a (history of) mental disorder, including personality and substance use disorders. In half (50.4%) of all prisoners for whom information was available in their case files, a history of at least one suicide attempt was documented; 17.7% attempted suicide before imprisonment, and one in three did so while incarcerated (35.5%).

Location and Cell Accommodation Table 5 shows the circumstances of 249 prison suicides (95% of all identified cases), since details were not available for 13 cases. In total, 211 suicides took place in the prisoners’ regular cells (84.7%). Over half of all suicides (60.6%) occurred in single-occupant cells, and 60 suicides (24.1%) in shared cells. In the latter case, cellmates were absent from the cells (e.g., due to visitation or airing) in 63.3% of the suicides that occurred in multiple-occupancy cells. Of the suicides that did not occur in regular cells (15.3%), 28 (11.2% of all suicides) took place in solitary confinement.1 The remaining suicides occurred outside the prison cell but within the correctional facility (e.g., in the laundry room or in common areas such as the prison hallway or stairwell), and one suicide took place outside prison (more specifically, when the prisoner was temporarily transferred to a courthouse). Suicide Method The vast majority (90%) of suicides occurred as the result of hanging or self-strangulation (Table 5). Of these suicides (N = 224), the most common ligature was bedding (49.6%) followed by clothing (including shoelaces and belts; 35.3%). Cell window bars were the ligature point used in two thirds of all suicides by hanging (67.0%), followed by inmates’ bed or bunk (9.4%). In the other suicides by hanging, anchoring points consisted of easily accessible cell fittings such as ventilation or heating, pipes, sanitary devices (toilets and sinks) or doors. The remaining suicides (N = 25) involved jumping from a height in prison (2.8%), laceration (3.2%), and self-poisoning or intentional overdose (3.6%). Timing Overall, the timing of suicides was fairly spread by day and by month, and there was no distinguishable pattern. More specifically, suicides were evenly distributed throughout the week, ranging from 13.7% to 14.9% per day. Furthermore, there was no pronounced accumulation of suicides during certain months (range = 6.5–10.3%). There was, however, some discernible pattern by time of day. In all, 56 suicides (24.8%) occurred between 6:00 a.m. and 3:00 p.m., whereas 43.1% occurred during the night shift of prison staff (9:00 p.m. to 6:00 a.m.).

1 In Belgium, solitary confinement refers to the physical and social isolation of a prisoner for disciplinary reasons or protective purposes. Irrespective of reason, prisoners are temporarily housed alone in a specially designed bare-walled cell for 23 hr a day. Inmates placed in a disciplinary or security cell – in practice, a different label for the same type of cell – are subjected to a highly restricted regime and closer levels of monitoring by prison staff (see Shalev, 2015). Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

Significant Events Surrounding the Suicides Of all cases, 33 prisoners (12.6%) had been transferred from another prison within 6 months prior to suicide, with 24 prisoners (9.2%) being transferred to the facility where they died by suicide within 1 month of the suicide. Furthermore, 21.1% of all suicides occurred before or after 5 days of a significant event related to the criminal case, such as a court proceeding, conviction, reconstruction of the crime, or prolongation of the pre-trial detention by the Council Chamber. Increased Monitoring Measure At the time of suicide, 41.8% of all prisoners were subject to conditions of increased monitoring due to a disciplinary or safety measure. This monitoring measure encompasses a visual check by prison staff every 15 or 30 min. Prisoners who are subjected to this measure can be housed in their regular cells in the general prison population, or can be physically and socially isolated from the general prison population in solitary confinement; as was the case for, respectively, 76 and 28 prisoners in the study population.

Discussion The present study reports findings from a review of all documented suicides in the whole Belgian prison estate within a 17-year period. During 2000–2016, suicide was a leading cause of mortality in Belgian prisons, accounting for one third of all deaths during imprisonment. Based on ADP figures, the average annual suicide rate was estimated to be 156.2 per 100,000 prisoners – echoing the recent finding that Belgium has one of the highest prison suicide rates in Western Europe (Fazel et al., 2017). Although not standardized by age and gender, the annual prison suicide rates calculated here are substantially higher than those recorded in the general population in Belgium (see Appendix), and such a large discrepancy can and should not be disregarded. The general demographic profile of the prisoner who died by suicide was that of a single, young (aged 25–34) male of Belgian nationality, which resonates with prior findings from Europe (Gauthier et al., 2015; Radeloff et al., 2017) and Australia (O’Driscoll et al., 2007). This finding is, however, not surprising since this profile mirrors the vast majority of prisoners throughout Belgium, making reliance on such demographic indicators of limited use for identifying at-risk prisoners (Hayes, 2012). For example, 5% of all cases comprised women, reflecting the relative size of the female prison population. Half of all prisoners who died by suicide were charged with, or convicted of, homicide or sexual offences. In light © 2018 Hogrefe Publishing

of data provided by the Council of Europe (Aebi & Delgrande, 2014) and a Belgian study (Favril et al., 2017c), indicating that 20–25% of offenders in Belgium are incarcerated because of such crimes, we can conclude that violent index offences were proportionally more common among suicide cases, which is in keeping with prior international findings (Fruehwald et al., 2004; Webb et al., 2012). Furthermore, nearly one in five suicides involved ODCI, whereas this specific group of offenders only accounts for roughly 8% of the total prison population (DG EPI, 2016). It has been repeatedly stressed (among others by the European Court of Human Rights) that ODCI should not be imprisoned, but rather should receive appropriate care within psychiatric facilities outside of prison (Meysman, 2016; Vandevelde et al., 2011). Although the first high-risk forensic psychiatric hospital in Belgium became operational in November 2014, approximately 900 ODCI are still (unlawfully) residing in prison to date. Hence, reducing the incarceration of ODCI – and offenders with severe mental illness more generally – by diverting them to appropriate facilities for (forensic) mental health care would be a legitimate prevention strategy, both ethically and judicially. Mental illness was, however, not exclusive to the group of ODCI. Clinically, more than 80% of all cases (82% of non-ODCI) had a charted lifetime psychiatric disorder and approximately 50% (47% of non-ODCI) had a documented history of a prior suicide attempt. As a basis of comparison, recent findings from a representative sample of more than 1,000 prisoners in Belgium suggest that lifetime prevalence rates of psychiatric diagnoses (46.3%) and suicide attempts (21.8%) are – albeit highly elevated in reference to their nonincarcerated counterparts – markedly lower compared with those found in the current population of prisoners who died by suicide (Favril, Vander Laenen, & Audenaert, 2017a, 2017b). This resonates with findings from a 2008 meta-analysis identifying such clinical vulnerabilities as significant risk factors for suicide in prisoners (Fazel et al., 2008). Therefore, adequate treatment and management of psychiatric disorders by means of psychological interventions should be provided (Bolton, Gunnell, & Turecki, 2015; Marzano et al., 2016), which is insufficiently the case in Belgian prisons to date (Favril et al., 2017a; Mistiaen et al., 2017). As previously highlighted (O’Driscoll et al., 2007; Shaw et al., 2004), the first months of imprisonment were found to represent a critical risk period for suicide, which underscores the importance of screening for suicide risk during the early stages of custody (for a review, see Gould, McGeorge, & Slade, 2018). Since suicidality is a dynamic process fluctuating over time rather than a static phenomenon (Sveticic & De Leo, 2012), screening for suicide risk should not be limited to the point of reception at prison, but must be an ongoing and systematic process, at regular Crisis (2019), 40(1), 42–53

intervals throughout the period of detention. Such subsequent re-assessments are especially indicated at critical times during their incarceration, when inmates’ circumstances or conditions change while detained (Marzano et al., 2016). As our results suggest, events such as interfacility transfers, periods shortly before and after significant court appearances, and changes in custodial status (after sentencing or appeal) should prompt a screening protocol. In the present study, there was no discernible pattern of suicides by month, nor was there a clear accumulation on particular days of the week (e.g., weekends). However, time of day seems to have some significance, in that two out of five suicides (43.1%) occurred between 9:00 p.m. and 6:00 a.m. Since this period roughly corresponds to the interval between lock-up and reveille of prisoners, this disproportionate number of suicides is possibly due to lower staff supervision during the overnight shift. Not surprisingly (Gauthier et al., 2015), the large majority of all suicides occurred in prisoners’ own cells, in the absence of a cell mate, by means of hanging (most commonly from the cell window bars using bedding or clothing as ligatures). A recent US-based study found that, compared with suicide attempters, prisoners who died by suicide were more likely to be accommodated in single cells, to act during overnight hours, and to use more lethal methods such as hanging (Boren et al., 2017). The latter finding relates to the most accessible method in custody, and further supports the need for restriction of lethal means in correctional settings, especially for prisoners who are identified as suicidal (Gunnell, Bennewith, Hawton, Simkin, & Kapur, 2005). With regard to cell accommodation, single-cell occupancy has been consistently identified as a major risk factor for suicide (Fazel et al., 2008; Humber et al., 2013). Since prisoners housed in single cells may lack informal monitoring provided by cell mates, allowing for more undisturbed opportunities to engage in (lethal) suicidal behavior, placement in shared accommodation has been put forward as a prevention strategy (WHO, 2007). In the present study, however, two thirds of all 60 suicides in multi-occupant cells occurred when prisoners were alone in their cell, even if they were technically sharing one. Similarly, a national study in the UK found that, when prisoners ended their life in shared accommodation, half of them did so in the absence of their cell mate (Shaw et al., 2004). Hence, although we were not able to investigate the number of suicides that may have been prevented by this measure, such a prevention strategy should only be considered a supple-

L. Favril et al., Prison Suicides in Belgium

ment to, and not a substitute for, other prevention efforts such as social support by trained correctional staff (Konrad et al., 2007). At the time of suicide, one in ten prisoners was residing in solitary confinement. Since precise data on utilization of solitary confinement across the Belgian prison estate are elusive, it was not possible to generalize whether this number was disproportionate. However, given the specific Belgian setting2, one in 10 suicides occurring is such restrictive conditions is worrying at the very least. As it inherently reduces protective factors against suicidal behavior (such as purposeful activity and social support), the use of solitary confinement can be detrimental to prisoners’ mental health and well-being (Grassian, 2006; Haney, 2018; Smith, 2006) and has been shown to increase the risk of suicide (Bonner, 2006; Duthé, Hazard, Kensey, & Pan Ké Shon, 2013; Kaba et al., 2014). Therefore, although inappropriate for those at risk of suicide, solitary confinement should only be adopted in exceptional circumstances for those known to pose an acute danger to themselves, for a period as short as necessary. Furthermore, 40% of all prisoners were the subject of an increased monitoring measure, irrespective of cell accommodation. In this regard, it should be emphasized that mere isolation and monitoring, in the absence of any psychosocial support or clinical intervention, does not suffice and should not be considered a stand-alone intervention. Moreover, prisoners may fear such restrictive measures owing to their harsh and often punitive nature, which may discourage the communication of suicidal thoughts or plans to prison staff (Kerkhof & Bernasco, 1990; Way, Kaufman, Knoll, & Chlebowski, 2013). As such, denial or nondisclosure of suicidal feelings impedes the early identification of at-risk prisoners. Collectively, most of the findings outlined here corroborate prior international research on prison suicides, indicating that existing national standards and guidelines for the prevention of prison suicide in other countries (see Daigle et al., 2007) could serve as an evidence-based starting point for the development of a national blueprint for suicide prevention in Belgian prisons, tailored to its specific prison system.

Methodological Limitations The present study was the first endeavor to examine characteristics of the total Belgian penitentiary suicide population, by reviewing official prison files of suicide cases over

2 Solitary confinement in Belgian (and by extension European) prisons is a rather exceptional form of short-term isolation (Shalev, 2015). By international standards (e.g., the United States), Belgium sets a high bar for the use of solitary confinement and restricts its application to no more than seven to nine consecutive days. Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

a relatively long period. The strength of this study lies in the complete nature of a national population (rather than a sample) of prisoners who died by suicide, and the fact that detailed reports were obtained from DG EPI – of which the large majority were available for scrutiny. When interpreting the study’s findings, four important methodological issues should nonetheless be considered. First, only deaths receiving a coroner’s verdict of suicide were included in the study. There may be some degree of inaccuracy in such data; for instance, it is possible that some suicides were not recognized as such. For example, in the absence of a suicide note from which to infer suicidal intent, research has shown that suicides by self-poisoning are not uncommonly misclassified as accidental overdoses or undetermined deaths, especially for individuals with a history of substance use disorders (Bohnert et al., 2013; Olsson, Bradvik, Öjehagen, & Hakansson, 2016; Stone et al., 2017). As such, official figures possibly represent an underestimation of the actual number of suicides in prison. Relatedly, since prison files provided the sole source of information, we were highly reliant on the quantity and quality of archival data available in prisoners’ official records – which are initially not collected for scientific purposes. As file information may be incorrect, biased, or incomplete, a problem inherent to retrospective record-based research is the potential unreliability of the information available in case records (Liebling, 1992). In a similar vein, information relating to the subjective experience of imprisonment and the context in which suicides arise (e.g., social isolation or bullying) was marginally reported in prisoners’ case files. An alternative methodological approach, that of interviewing prisoners who have engaged in near-lethal suicide attempts (Marzano, Rivlin, Fazel, & Hawton, 2009), allows for a broader and deeper range of risk and protective factors to be assessed, as well as the investigation of the psychological processes leading up to suicidal behavior, which is not possible through analyses of official records. A third issue that warrants attention relates to the calculation of suicide rates among prisoners. Several issues have been raised as to whether the prison suicide rate should be calculated based on the ADP, or rather on the number of receptions (i.e., the annual number of unique individuals who entered prison) as denominator (for a discussion, see O’Mahony, 1994). Here, we calculated suicide rates based on ADP figures, which does not factor in the number of receptions into prison. As many prisoners do not stay for a 12-month period, ADP figures will invariably underrepresent the number of people passing through the prison system, consequently leading to the overestimation of the actual rate (Fazel et al., 2017), especially in populations with substantial turnover (Gallagher & Dobrin, 2007). However, © 2018 Hogrefe Publishing

in spite of this issue of overestimation, suicide rates calculated using the ADP are commonly reported in the literature (e.g., Joukamaa, 1997; O’Driscoll et al., 2007; Preti & Cascio, 2006), enabling fair comparisons across studies and countries (Pratt, 2016). Last, and most importantly, the current descriptive study examined characteristics of prisoners who died by suicide without the inclusion of matched controls for each suicide case. Since it is not possible to ascertain risk factors when studying only the suicide group itself, we were unable to delineate which variables were significantly more common among suicide cases, and which characteristics merely reflect those of the whole prison population from which the suicide group was drawn (Liebling, 1992; Pratt, 2016). To partially overcome this limitation, we benchmarked aggregated data of all cases against information about the general Belgian prison population, when available. Ideally, however, our next step should encompass a case-control study in order to elucidate the unique contribution of risk factors to the outcome of suicide in custody.

Implications for Suicide Prevention As in other countries, suicide is common among prisoners in Belgium, underscoring the seriousness of this preventable cause of death. By stark contrast, a national strategy for suicide prevention is currently lacking in Belgian prisons (Favril et al., 2017c). Yet, several prominent studies have unambiguously concluded that multifaceted and prison-wide prevention strategies are of paramount importance in order to reduce the incidence of suicidal behavior in custodial settings (Barker, Kõlves, & De Leo, 2014; Konrad et al., 2007; Marzano et al., 2016). To date, however, the implemented prevention measures in Belgian prisons are for the most part restrictive in nature (such as increased monitoring and placement in solitary confinement). As our results indicate, the fact that inmates residing in solitary confinement – conditions that are supposed to be safe and subject to increased levels of monitoring – still had both the means and the opportunity to engage in fatal suicidal behavior represents a structural frailty in Belgian prisons’ approach to suicide prevention. Hence, although temporarily useful to a certain degree, suicide prevention should not be limited to such situational prevention strategies, as these do not address the underlying problem (Kerkhof & Bernasco, 1990; Liebling, 1992). Therefore, prevention efforts should equally target the psychosocial etiology of suicidal thoughts and behavior. In this respect, international good practices include the Assessment, Care in Custody and Teamwork (ACCT) procedure as implemented in UK prisons (Humber, Hayes, Senior, Fahy, & Shaw, 2011), the provision of appropriate mental health care (Marzano Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

et al., 2016), peer support initiatives (Bagnall et al., 2015; Hall & Gabor, 2004), and cognitive-behavioral therapy for suicidal inmates (Pratt et al., 2015). In sum, rather than relying on physical methods of suicide prevention, more emphasis should be placed on screening and risk assessment, providing training for prison staff, promoting suicide awareness, and ensuring that multi-disciplinary care and support are available for prisoners at risk. Considering the ongoing austerity measures in the Federal Public Service of Justice, the persistent problem of overcrowding, and the subsequent increasing workload for prison staff, the prevention of suicide in prison currently does not constitute a pronounced policy priority in Belgium. Acknowledging this unfortunate reality, the authors nonetheless emphasize that (more) policy attention toward suicide and its prevention is warranted – not only to reduce the incidence of suicidal behavior among prisoners and its associated costs (Howard League, 2016), but as much as part of a safe and humane execution of the custodial measure in Belgium. Only a holistic and whole-prison approach toward suicide prevention will be able to reduce the number of suicides in prisoners as well as the burden and suffering caused for relatives, incarcerated peers, and prison staff.

Conclusion Suicide is a leading cause of death in Belgian prisons, warranting a wide implementation of preventive interventions. Given the lives at stake, it is imperative that such prevention strategies are maximally underpinned by high-quality research evidence. Whereas the present study – the first of its kind in Belgium – provides a modest contribution in this direction, further research (both quantitative and qualitative) is needed in order to elucidate causal risk factors and possible points of intervention in the early phases of the suicidal process. The authors firmly emphasize the timely need for elaborating national standards and guidelines for suicide prevention in Belgian prisons, equivalent to the general population (WHO, 2012). Acknowledgments This project was supported by the Special Research Fund (BOF) of Ghent University. The funding agency did not have any involvement in the study design, data collection, data analysis, and the interpretation of results, nor in the writing or decision to submit the manuscript for publication. Our sincere appreciation is extended to Dirk Janssens and Patrick Franssen (DG EPI) for facilitating the study and for providing the necessary data. The authors declare that the research was conducted in the absence of any commerCrisis (2019), 40(1), 42–53

cial or financial relationships that could be construed as a potential conflict of interest.

References Aebi, M. F., & Delgrande, N. (2014). SPACE I – Council of Europe annual penal statistics: Prison populations. Survey 2012. Strasbourg, France: Council of Europe. Bagnall, A. M., South, J., Hulme, C., Woodall, J., Vinall-Collier, K., Raine, G., … Wright, N. M. (2015). A systematic review of the effectiveness and cost-effectiveness of peer education and peer support in prisons. BMC Public Health, 15, 290. Barker, E., Kõlves, K., & De Leo, D. (2014). Management of suicidal and self-harming behaviors in prisons: Systematic literature review of evidence-based activities. Archives of Suicide Research, 18(3), 227–240. Blaauw, E., Kerkhof, A. J. F. M., & Hayes, L. M. (2005). Demographic, criminal, and psychiatric factors related to inmate suicide. Suicide and Life-Threatening Behavior, 35(1), 63–75. Blaauw, E., Winkel, F. W., & Kerkhof, A. J. F. M. (2001). Bullying and suicidal behavior in jails. Criminal Justice and Behavior, 28(3), 279–299. Bohnert, A. S. B., McCarthy, J. F., Ignacio, R. V., Ilgen, M. A., Eisenberg, A., & Blow, F. C. (2013). Misclassification of suicide deaths: Examining the psychiatric history of overdose decedents. Injury Prevention, 19(5), 326–330. Bolton, J. M., Gunnell, D., & Turecki, G. (2015). Suicide risk assessment and intervention in people with mental illness. BMJ, 351, h4978. Bonner, R. L. (2006). Stressful segregation housing and psychosocial vulnerability in prison suicide ideators. Suicide and Life-Threatening Behavior, 36(2), 250–254. Boren, E. A., Folk, J. B., Loya, J. M., Tangney, J. P., Barboza, S. E., & Wilson, J. S. (2017). The suicidal inmate: A comparison of inmates who attempt versus complete suicide. Suicide and Life-Threatening Behavior. https://doi.org/10.1111/sltb.12374 Daigle, M. S., Daniel, A. E., Dear, G. E., Frottier, P., Hayes, L. M., Kerkhof, A., … Sarchiapone, M. (2007). Preventing suicide in prisons, part II. Crisis, 28(3), 122–130. https://doi.org/10.1027/02275910.28.3.122 De Smet, S., De Keyzer, W., De Donder, L., Ryan, D., Verté, D., Broekaert, E., & Vandevelde, S. (2016). Older offenders deemed criminally irresponsible in Flanders (Belgium): Descriptive results from a retrospective case note study. International Journal of Law and Psychiatry, 46, 35–41. https://doi.org/10.1016/j. ijlp.2016.02.025 Directory-General of Penitentiary Institutions. (2016). Annual report 2015. Brussels, Belgium: Author. Duthé, G., Hazard, A., Kensey, A., & Shon, J. L. P. K. (2013). Suicide among male prisoners in France: A prospective population-based study. Forensic Science International, 233(1-3), 273–277. Dye, M. H. (2010). Deprivation, importation, and prison suicide: Combined effects of institutional conditions and inmate composition. Journal of Criminal Justice, 38(4), 796–806. Enggist, S., Møller, L., Galea, G., & Udesen, C. (2014). Prisons and health. Geneva, Switzerland: World Health Organization. Favril, L., Vander Laenen, F., & Audenaert, K. (2017a). Psychiatrische morbiditeit bij gedetineerden in Vlaanderen [Psychiatric morbidity among prisoners in the Flanders region of Belgium]. Panopticon, 38(4), 231–245. Favril, L., Vander Laenen, F., & Audenaert, K. (2017b). Suïcidaal gedrag bij gedetineerden in Vlaanderen: Prevalentie en sa© 2018 Hogrefe Publishing

L. Favril et al., Prison Suicides in Belgium

menhang met psychische distress [Suicidal behavior among prisoners: Prevalence and association with psychological distress in Flemish prisons]. Tijdschrift voor Psychiatrie, 59(4), 203–211. Favril, L., Vander Laenen, F., Vandeviver, C., & Audenaert, K. (2017c). Suicidal ideation while incarcerated: Prevalence and correlates in a large sample of male prisoners in Flanders, Belgium. International Journal of Law and Psychiatry, 55, 19–28. Fazel, S., & Baillargeon, J. (2011). The health of prisoners. Lancet, 377(9769), 956–965. Fazel, S., Cartwright, J., Norman-Nott, A., & Hawton, K. (2008). Suicide in prisoners: A systematic review of risk factors. Journal of Clinical Psychiatry, 69(11), 1721–1731. Fazel, S., Hayes, A.J., Bartellas, K., Clerici, M., & Trestman, R. (2016). Mental health of prisoners: Prevalence, adverse outcomes, and interventions. Lancet Psychiatry, 3(9), 871–881. Fazel, S., Ramesh, T., & Hawton, K. (2017). Suicide in prisons: An international study of prevalence and contributory factors. Lancet Psychiatry, 4(12), 946–952. Fruehwald, S., Matschnig, T., Koenig, F., Bauer, P., & Frottier, P. (2004). Suicide in custody: Case-control study. British Journal of Psychiatry, 185(6), 494–498. Gallagher, C. A., & Dobrin, A. (2007). Risk of suicide in juvenile justice facilities: The problem of rate calculations in high-turnover populations. Criminal Justice and Behavior, 34(10), 1362–1376. Gauthier, S., Reisch, T., & Bartsch, C. (2015). Swiss prison suicides between 2000 and 2010. Crisis, 36(2), 110–116. https://doi. org/10.1027/0227-5910/a000302 Goomany, A., & Dickinson, T. (2015). The influence of prison climate on the mental health of adult prisoners: A literature review. Journal of Psychiatric and Mental Health Nursing, 22(6), 413–422. Gould, C., McGeorge, T., & Slade, K. (2018). Suicide screening tools for use in adult offenders: A systematic review. Archives of Suicide Research, 22(3), 345–364. https://doi.org/10.1080/13811118.20 17.1334611 Grassian, S. (2006). Psychiatric effects of solitary confinement. Journal of Law & Policy, 22, 325–383. Gunnell, D., Bennewith, O., Hawton, K., Simkin, S., & Kapur, N. (2005). The epidemiology and prevention of suicide by hanging: A systematic review. International Journal of Epidemiology, 34(2), 433–442. Hall, B., & Gabor, P. (2004). Peer suicide prevention in a prison. Crisis, 25(1), 19–26. https://doi.org/10.1027/0227-5910.25.1.19 Haney, C. (2018). Restricting the use of solitary confinement. Annual Review of Criminology, 1, 285–310. Hayes, L. M. (2012). National study of jail suicide: 20 years later. Journal of Correctional Health Care, 18(3), 233–245. Howard League (2016). The cost of prison suicide. London, UK: Howard League for Penal Reform. Huey, M. P., & McNulty, T. L. (2005). Institutional conditions and prison suicide: Conditional effects of deprivation and overcrowding. The Prison Journal, 85(4), 490–514. Humber, N., Hayes, A., Senior, J., Fahy, T., & Shaw, J. (2011). Identifying, monitoring and managing prisoners at risk of self-harm/ suicide in England and Wales. The Journal of Forensic Psychiatry & Psychology, 22(1), 22–51. Humber, N., Webb, R., Piper, M., Appleby, L., & Shaw, J. (2013). A national case-control study of risk factors among prisoners in England and Wales. Social Psychiatry and Psychiatric Epidemiology, 48(7), 1177–1185. Joukamaa, M. (1997). Prison suicide in Finland, 1969–1992. Forensic Science International, 89(3), 167–174. Kaba, F., Lewis, A., Glowa-Kollisch, S., Hadler, J., Lee, D., Alper, H., … Venters, H. (2014). Solitary confinement and risk of self-harm among jail inmates. American Journal of Public Health, 104(3), 442–447. https://doi.org/10.2105/AJPH.2013.301742 © 2018 Hogrefe Publishing

Kerkhof, A. J. F. M., & Bernasco, W. (1990). Suicidal behavior in jails and prisons in the Netherlands: Incidence, characteristics, and prevention. Suicide and Life-Threatening Behavior, 20(2), 123– 137. Konrad, N., Daigle, M. S., Daniel, A. E., Dear, G. E., Frottier, P., Hayes, L. M., … Sarchiapone, M. (2007). Preventing suicide in prisons, part I. Crisis, 28(3), 113–121. https://doi.org/10.1027/02275910.28.3.113 Leese, M., Thomas, S., & Snow, L. (2006). An ecological study of factors associated with rates of self-inflicted death in prisons in England and Wales. International Journal of Law and Psychiatry, 29(5), 355–360. Liebling, A. (1992). Suicides in prison. London, UK: Routledge. Liebling, A. (2006). The role of the prison environment in prison suicide and prisoner distress. In G. E. Dear (Ed.), Preventing suicide and other self-harm in prison (pp. 16–28). Houndmills, UK: Palgrave Macmillan. Liebling, A., & Ludlow, A. (2016). Suicide, distress and the quality of prison life. In Y. Jewkes, B. Crewe, & J. Bennett (Eds.), Handbook on prisons (pp. 224–245). London, UK: Routledge. Marzano, L., Hawton, K., Rivlin, A., Smith, E. N., Piper, M., & Fazel, S. (2016). Prevention of suicidal behavior in prisons. Crisis, 37(5), 323–334. https://doi.org/10.1027/0227-5910/a000394 Marzano, L., Rivlin, A., Fazel, S., & Hawton, K. (2009). Interviewing survivors of near-lethal self-harm: A novel approach for investigating suicide amongst prisoners. Journal of Forensic and Legal Medicine, 16(3), 152–155. Meysman, M. (2016). The tension between cross-border cooperation in the European Area of Freedom, Security and Justice and the fundamental rights of mentally ill offenders in detention. International Journal of Law and Psychiatry, 47, 136–147. Mistiaen, P., Dauvrin, M., Eyssen, E., Roberfroid, D., San Miguel, L., & Vinck, I. (2017). Health care in Belgian prisons. Brussels, Belgium: Belgian Health Care Knowledge Centre (KCE). National Institute for Health and Care Excellence. (2017). Mental health of adults in contact with the criminal justice system. London, UK: Author. O’Connor, R.C., & Nock, M.K. (2014). The psychology of suicidal behaviour. Lancet Psychiatry, 1(1), 73–85. O’Driscoll, C., Samuels, A., & Zacka, M. (2007). Suicide in New South Wales prisons, 1995-2005: Towards a better understanding. Australian and New Zealand Journal of Psychiatry, 41(6), 519–524. O’Mahony, P. (1994). Prison suicide rates: What do they mean? In A. Liebling & T. Ward (Eds.), Deaths in custody: International perspectives (pp. 45–57). London, UK: Whiting & Birch. Olsson, M. O., Bradvik, L., Ojehagen, A., & Hakansson, A. (2016). Risk factors for unnatural death: Fatal accidental intoxication, undetermined intent and suicide: register follow-up in a criminal justice population with substance use problems. Drug and Alcohol Dependence, 162, 176–181. Organisation for Economic Cooperation and Development. (2017). Health at a glance 2017. Paris, France: Author. Pratt, D. (2016). The prevention of suicide in prison. New York, NY: Routledge. Pratt, D., Tarrier, N., Dunn, G., Awenat, Y., Shaw, J., Ulph, F., & Gooding, P. (2015). Cognitive-behavioural suicide prevention for male prisoners: A pilot randomized controlled trial. Psychological Medicine, 45(16), 3441–3451. https://doi.org/10.1017/ S0033291715001348 Preti, A., & Cascio, M. T. (2006). Prison suicides and self-harming behaviours in Italy, 1990-2002. Medicine, Science and the Law, 46(2), 127–134. Rabe, K. (2012). Prison structure, inmate mortality and suicide risk in Europe. International Journal of Law and Psychiatry, 35(3), 222–230. Crisis (2019), 40(1), 42–53

Radeloff, D., Lempp, T., Kettner, M., Rauf, A., Bennefeld-Kersten, K., & Freitag, C. M. (2017). Male suicide rates in German prisons and the role of citizenship. PLoS ONE, 12(6), e0178959. Rivlin, A., Hawton, K., Marzano, L., & Fazel, S. (2013). Psychosocial characteristics and social networks of suicidal prisoners: Towards a model of suicidal behaviour in detention. PLoS ONE, 8(7), e68944. Shalev, S. (2015). Solitary confinement: The view from Europe. Canadian Journal of Human Rights, 4(1), 143–165. Shaw, J., Baker, D., Hunt, I.M., Moloney, A., & Appleby, L. (2004). Suicide by prisoners: National clinical survey. British Journal of Psychiatry, 184(3), 263–267. Smith, S.P. (2006). The effects of solitary confinement on prison inmates: A brief history and review of the literature. Crime and Justice, 34(1), 441–528. Stoliker, B. E. (2018). Attempted suicide: A multilevel examination of inmate characteristics and prison context. Criminal Justice and Behavior, 45(5), 589–611. Stone, D. M., Holland, K. M., Bartholow, B., Logan, J., LiKamWa McIntosh, W., Trudeau, A., … Rockett, I. R. H. (2017). Deciphering suicide and other manners of death associated with drug intoxication: A Centers for Disease Control and Prevention consultation meeting summary. American Journal of Public Health, 107(8), 1233–1239. https://doi.org/10.2105/AJPH.2017.303863 Sveticic, J., & De Leo, D. (2012). The hypothesis of a continuum in suicidality: A discussion on its validity and practical implications. Mental Illness, 4(2), 73–78. Turecki, G., & Brent, D.A. (2016). Suicide and suicidal behaviour. Lancet, 387(10024), 1227–1239. van Heeringen, K. (2012). Stress-diathesis model of suicidal behavior. In Y. Dwivedi (Ed.), The neurobiological basis of suicide (pp. 113–123). Boca Raton, FL: CRC Press. Vandevelde, S., Soyez, V., Vander Beken, T., De Smet, S., Boers, A., & Broekaert, E. (2011). Mentally ill offenders in prison: The Belgian case. International Journal of Law and Psychiatry, 34(1), 71–78. Walmsley, R. (2016). World prison population list (11th ed.). London, UK: Institute for Criminal Policy Research. Way, B. B., Kaufman, A. R., Knoll, J. L., & Chlebowski, S. M. (2013). Suicidal ideation among inmate-patients in state prison: Prevalence, reluctance to report, and treatment preferences. Behavioral Sciences & the Law, 31(2), 230–238. Webb, R. T., Shaw, J., Stevens, H., Mortensen, P. B., Appleby, L., & Qin, P. (2012). Suicide risk among violent and sexual criminal offenders. Journal of Interpersonal Violence, 27(17), 3405–3424.

Crisis (2019), 40(1), 42–53

L. Favril et al., Prison Suicides in Belgium

World Health Organization. (2007). Preventing suicide in jails and prisons. Geneva, Switzerland: Author. World Health Organization. (2012). Public health action for the prevention of suicide: a framework. Geneva, Switzerland: Author.

Received July 14, 2017 Revision received December 24, 2017 Accepted January 2, 2018 Published online July 27, 2018 Louis Favril, MSc Clinical Psychology and MSc Criminology, is a PhD researcher and academic assistant at the Faculty of Law and Criminology, Ghent University, Belgium. His PhD research focuses on suicidal thoughts and behavior in prison. Other research interests include forensic psychiatry, substance use, and drug policy. Ciska Wittouck, MSc Clinical Psychology and MSc Criminology, is a PhD researcher at the Faculty of Medicine and Health Sciences, Ghent University, Belgium. Her PhD research focuses on procedural justice in mentally ill offenders. Other research interests include prison suicide and drug treatment courts. Kurt Audenaert, MD, PhD, is full professor of psychiatry at the Faculty of Medicine and Health Sciences, Ghent University, Belgium. His expertise focuses on forensic psychiatry, neuroscience, affective disorders, and impulsive behavior. Freya Vander Laenen, PhD, is professor of criminology at the Faculty of Law and Criminology, Ghent University, Belgium. Her expertise focuses on vulnerable individuals (because of substance use, mental illness, and/or social exclusion) in contact with the criminal justice system.

Louis Favril Ghent University, Campus Aula Universiteitstraat 4 9000 Ghent Belgium louis.favril@ugent.be

L. Favril et al., Prison Suicides in Belgium

Appendix Table A1. Prison suicide rates compared with the general population, Belgium (2000–2015) Suicide rate (per 100,000) Year

General population

Prison

Crude relative ratea

2000

20.9

224.5

10.7

2001

21.1

270.7

12.8

2002

20.1

250.9

12.5

2003

19.8

122.1

6.2

2004

19.1

86.9

4.5

2005

19.4

140.7

7.2

2006

18.4

146.2

7.9

2007

17.5

141.8

8.1

2008

18.7

161.8

8.6

2009

18.7

127.0

6.8

2010

18.5

189.8

10.3

2011

19.0

109.4

5.7

2012

18.3

114.7

6.3

2013

17.1

128.8

7.5

2014

17.0

172.7

10.2

2015

16.6

144.9

8.7

18.8

158.3

8.4

Note. Data on suicide rates in the general population were not yet available for 2016. aPrison suicide rates compared with general population suicide rates.

Crisis (2019), 40(1), 42–53

Research Trends

Self-Harm and Suicide Coverage in Sri Lankan Newspapers An Analysis of the Compliance With Recommended Guidelines Jane Brandt Sørensen1,6, Melissa Pearson5,6, Martin Wolf Andersen1, Manjula Weerasinghe2,6, Manjula Rathnaweera3, D. G. Chathumini Rathnapala4, Michael Eddleston5,6, and Flemming Konradsen1,6 Department of Public Health, University of Copenhagen, Denmark Department of Community Medicine, Faculty of Medicine and Allied Sciences, Rajarata University of Sri Lanka, Sri Lanka 3 Thalawa Maha Vidyalaya, Thalawa, Sri Lanka 4 Department of English, University of Colombo, Sri Lanka 5 Pharmacology, Toxicology, & Therapeutics, University/British Heart Foundation Centre for Cardiovascular Science, University of Edinburgh, UK 6 South Asian Clinical Toxicology Research Collaboration (SACTRC), Faculty of Medicine, University of Peradeniya, Sri Lanka 1 2

Abstract. Background: Irresponsible media reporting may influence suicidal behavior. Adherence to guidelines for responsible reporting of suicide has not been examined in Sri Lanka in recent times. Aims: To examine the quality of reporting on self-harm and suicide in Sri Lankan newspapers and compare the quality between Sinhala and English newspapers. Method: From December 1, 2014 to January 31, 2015, 407 editions of newspapers were screened. Reporting quality was measured using the PRINTQUAL tool. Results: We identified 68 articles covering an episode of self-harm or suicide (42 Sinhala and 26 English). The majority of articles were noncompliant with guidelines for sensitive reporting. Indicators of noncompliance included that newspaper articles frequently reported method in the headline (53%), included detailed characteristics of the individual (100%), used insensitive language (58% of English articles), and attributed a single-factor cause to the self-harm (52%). No information about help-seeking was included. Limitations: The study involved a relatively short period of data collection. Including social media, Tamil language newspapers, and online publications would have provided additional understanding of reporting practices. Conclusion: The majority of Sri Lankan newspapers did not follow the principles of good reporting, indicating a need for further training of journalists. Keywords: suicide, self-harm, suicide reporting, media guidelines, Sri Lanka

A number of studies have shown a relationship between practices in media reporting on suicide and subsequent suicidal behavior in vulnerable individuals (Pirkis, Mok, Robinson, & Nordentoft, 2016). There is evidence that reporting can introduce or spread new means of suicide (Chen, Chen, Gunnell, & Yip, 2013) and lead to the socalled Werther effect, where readers imitate the suicides portrayed in the media (Pirkis, Blood, Beautrais, Burgess, & Skehan, 2006). However, responsible coverage of suicide can have a preventive effect, help de-stigmatize and educate, as well as encourage help-seeking – in addition to ensuring decency and respect for the bereaved (Bohanna & Wang, 2012; Niederkrotenthaler et al., 2010; World Health Organization [WHO], 2008). Several guidelines for responsible reporting on suicide have been developed by governments, nongovernmental Crisis (2019), 40(1), 54–61 https://doi.org/10.1027/0227-5910/a000534

organizations, and media entities at national and global levels, in particular the WHO (Pirkis et al., 2006). In 2000, the WHO produced a set of media guidelines encouraging sensitive reporting of suicide (WHO, 2000) that was updated in 2008 (WHO, 2008) and again in 2017 (WHO, 2017). Recommendations for the media formed part of the WHO guidelines for the development of national suicide prevention strategies (WHO, 2014). The majority of media guidelines for responsible reporting of suicide are similar in their approach and include (a) advice on avoiding sensationalism and explicit descriptions of the method and site associated with the suicide as well as undue repetition of stories about suicide and (b) recommendations to include information about where to seek help. Over 800,000 people die by suicide annually (WHO, 2014). Low- and middle-income countries bear the high© 2018 Hogrefe Publishing

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

est burden and almost 40% of all global suicides occur in such countries in East and South East Asia (WHO, 2014). Sri Lanka – where this study was conducted – recorded one of the highest suicide rates worldwide in 1995 with 51.5/100,000 population (Knipe, Metcalfe, & Gunnell, 2015). This has since decreased to an estimated 14/100,000 population in 2015 (Knipe, Chang et al., 2017). In addition, a considerable number of individuals still self-harm – it was recently estimated that 339/100,000 population self-harmed in the Anuradhapura area of Sri Lanka (Knipe, Gunnell et al., 2017). In Sri Lanka, a few resources guiding the media in their reporting of suicides exist. The Centre for Policy Alternatives, Sri Lanka, and PressWise Trust, UK, developed the Suicide Sensitive Journalism Handbook in 2003. This handbook included an analysis of 84 printed newspaper articles reporting on suicide as well as recommendations for responsible reporting for Sri Lankan media professionals (Deshapriya, Hattotuwa, & Jempson, 2003). The Editor’s Guild of Sri Lanka includes one recommendation for reporting on suicide: to avoid presenting excessive details of methods used (The Editors’ Guild of Sri Lanka, 2014). To our knowledge, no peer-reviewed analysis has been conducted on the quality of reporting on suicide in Sri Lanka. This study aims to examine the quality of reporting on episodes of self-harm and suicide in Sri Lankan newspapers. More specifically, objectives were to: 1. Assess how reporting on self-harm and suicide in Sri Lankan newspapers complies with suicide reporting guidelines utilizing the PRINTQUAL tool. 2. Compare the quality of reporting on self-harm and suicide between Sinhala and English newspapers.

Method Search Strategy A media analysis was carried out over a 2-month period from December 1, 2014 to January 31, 2015. Seven national printed newspapers were searched, chosen due to their popularity and wide circulation: three out of seven available Sinhala newspapers (Lankadeepa, Divaina, and Ada) and four out of four available English newspapers (Daily News, Daily Mirror, The Island, and Ceylon Today). All newspapers had daily editions, except for Ada that was sold 5 days of the week. All newspaper articles were hand searched for reports on self-harm and suicide. M. W. and M. R. screened the Sinhala newspapers while M. W. A. screened the English newspapers for articles. All Sinhala articles were translated from Sinhala to English by C. R., a graduate student in © 2018 Hogrefe Publishing

English literature and who is fluent in both Sinhala and English. Following translation, J. B. S. screened all articles in accordance with the inclusion and exclusion criteria. Newspaper articles were included when they reported on a specific case of self-harm or suicide and excluded when they did not report on a specific case of self-harm or suicide. Excluded articles comprised reports on suicide bombing or reports where neither suicide nor self-harm was mentioned. Additionally, reports of deceased under the age of 10 years were excluded owing to the difficulties in assessing intent in line with most official suicide statistics. For an overview of the newspaper article selection process, refer to Figure 1. Each included Sinhala article was independently rated by both C. R. and J. B. S. utilizing the PRINTQUAL tool. Ten of the English articles were rated by F. K. to ensure agreement of the coding.

Quality Assessment International studies have found that while strategies in the form of guidelines on suicide reporting are effective, implementation is still lacking in many national contexts (Pirkis et al., 2006). The PRINTQUAL tool was developed to create a common reference system for measuring the quality of suicide reporting and compliance to guidelines (John et al., 2014). It comprises two scales of poor and good reporting on suicides (19 poor-quality and four good-quality items) as well as a weighted scoring system to measure the perceived degree of negative or positive impact each item has on the general population (see Table 1). Following the example of John et al. in their study of cluster suicides in the UK, we in this study only measured the frequency of nonweighted items (John et al., 2017). The weighting for each item is, however, still reflected in Table 1. Considering the high number of individuals who selfharm in Sri Lanka, we included articles that covered such episodes. Although the PRINTQUAL tool was developed for the reporting of suicide (John et al., 2014), we assessed that the tool was also relevant to use for reporting on episodes of self-harm. In this report it should be noted that we use the term self-harm and not suicide attempt to reflect the context of suicide in Sri Lanka as it does not necessarily entail an intent to die (Eddleston, Sheriff, & Hawton, 1998). The PRINTQUAL tool includes recommended phrases to be used or avoided in Items 6 and 20. For the English newspaper articles, we used the list of suggestions from the Suicide Sensitive Journalism Handbook (Deshapriya et al., 2003) that are in line with other guidelines, that is, the Samaritans’ guidelines also used by the PRINTQUAL authors (Samaritans, 2008; John et al., 2014). No such list of appropriate Sinhala wording was available. Although the translation for this study was thorough, we were conCrisis (2019), 40(1), 54–61

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

cerned that it might not have fully captured how specific words and phrases contain different meanings in the Sinhala and English languages. We therefore decided not to include a rating of the appropriateness of wording for the Sinhala newspaper articles. For all articles, we recorded: (a) whether the articles reported on a case of self-harm or suicide; (b) the themes explained to cause the self-harm/suicide; and (c) the reporting of personal characteristics (gender, name, place of residence, and occupation). Drawings of the deceased and means were a feature in the articles and we therefore also recorded this.

Results Screening and Selection of Newspaper Articles The selection process is depicted in Figure 1. In total, 397 editions of seven newspapers were screened; 164 Sinhala and 233 English. Screening revealed 138 articles of which 70 were excluded because they did not report on a specific case of self-harm or suicide. In total, 68 articles were included in the study: 26 were English-language and 42 were Sinhala-language articles.

Newspaper and Reporting Characteristics One newspaper had not published any articles on selfharm or suicide within the set time frame. The majority of the included articles were found in the Sinhala newspapers

Figure 1. Newspaper article selection process.

Divaina (n = 27) and Lankadeepa (n = 15) and in the English newspaper Ceylon Today (n = 16) – all privately owned, independent newspapers (Press Reference, 2017). Of the 68 newspaper articles, 53 reported on a specific case of self-harm or suicide. In the remaining 15 newspaper articles the words self-harm or suicide were mentioned as possible explanations of an episode that could also be found to be an accident or homicide. Of all 68 newspaper articles, 79% were reports of suicides. In three of the Sinhala-language newspaper articles (7%), more than one episode of self-harm and suicide were mentioned. Five of the English-language newspaper articles (19%) reported on suicides outside of Sri Lanka. The most common format of newspaper reporting was brief police reports, with 79% of Sinhala (n = 33) and 58% of English (n = 15) newspaper articles using this presentation. These reports only briefly explained the episode of self-harm or suicide. More in-depth coverage of the events was provided in 25% (n = 17) of articles. Of these, seven included hypothetical emotional states and conversations leading up to the case of self-harm or suicide.

Reporting Quality Table 1 shows the frequency of the PRINTQUAL poorquality and good-quality items for each category of articles included. The poor-quality items rated per newspaper article ranged from zero to eight out of the 19 possible items. Of the 68 newspaper articles, 4% had no poor-quality scorings. The range of good-quality items per newspaper article ranged from zero to one out of the four possible items. A total of 85% of newspaper articles had no good-quality scorings. The method of self-harm and suicide was mentioned in the headline in 53% of newspaper articles and technical Figure 1. Newspaper article selection process.

Total newspaper editions screened: 233 English newspapers (59 Daily News, 58 Daily Mirror, 58 The Island, and 58 Ceylon Today) and 164 Sinhala newspapers (58 Lankadeepa, 57 Divaina, and 49 Ada). Total newspaper articles on selfharm and suicide screened: 74 English and 64 Sinhala newspaper articles (N = 138)

Total newspaper articles on selfharm and suicide included: 26 English and 42 Sinhala newspaper articles (N = 68)

Crisis (2019), 40(1), 54–61

Total newspaper articles excluded: Self-harm/suicide not mentioned (9 English and 5 Sinhala) Suicide bombing (27 English and 10 Sinhala) Children under the age of 10 (3 Sinhala) Court case reporting of a criminal offense (3 Sinhala) Political or educational views (6 English and 1 Sinhala) Fiction (1 English) International news not reporting on specific self-harm/suicide case (5 English) (N = 70)

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

Table 1. PRINTQUAL: frequency of each item in total and for Sinhala and English articles Weighting

Frequency present (%), All

Sinhala articles on suicide/self-harm (%)

English articles on suicide/self-harm (%)

n = 68

n = 42

n = 26

8 (11.8)

4 (9.5)

4 (15.4)

Poor-quality items Is the article on the front page?

Is the main headline on the front page?

Is the method mentioned in the headline?

36 (52.9)

26 (61.9)

10 (38.5)

Does the article cover over 50% of the page?

7 (10.3)

5 (11.9)

2 (7.7)

Is it on Page 3?

14 (20.6)

13 (31.0)

1 (3.8)

Does the article use phrases to be avoided as stated guidelines?

15 (57.7)

–

15 (57.7)

Are explicit or technical details of the method described?

5 (7.4)

2 (4.8)

3 (11.5)

Are technical details of an unusual method for the locality described?

4 (5.9)

1 (2.4)

3 (11.5)

Are the contents of a suicide note described?

1 (1.5)

1 (3.8)

Does it mention or refer to a suicide hotspot?

Does it report positive outcomes from the death?

2 (2.9)

2 (7.7)

Is the cause of the suicide attributed to a single factor?

35 (51.5)

22 (52.4)

13 (50.0)

Is there repeated reporting of earlier suicides in the article?

2 (2.9)

1 (2.4)

1 (3.8)

Does the article report whether the person knew previous suicides or that the timing implies a link?

3 (4.4)

2 (4.8)

1 (3.8)

Does the article highlight community expressions of grief?

7 (10.3)

3 (7.1)

4 (15.4)

Does the article include interviews with the bereaved?

15 (22.1)

12 (28.6)

3 (11.5)

Does the article include photographs of the scene, location, or method?

4 (5.9)

3 (7.1)

1 (3.8)

Does the article include a photograph of the deceased?

14 (20.6)

10 (23.8)

4 (15.4)

Does the article mention a celebrity suicide?

Good-quality items Does the article include recommended language as based on guidelines?

1 (3.8)

–

1 (3.8)

Does the article describe complex or multifactorial causes of death?

1 (1.5)

1 (3.8)

Does it include sources of information or advice?

Does it take the opportunity to educate the reader?

details of the method were provided in 7%. Table 2 presents an overview of the methods used in the headline or text of reported cases of self-harm and suicide as well as the frequency of reporting. In seven newspaper articles, several methods were described. Of all 68 articles, 6% described an unusual method for the area, which included cutting one’s throat with a knife and swallowing a number of substances and materials like hair clips and ashes mixed with soap water. In Table 3 the frequency of personal characteristics reported in the newspaper articles is highlight© 2018 Hogrefe Publishing

ed. Notably, Sinhala-language newspaper articles more frequently included personal information than did English-language newspaper articles. Of all the 68 newspaper articles, 21% (n = 14) were printed on Page 3, which is considered prominent placement. Further, 12% of articles (n = 8) were on the front page; however, the headline was never the main headline on the front page. A photograph of the deceased or the location of the episode was included in 26% (n = 18) of the 68 newspaper Crisis (2019), 40(1), 54–61

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

Discussion

Table 2. Method of self-harm or suicide portrayed in articles All

Sinhala articles on suicide/selfharm

English articles on suicide/selfharm

Hanging

Drowning

Self-immolation

Pesticide poisoning

Jumping into a well

Jumping from a height

Knife wounds

Firearms

Medicine

Jumping in front of a train

Unusual events

Table 3. Personal characteristics of individuals portrayed in articles

Gender

Frequency present (%), All

Sinhala articles on suicide/ self-harm

English articles on suicide/ self-harm

n = 68

n = 42

n = 26

68 (100.0)

42 (100.0)

26 (100.0)

Age

52 (76.5)

34 (81.0)

18 (69.2)

Place of residence

54 (79.4)

34 (81.0)

20 (76.9)

Name

51 (75.0)

35 (83.3)

16 (61.5)

Occupation

31 (45.6)

21 (50.0)

10 (38.5)

articles; the deceased was pictured in 21% (n = 14) of newspaper articles – primarily in the Sinhala articles. The English articles frequently used drawings of the deceased and/or method used – this was the case in 42% of English newspaper articles (n = 11). In 58% (n = 15) of the English newspaper articles, phrases were used that should be avoided when reporting on suicide. Phrases recommended when describing suicide were used in one English newspaper article. No newspaper articles included information about where to seek help or educated the reader about suicide. In 22% (n = 15) of the 68 articles an interview with a bereaved person was included – the large majority of which were in Sinhala-language articles. The underlying causes of self-harm or suicide were attributed to a single factor in 52% of all the 68 articles; only one English-language newspaper article described a multifactorial cause of death. Where a motive was reported, the main explanations were a family fight, a failed love affair, financial difficulties, and/or diseases.

Crisis (2019), 40(1), 54–61

This study reveals that the majority of Sri Lankan printed newspaper articles reporting suicides and self-harm in a 2-month period during 2014–2015 were noncompliant with guidelines for sensitive reporting. Recommendations especially likely to be violated comprised: reporting of detailed information about the individual and method used and lack of information about help-seeking or the complexities of suicide. Conversely, recommendations likely to be followed included: not publishing an article about self-harm or suicide on the front page and not mentioning a suicide hotspot or earlier suicides. Although the Sinhala and English articles scored differently on selected items in the PRINTQUAL tool, they were equally noncompliant with guidelines.

Visual and Written Reporting of Personal Characteristics and Methods Suicide reporting is known to be explicit in some Asian countries (Beautrais et al., 2008). We found frequent presentation of personal characteristics and methods involved in cases of self-harm or suicide. This is concerning and might provide vulnerable individuals with models to imitate. In 26% of newspaper articles a photograph of the deceased or the location was shown – primarily in Sinhala-language articles. By comparison, a study carried out in India found that 19% of 341 newspaper articles included photographs (Chandra, Doraiswamy, Padmanabh, & Philip, 2014), whereas a Chinese study found a much higher use of photographs (58% of 2,279 articles; Fu, Chan, & Yip, 2011). Although English newspapers more commonly made use of drawings instead of photographs, they were still explicit and clearly illustrated the method used in the episode of self-harm or suicide. Including the method of suicide in the headline is discouraged to avoid promoting and perpetuating methods of suicide (Samaritans, 2008), but this was done in 53% of the articles in this study – especially those in Sinhala – a proportion that is higher than in the study from India (23%; Chandra et al., 2014), but considerably lower than what was found in the study from China (81%; Fu et al., 2011). Furthermore, 7% of all articles included technical details of the method. This is in contrast to the study from India where 19% of articles gave step-by-step guides to the method used (Chandra et al., 2014) as well as the study on Sri Lankan media coverage explained in the Suicide Sensitive Journalism Handbook from 2003, where all 84 articles clearly specified in-depth details of the methods used (Deshapriya et al., 2003). © 2018 Hogrefe Publishing

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

Ingestion of pesticides and hanging are the most common methods of suicide in Sri Lanka (Knipe et al., 2014), but while hanging was the means most frequently mentioned in the included newspaper articles, pesticide ingestion rarely featured. Others have noted overreporting of suicides involving more dramatic methods (Pirkis, Burgess, Blood, & Francis, 2007) and it can be argued that hanging is visually quite dramatic. Although pesticide ingestion might also appear dramatic, the pervasive use of it in the Sri Lankan context might have normalized its occurrence.

that Sri Lankan individuals who previously self-harmed explained a range of seemingly unresolvable daily life stressors, appearing as accumulating, deeper causes of the self-harm (Sørensen et al., 2017). Sri Lankan newspaper outlets could play a role in dispelling myths and highlighting the series of events and personal circumstances behind suicide and self-harm, while respecting the privacy of the deceased and bereaved.

Language Use

Guidelines for sensitive reporting of suicide do exist in Sri Lanka (Deshapriya et al., 2003); however, they could be refined, for instance, by including a list of appropriate Sinhala and Tamil phrases to use for sensitive reporting. Further training of Sri Lankan media professionals would be relevant, bearing in mind the specific gaps for Sinhala- and English-language newspaper reporting. The Suicide Sensitive Journalism Handbook includes a report of a training activity (Deshapriya et al., 2003), but whether this routinely occurs was not possible to ascertain. Media professionals should be closely involved in a process of refining guidelines and training of journalists, since adherence has been found to be closely linked to the degree of involvement and ownership (Bohanna & Wang, 2012). Additionally, schools of journalism should further emphasize responsible media reporting on suicide. Qualitative research examining awareness, use, and perception of the guidelines by Sri Lankan media professionals would also be useful (Bohanna & Wang, 2012; Cheng, Fu, Caine, & Yip, 2014).

We decided to exclude the PRINTQUAL items pertaining to appropriate language use for the Sinhala newspaper articles included in this study, because we were concerned that the translation could not fully capture the necessary nuances. However, the English newspaper articles included in this study did use phrases that should be avoided (58%). In particular, the term to commit suicide was frequently used. Only one English article included recommended language as based on guidelines (Deshapriya et al., 2003; Samaritans, 2008).

Sources of Help and Information About Suicide A common recommendation of guidelines is to provide information about options for help-seeking at the end of an article covering suicide (WHO, 2008). This was not included in any of the articles in the current study, which is similar to the findings from a study carried out in Bangalore, India, where only 1% of articles included information about help-seeking (Chandra et al., 2014). Although the level of basic mental health support services is limited in Sri Lanka, especially in rural areas, a number of nongovernmental-driven suicide hotlines do exist – for example, Sumithrayo (Sri Lanka Sumithrayo, 2013) and CCCline (CCC Foundation, 2009). Sri Lankan newspapers could play an important role in creating awareness about the existence of these services and other relevant institutions offering help to suicidal individuals. Guidelines further recommend that newspaper outlets educate the reader about the complexities of self-harm and suicide (WHO, 2008) – an item that was included in only one article. By contrast, articles typically reported mono-causality behind the self-harm or suicide (52%). Self-harm in Sri Lanka has been described as a sudden, impulsive action (Pearson et al., 2014), which also explains why no suicide note was mentioned in any of the newspaper articles. At the same time, it has also been found © 2018 Hogrefe Publishing

Policy Considerations

Limitations The study has several limitations including the brief monitoring period that resulted in a relatively low number of articles for further analysis. Furthermore, the months selected for the media analysis were leading up to the presidential election – an event that was covered in depth leaving less room for other types of news. The item of whether a celebrity suicide was reported was not relevant for this time period. Hence, under other circumstances there might have been more and other types of articles on selfharm and suicide. It is, however, the authors’ opinion that the period of analysis still provides a relevant data set to shed light on compliance to guidelines in print newspapers in Sri Lanka. The PRINTQUAL tool was developed in a Western setting and has, to the best of our knowledge, not been used in a middle-income country. In general, we found it to be apCrisis (2019), 40(1), 54–61

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

propriate to the Sri Lankan context, although we measured additional items not captured in PRINTQUAL. The item pertaining to a suicide note was not relevant for the articles included in this study. There is no consensus of language use in Sinhala and thus it was not possible to include the items pertaining to appropriate language use in the Sinhala newspaper articles. PRINTQUAL was developed to measure the quality of suicide reporting; however, we found it to also be a useful instrument to capture cases of nonfatal self-harm. A more comprehensive explanation of suicide might have been present in articles excluded from this study. We only included articles from printed newspapers; however, more Sri Lankan news outlets are shifting from print to digital publications. Further, only English- and Sinhala-language newspapers were included – the languages of the majority of the population. Including Tamil newspaper articles in the search strategy would likely have provided an additional understanding of the portrayal of self-harm and suicide in Sri Lanka. Future research could also benefit from an online analysis; social media is extremely popular in Sri Lanka and how suicide is portrayed on, for example, Facebook would be relevant to explore. Although traditional media outlets are losing ground to new online alternatives, regulation of social media can be supported by already existing quality assessment tools, such as PRINTQUAL.

Conclusion This analysis of self-harm and suicide reporting practices found that Sri Lankan newspapers might expose vulnerable readers to harmful influences. Dialogue with and training of Sri Lankan media professionals would be relevant to improve the portrayal of self-harm and suicide in Sri Lankan printed newspapers. We furthermore call for more research on reporting practices in other media outlets. Acknowledgments We thank Mala Ranawake and Surani Neangoda for their advice in the development of this study. Further, we thank the American Foundation for Suicide Prevention for M. W.’s time spent on this study. Funding sources: The University of Copenhagen, Denmark and the Wellcome Trust Safe Storage grant (GR090958) funded this study. The authors have nothing to disclose.

Crisis (2019), 40(1), 54–61

References Beautrais, A., Hendin, H., Yip, P., Takahashi, Y., Chia, B. H., Schmidtke, A., & Pirkis, J. (2008). Improving portrayal of suicide in the media in Asia. In H. Hendin, M. R. Phillips, L. Vijayakumar, J. Pirkis, H. Wang, P. Yip, D. Wasserman, J. M. Bertolote, & A. Fleischmann (Eds.), Suicide and suicide prevention in Asia (pp. 39– 50). Geneva: WHO. Bohanna, I., & Wang, X. (2012). Media guidelines for the responsible reporting of suicide: A review of effectiveness. Crisis, 33(4), 190–198. https://doi.org/10.1027/0227-5910/a000137 CCC Foundation. (2009). CCCline. Retrieved from http://cccfoundation.org.au/cccline/ Chandra, P. S., Doraiswamy, P., Padmanabh, A., & Philip, M. (2014). Do newspaper reports of suicides comply with standard suicide reporting guidelines? A study from Bangalore, India. International Journal of Social Psychiatry, 60(7), 687–694. Chen, Y.-Y., Chen, F., Gunnell, D., & Yip, P. S. F. (2013). The impact of media reporting on the emergence of charcoal burning suicide in Taiwan. PLoS ONE, 8(1), e55000. https://doi.org/10.1371/journal.pone.0055000 Cheng, Q., Fu, K., Caine, E., & Yip, P. S. F. (2014). Why do we report suicides and how can we facilitate suicide prevention efforts? Perspectives of Hong Kong media professionals. Crisis, 35(2), 74–81. https://doi.org/10.1027/0227-5910/a000241 Deshapriya, S., Hattotuwa, S., & Jempson, M. (2003). Suicide sensitive journalism handbook. Colombo, Sri Lanka: Centre for Policy Alternatives & PressWise Trust (UK). Eddleston, M., Sheriff, M. H., & Hawton, K. (1998). Deliberate self harm in Sri Lanka: An overlooked tragedy in the developing world. British Medical Journal, 317(7151), 133. Fu, K.-W., Chan, Y.-Y., & Yip, P. S. F. (2011). Newspaper reporting of suicides in Hong Kong, Taiwan and Guangzhou: Compliance with WHO media guidelines and epidemiological comparisons. Journal of Epidemiology & Community Health, 65(10), 928–933. https://doi.org/10.1136/jech.2009.105650 John, A., Hawton, K., Gunnell, D., Lloyd, K., Scourfield, J., Jones, P. A., … Dennis, M. S. (2017). Newspaper reporting on a cluster of suicides in the UK: A study of article characteristics using PRINTQUAL. Crisis, 38(1), 17–25. https://doi.org/10.1027/02275910/a000410 John, A., Hawton, K., Lloyd, K., Luce, A., Platt, S., Scourfield, J., … Dennis, M. S. (2014). PRINTQUAL – a measure for assessing the quality of newspaper reporting of suicide. Crisis, 35(6), 431–435. https://doi.org/10.1027/0227-5910/a000276 Knipe, D. W., Chang, S.-S., Dawson, A., Eddleston, M., Konradsen, F., Metcalfe, C., & Gunnell, D. (2017). Suicide prevention through means restriction: Impact of the 2008-2011 pesticide restrictions on suicide in Sri Lanka. PLOS ONE, 12(3), e0172893. https://doi.org/10.1371/journal.pone.0172893 Knipe, D. W., Gunnell, D., Pieris, R., Priyadarshana, C., Weerasinghe, M., Pearson, M., … Metcalfe, C. (2017). Is socioeconomic position associated with risk of attempted suicide in rural Sri Lanka? A cross-sectional study of 165 000 individuals. BMJ Open, 7(3), e014006. Knipe, D. W., Metcalfe, C., Fernando, R., Pearson, M., Konradsen, F., Eddleston, M., & Gunnell, D. (2014). Suicide in Sri Lanka 19752012: Age, period and cohort analysis of police and hospital data. BMC Public Health, 14(1), 839. Knipe, D. W., Metcalfe, C., & Gunnell, D. (2015). WHO suicide statistics – a cautionary tale. Ceylon Medical Journal, 60(1), 35. https://doi.org/10.4038/cmj.v60i1.7464 Niederkrotenthaler, T., Voracek, M., Herberth, A., Till, B., Strauss, M., Etzersdorfer, E., … Sonneck, G. (2010). Role of media reports in completed and prevented suicide: Werther v. Papageno ef© 2018 Hogrefe Publishing

J. B. Sørensen et al., Self-Ha rm and Suicide Coverage in Sri Lankan Newspapers

fects. The British Journal of Psychiatry, 197(3), 234–243. https:// doi.org/10.1192/bjp.bp.109.074633 Pearson, M., Zwi, A. B., Rouse, A. K., Fernando, R., Buckley, N. A., & McDuie-Ra, D. (2014). Taking stock – what is known about suicide in Sri Lanka: A systematic review of diverse literature. Crisis, 1(1), 1–12. https://doi.org/10.1027/0227-5910/a000244 Pirkis, J., Blood, R. W., Beautrais, A., Burgess, P., & Skehan, J. (2006). Media guidelines on the reporting of suicide. Crisis, 27(2), 82– 87. https://doi.org/10.1027/0227-5910.27.2.82 Pirkis, J., Burgess, P., Blood, R. W., & Francis, C. (2007). The newsworthiness of suicide. Suicide and Life-Threatening Behavior, 37, 278–283. https://doi.org/10.1521/suli.2007.37.3.278 Pirkis, J., Mok, K., Robinson, J., & Nordentoft, M. (2016). Media Influences on suicidal thoughts and behaviors. In R. O’Connor & J. Pirkis (Eds.), The international handbook of suicide prevention (pp. 743–757). Chichester, UK: John Wiley & Sons. Press Reference. (2017). Press reference Sri Lanka. Retrieved from http://www.pressreference.com/Sa-Sw/Sri-Lanka.html Samaritans. (2008). Media guidelines for reporting suicide and selfharm. Retrieved from http://www.samaritans.org/sites/default/ files/kcfinder/files/press/Samaritans%20Media%20Guidelines%202013%20UK.pdf Sri Lanka Sumithrayo. (2013). Sri Lanka Sumithrayo. Retrieved from http://www.srilankasumithrayo.org/talk-to-us Sørensen, J., Agampodi, T., Sørensen, B. R., Siribaddana, S., Konradsen, F., & Rheinländer, T. (2017). ‘We lost because of his drunkenness’ – the social processes linking alcohol use to selfharm in the context of daily life stress in marriages and intimate relationships in rural Sri Lanka. BMJ Global Health, 2(4), e000462. https://doi.org/10.1136/bmjgh-2017-000462 The Editors’ Guild of Sri Lanka. (2014). The Editors’ Guild of Sri Lanka – code of professional practice. Retrieved from http://www. pccsl.lk/sites/default/files/Code-English.pdf World Health Organization. (2000). Preventing suicide: A resource for media professionals. Geneva, Switzerland: Author. World Health Organization. (2008). Preventing suicide: A resource for media professionals. Geneva, Switzerland: Author. World Health Organization. (2014). Preventing suicide, a global imperative. Retrieved from http://apps.who.int/iris/bitstre am/10665/131056/1/9789241564779_eng.pdf?ua=1 World Health Organization. (2017). Preventing suicide: A resource for media professionals. Geneva, Switzerland: Author.

Received August 2, 2017 Revision received January 17, 2018 Accepted January 19, 2018 Published online July 20, 2018 Jane Brandt Sørensen is a PhD fellow at the Department of Public Health, University of Copenhagen, Denmark. Her main fields of interests lie within self-harm and its association with alcohol use and the media. She has carried out extensive field work in Sri Lanka on these topics. Dr. Melissa Pearson, PhD, has carried out extensive research on selfharm and suicide, especially in improving evidence available for policymaking in Sri Lanka and other lower- and middle-income countries in the region. Martin Wolf Andersen has carried out research through the Department of Public Health, University of Copenhagen, Denmark. He is currently associated with the Danish Refugee Council. Manjula Weerasinghe is a PhD fellow at the Faculty of Medicine and Allied Sciences at Rajarata University of Sri Lanka. He has carried out extensive research on the role of private pesticide vendors in preventing access to pesticides for self-poisoning in rural Sri Lanka. Manjula Rathnaweera works as a school teacher at Thalawa Maha Vidyalaya in Sri Lanka. Chathumini Rathnapala is a student at the Department of English, University of Colombo, Sri Lanka. Michael Eddleston is Professor of Clinical Toxicology at the University of Edinburgh, UK. He has carried out extensive research to reduce deaths from pesticides and self-poisoning in rural Asia, including Sri Lanka. Prof. Flemming Konradsen has more than 20 years of experience working in global health. He has been involved in research on policies and strategies aimed at reducing acute pesticide poisoning in South Asia, specifically Sri Lanka. Jane Brandt Sørensen Øster Farimagsgade 5 1353 Copenhagen Denmark janebs@sund.ku.dk

Crisis (2019), 40(1), 54–61

Short Report

Is Suicidal Behavior in Mood Disorders Altered by Comorbid PTSD? René M. Lento1,2,3, Amanda Carson-Wong1,2, Jonathan D. Green1,4, Christopher G. AhnAllen1,2,5, and Phillip M. Kleespies1,4 VA Boston Healthcare System, Boston, MA, USA Harvard Medical School, Boston, MA, USA

1 2

Massachusetts General Hospital, Boston, MA, USA Boston University School of Medicine, Boston, MA, USA 5 Brigham and Women’s Faulkner Hospital, Boston, MA USA 3 4

Abstract. Background: Suicide is a leading cause of death among US veterans. Associations between depression, posttraumatic stress disorder (PTSD), and suicidal behaviors have been found in this population, yet minimal research has explored how manifestations of self-injurious behavior (SIB) may vary among different diagnostic presentations. Aims: This study aimed to identify clinically useful differences in SIB among veterans who experience comorbid mood disorder and PTSD (CMP) compared with those who experience a mood disorder alone (MDA). Method: Participants were 57 US military veterans who reported an incident of intentional SIB. The semistructured Post Self-Injury/Attempted Self-Injury Debriefing Interview was used to examine characteristics of the SIB. Results: Veterans diagnosed with CMP were more likely than those with MDA to (a) report that the SIB was impulsive and (b) to be under the influence of substances at the time of self-injury. Limitations: Generalizability may be limited by small sample size and predominantly European American, male demographics. While highly relevant to routine clinical practice, caution is recommended, as study diagnoses were attained from medical records rather than structured interviews. Conclusion: Safety planning that emphasizes protection against impulsive suicide attempts (e.g., means restriction) may be especially important among veterans with comorbid mood disorder and PTSD. Keywords: veteran, suicide, mood disorder, PTSD, comorbid PTSD and mood disorder

Recent estimates indicate that an average of 20 US veterans die by suicide each day (US Department of Veterans Affairs, 2016). Numerous risk factors for self-injurious behavior have been identified, including psychosocial stressors (e.g., stressful life events, social withdrawal), negative internal states (e.g., hopelessness, shame), and substance use (see Franklin et al., 2017 for a full review). Additionally, a strong association between clinical depression and suicidal ideation has been established in this population (Pfeiffer et al., 2014); however, depressive symptoms alone do not adequately distinguish which individuals will transition from ideation to suicidal behavior (Klonsky & May, 2014). Accordingly, the contributing role of comorbid diagnoses has gained attention (Busch, Fawcett & Jacobs, 2003; Capron et al., 2012; Conner et al., 2014). Nock, Hwang, Sampson, and Kessler (2010) proposed that while factors associated with depression (e.g., hopelessness) may increase the propensity to experience suicidal ideation, the added effects of disorders characterized by anxiety/agitation, such as posttraumatic stress disorder Crisis (2019), 40(1), 62–66 https://doi.org/10.1027/0227-5910/a000532

(PTSD), may contribute to the impetus for escape often associated with suicidal acts. To date, studies of comorbid depression and PTSD have produced equivocal findings. Veterans diagnosed with PTSD demonstrate high levels of suicidal ideation and behaviors (Oquendo et al., 2005; Wisco et al., 2014). Several studies also support a higher suicide rate among veterans with only PTSD diagnoses compared with the general population (e.g., Bullman & Kang, 1994; Conner, et al., 2013). Yet, in numerous recent studies, PTSD has been associated with lower risk for suicide deaths in VA health-care users after accounting for comorbid diagnoses including depression (Britton et al., 2017; Conner et al., 2014; Shen, Cunha, & Williams, 2016; Zivin et al., 2007). Proposed explanations for this finding include greater availability of treatment and financial resources for veterans with PTSD (Zivin et al. 2007), as well as methodological considerations (Gradus, 2017). On one hand, studies that did not account for psychiatric comorbidity may have inflated estimates for the PTSD–suicide association. On the other © 2018 Hogrefe Publishing

R. M. Lento et al., Suicidal Behavior Mood Disorder PTSD

hand, attempts to statistically adjust for comorbid depression may have obscured part of the impact of PTSD, leading to an underestimate of the PTSD–suicide association. Age may also play a moderating role, as younger depressed veterans with PTSD have been found to have a higher rate of suicide than older veterans with similar presentations (Zivin et al., 2007). Still, research has yet to fully investigate the specific, characteristic differences in how self-injurious behaviors (SIB) may manifest among individuals with different diagnostic presentations. That is, while some combinations of diagnoses might be additive in terms of SIB risk, other combinations may diminish risk. Consequently, the present study aimed to identify how veterans diagnosed with comorbid mood disorder and PTSD might differ from those having a mood disorder alone in terms of charac-

teristics that effect degree of SIB risk (e.g., level of intent, planning).

Method Participants Participants were US military veterans enrolled in VA health care in a large, urban, academic medical center between January 1, 2006 and June 30, 2008 (see Kleespies et al., 2011). Participants included all patients who reported an incident of SIB to VA clinicians. SIB was defined as any intentional, self-inflicted harm or any attempted, intentional, physical harm in which destruction of body tis-

Table 1. Demographic characteristics of self-injury, M (SD) or %

Gender (% male)

Full sample

CMP groupa

MDA groupb

(n = 57)

(n = 23)

(n = 34)

84.2

73.9

91.2

49.5 (13.3)

46.4 (12.1)

51.6 (13.9)

91.2

95.7

88.2

African American (%)

7.0

4.3

8.8

Hispanic American (%)

1.8

2.9

Age (years) Race and ethnicity European American (%)

Marital status Never married (%)

19.6

9.1

26.5

Married (%)

21.4

31.8

14.7

Divorced (%)

44.6

45.5

44.1

Separated (%)

10.7

13.6

8.8

Widowed (%) Combat exposure (%)

3.6

5.9

29.8

39.1

23.5

96.5

100

94.1

3.6

5.8

Patient status at time of SIB Outpatient (%) Residential (VA/non-VA) (%) Method of self-injury Overdose (%)

57.9

43.5

67.6

Cutting (%)

21.1

26.1

17.6

8.8

8.6

8.8

Hanging/strangulation (%) Jump (%)

1.8

2.9

Asphyxiation (%)

1.8

4.3

Poison (%) Other harm (%)

3.5

4.3

2.9

10.5

17.4

5.9

Substance use diagnoses (%)

68.4

73.9

64.7

Alcohol use disorder (%)

45.6

52.2

41.2

Cocaine use disorder (%)

8.8

13.0

5.9

Cannabis use disorder (%)

8.8

4.3

11.8

Note. aComorbid mood and PTSD group. bMood alone group. SIB = self-injurious behavior.

Crisis (2019), 40(1), 62–66

R. M. Lento et al., Suicidal Behavior Mood Disorder PTSD

sue or drug overdose would have been likely to occur if not for immediate interference or intervention. SIBs may be classified as either suicidal (i.e., suicidal intent is present) or nonsuicidal (i.e., no intent; Nock & Favazza, 2009). The present study examined a subsample of 57 participants who reported at least some intent to die when engaging in SIB. Sample characteristics are presented in Table 1.

is the sample size of the first group, and s2 is the sample size of the second group. Both veteran rating and clinician rating of the veteran’s intent to die were dichotomized into high intent (i.e., rating of 4 or above, indicating that intent to die was greater than desire to live) or low intent (i.e., rating of 3 or below, indicating that intent to die was equal to or lower than desire to live). A t test was used to examine whether participant age varied across the two groups.

Measures The Post Self-Injury/Attempted Self-Injury Debriefing Interview is a clinician-administered, semistructured interview developed by a study author (P.K.), which assesses veteran and clinician ratings of veteran’s intent to die (7-point Likert scales ranging from 6 = definite intent to 0 = no intent), and other characteristics of the SIB. Intercorrelations among the scales assessing intent to die and perceived lethality have been found to be significant (range r = .56–.76; see Kleespies et al., 2011). Supplemental clinical and demographic factors at the time of the SIB were collected from participants’ VA electronic medical records (see Table 1).

Procedure Full procedures are documented in the publication by Kleespies and colleagues (2011). This project was initially supported as a quality improvement study, with all study procedures later receiving approval by the VA medical center’s Institutional Review Board as archival research. During the study period, all VA clinicians were required to conduct the semistructured clinical interview upon learning of a patient’s SIB, using a uniform template available in the VA electronic medical record. Approximately 83% of the interviews were completed within 1 week of the SIB. Diagnoses at time of SIB were attained through patients’ hospital discharge summaries.

Analyses Tests of comparative error (CE) were used to investigate differences between two groups: participants diagnosed with (a) comorbid mood and PTSD disorders (CMP) and (b) mood disorders alone (MDA). CE was calculated using the following equation: CE = 1.96x √(r1(100 – r1) ÷ s1)+(r2(100 – r2) ÷ s2) In this equation, r1 is the percentage response for the first group, r2 is percentage response for the second group, s1 Crisis (2019), 40(1), 62–66

Results Of the 57 participants, 23 (40.4%) were in the CMP group, 34 (59.6%) were in the MDA group. Unipolar depression was the most frequent mood diagnosis (77.2%). No significant differences were found among the groups in regard to age, t(55) = 1.48, p = .143, or gender, χ2 (1, N = 57) = 3.08, p = .080. A greater proportion of participants identified as European American than as Hispanic Americans in the CMP group. The majority of participants in both diagnostic groups had at least one co-occurring substance use disorder (see Table 1). There were significant differences between groups on perceived impulsivity and on use of substances at the time of SIB. A greater proportion of participants who reported their self-injury as being impulsive (as opposed to planned) were in the CMP group (CE = 22.64, p < .050). Additionally, a greater proportion of those who reported being under the influence of substances at the time of self-injury were in the CMP group (CE = 24.50, p < .050). No significant differences were found between the groups regarding reason for self-injury, method(s) of SIB, hope/wish to be saved, veteran or clinician rating of veteran’s intent to die, duration of thinking about SIB, or perceived lethality of the SIB.

Discussion Results support the hypothesis that SIB manifests differently among veterans with comorbid mood-and-PTSD presentations. Although both groups had comparable levels of intent to die, similar methods of self-injury, and similar prevalence rates of substance use disorders, veterans with CMP were more likely to be under the influence of substances at the time of the SIB and were more likely to perceive their SIB as impulsive. One possible explanation for these findings is that hyperarousal and avoidance symptoms characteristic of PTSD led to more reactive responses to stressors, resulting in concurrent substance use and less serious preplanning © 2018 Hogrefe Publishing

R. M. Lento et al., Suicidal Behavior Mood Disorder PTSD

of SIB. This logic is consistent with Nock and colleagues’ (2010) suggestion that added anxiety/agitation may contribute to suicidal behavior through a heightened impetus to escape. It is also possible that the combination of mood and PTSD symptoms led to more cumulative emotional or psychological pain, leading the individual to suicidal action (Shneidman, 1987). Finally, less serious preplanning might account for Zivin and colleagues’ (2007) finding that comorbid PTSD–depression groups died by suicide less frequently than depression-only groups. Prospective, longitudinal studies that compare veterans with varying diagnostic presentations on measures of substance use, impulsivity, and suicidal behavior are needed to further explore these hypotheses. Results also bring into question the stepwise fashion in which providers are often trained to conduct suicide risk screens (i.e., assess for ideation, then planning, then intent). The present study suggests that more erratic manifestations of SIB (i.e., concurrent substance use and perceived impulsiveness) were found in the CMP group. Moreover, patients with suicidal ideation do not always disclose ideation to providers during routine screening (Anestis & Green, 2015; Fawcett & Rosenblate, 2000; Freedenthal, 2008), thus making it difficult to gauge when a more extensive risk assessment is truly indicated. Clinicians may be better served by approaching safety planning (Stanley & Brown, 2012) with CMP patients by emphasizing protection against impulsive or unplanned action, restricting access to lethal means, and assessing for substance use. The importance of assessing for substance use, including marijuana, is highlighted by a prior meta-analysis by Serafini and colleagues (2012). In this study marijuana use was associated with an increased risk for SIB in individuals with additional risk factors, including mood disorder, stressful life events, interpersonal problems, and feelings of hopelessness. Accordingly, these assessment and safety planning practices may be advisable for CMP patients even in instances when active suicidal ideation or planning are denied.

Limitations and Future Directions The small number of participants, restricted demographic composition, and cross-sectional nature are important to note and may limit generalizability of this current study. Future research conducted with larger, more diverse samples, across different time-points will be important for increasing generalizability of the findings and investigating possible mediating or moderating variables. Additionally, diagnoses were attained through patients’ hospital discharge summaries rather than through formal diagnostic © 2018 Hogrefe Publishing

interviews. While embedding this study in routine clinical practice likely enhances the external validity of results, particularly among VA medical centers, it should be noted that these methods present limitations to diagnostic reliability.

Conclusion This study provides preliminary information on associations between diagnostic presentation and manifestation of self-injurious behavior in a population of treatment-seeking veterans. Results suggest that factors associated with comorbid PTSD may contribute to impulsively acting on suicidal thoughts. Results also provide support for the thorough assessment of substance use patterns and access to lethal means among veterans presenting with comorbid mood and PTSD symptoms. Future research is needed to investigate how factors such as intoxication may moderate or mediate impulsive action among suicidal veterans with comorbid mood disorders and PTSD. Acknowledgments Amanda Carson-Wong is now at Durham DBT, Inc. Jonathan D. Green is now at the O’Connor Professional Group.

References Anestis, M. D., & Green, B. A. (2015). The impact of varying levels of confidentiality on disclosure of suicidal thoughts in a sample of United States National Guard personnel. Journal of Clinical Psychology, 71, 1023–1030. https://doi.org/10.1002/jclp.22198 Britton, P. C., Bohnert, K. M., Ilgen, M. A., Kane, C., Stephens, B., & Pigeon, W. R. (2017). Suicide mortality among male veterans discharged from Veterans Health Administration acute psychiatric units from 2005 to 2010. Social Psychiatry and Psychiatric Epidemiology, 52, 1–7. Bullman, T. A., & Kang, H. K. (1994). Posttraumatic stress disorder and the risk of traumatic deaths among Vietnam veterans. The Journal of Nervous and Mental Disease, 182, 604–610. Busch, K., Fawcett, J., & Jacobs, D. (2003). Clinical correlates of inpatient suicide. Journal of Clinical Psychiatry, 64, 14–19. https:// doi.org/10.4088/JCP.v64n0105 Capron, D. W., Fitch, K., Medley, A., Blagg, C., Mallott, M., & Joiner, T. (2012). Role of anxiety sensitivity subfactors in suicidal ideation and suicide attempt history. Depression and Anxiety, 29, 195–201. https://doi.org/10.1002/da.20871 Conner, K. R., Bohnert, A. S., McCarthy, J. F., Valenstein, M., Bossarte, R., Ignacio, R., ... Ilgen, M. A. (2013). Mental disorder comorbidity and suicide among 2.96 million men receiving care in the Veterans Health Administration health system. Journal of Abnormal Psychology, 122, 256–263. Conner, K. R., Bossarte, R. M., He, H., Arora, J., Lu, N., Tu, X. M., & Katz, I. R. (2014). Posttraumatic stress disorder and suicide in 5.9 million individuals receiving care in the Veterans Health AdCrisis (2019), 40(1), 62–66

ministration health system. Journal of Affective Disorders, 166, 1–5. https://doi.org/10.1016/j.jad.2014.04.067 Fawcett, J., & Rosenblate, R. (2000). Suicide within 24 hours after assessment in the emergency department: Look for and manage anxiety. Psychiatric Annals, 30, 228–231. Franklin, J. C., Ribeiro, J. D., Fox, K. R., Bentley, K. H., Kleiman, E. M., Huang, X., … Nock, M. K. (2017). Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research. Psychological Bulletin, 143, 187–232. https://doi.org/ 10.1037/bul0000084 Freedenthal, S. (2008). Assessing the wish to die: A 30-year review of the suicide intent scale. Archives of Suicide Research, 12, 277–298. Gradus, J. L. (2017). PTSD and death from suicide. PTSD Research Quarterly, 28, 1–8. Kleespies, P. M., AhnAllen, C. G., Knight, J. A., Presskreischer, B., Barrs, K. L., Boyd, B. L., & Dennis, J. P. (2011). A study of self-injurious and suicidal behavior in a veteran population. Psychological Services, 8, 236–250. Klonsky, E. D., & May, A. M. (2014). Differentiating suicide attempters from suicide ideators: A critical frontier for suicidology research. Suicide and Life-Threatening Behavior, 44, 1–5. https:// doi.org/10.1111/sltb.12068 Nock, M., & Favazza, A. (2009). Nonsuicidal self-injury: Definition and classification. In M. Nock (Ed.), Understanding nonsuicidal self-injury: Origins, assessment, and treatment. Washington, DC: APA Books. Nock, M. K., Hwang, I., Sampson, N. A., & Kessler, R. C. (2010). Mental disorders, comorbidity and suicidal behavior: Results from the National Comorbidity Survey Replication. Molecular Psychiatry, 15, 868–876. https://doi.org/10.1038/mp.2009.29 Oquendo, M., Brent, D. A., Birmaher, B., Greenhill, L., Kolko, D., Stanley, B., ... Mann, J. J. (2005). Posttraumatic stress disorder comorbid with major depression: Factors mediating the association with suicidal behavior. American Journal of Psychiatry, 162, 560–566. https://doi.org/10.1176/appi.ajp.162.3.560 Pfeiffer, P. N., Brandfon, S., Garcia, E., Duffy, S., Ganoczy, D., Kim, H. M., & Valenstein, M. (2014). Predictors of suicidal ideation among depressed veterans and the interpersonal theory of suicide. Journal of Affective Disorders, 152, 277–281. https://doi. org/10.1016/j.jad.2013.09.025 Serafini, G., Pompili, M., Innamorati, M., Rihmer, Z., Sher, L., & Girardi, P. (2012). Can cannabis increase the suicide risk in psychosis? A critical review. Current Pharmaceutical Design, 18, 5165–5187. Shen, Y. C., Cunha, J. M., & Williams, T. V. (2016). Time-varying associations of suicide with deployments, mental health conditions, and stressful life events among current and former US military personnel: A retrospective multivariate analysis. The Lancet Psychiatry, 3, 1039–1048. https://doi.org/10.1016/S2215-0366(16) 30304-2 Shneidman, E. S. (1987). A psychological approach to suicide. In G. R. Van den Bos & B. K. Bryant (Eds.), Master lectures series. Cataclysms, crises, and catastrophes: Psychology in action (pp. 147–183). Washington, DC: American Psychological Association. Stanley, B., & Brown, G. K. (2012). Safety planning intervention: A brief intervention to mitigate suicide risk. Cognitive and Behavioral Practice, 19(2), 256–264. U.S. Department of Veterans Affairs. (2016, July 7). VA conducts nation’s largest analysis of veteran suicide [Press release]. Re-

Crisis (2019), 40(1), 62–66

R. M. Lento et al., Suicidal Behavior Mood Disorder PTSD

trieved from http://www.va.gov/opa/pressrel/pressrelease.cfm?id=2801 Wisco, B. E., Marx, B. P., Wolf, E. J., Miller, M. W., Southwick, S. M., & Pietrzak, R. H. (2014). Posttraumatic stress disorder in the US veteran population: Results from the National Health and Resilience in Veterans Study. The Journal of Clinical Psychiatry, 75, 1338–1346. https://doi.org/10.4088/JCP.14m09328 Zivin, K., Kim, H. M., McCarthy, J. F., Austin, K. L., Hoggatt, K. J., Walters, H., & Valenstein, M. (2007). Suicide mortality among individuals receiving treatment for depression in the Veterans Affairs health system: Associations with patient and treatment setting characteristics. American Journal of Public Health, 97, 2193–2198. https://doi.org/10.2105/AJPH.2007.115477

Received September 14, 2017 Revision received January 5, 2018 Accepted January 15, 2018 Published online July 20, 2018 Dr. René Lento is a postdoctoral fellow at the Massachusetts General Hospital and Red Sox Home Base Program. She completed her PhD in Clinical Psychology at the Catholic University of America and her clinical internship at the VA Boston Healthcare System, MA. Her research interests include suicide prevention and intervention, and treatment engagement. Amanda Carson-Wong, PhD, completed her undergraduate studies at Vanderbilt University in Neuroscience and went on to obtain master’s and doctorate degrees in Clinical Psychology from Rutgers University. She completed her psychology internship and postdoctoral training at the VA Boston Healthcare System and is currently in private practice in Durham, NC. Jonathan Green, PhD, is the Chief of Clinical Operations at the O’Connor Professional Group, a behavioral health consulting firm. He also conducts research in the National Center for PTSD at VA Boston Healthcare System. Dr. Green received his PhD from Clark University and completed his clinical internship and postdoctoral fellowship at the VA Boston. Dr. Christopher AhnAllen is Director of Inpatient Psychology and Psychology Education at Brigham and Women’s Faulkner Hospital, MA. He holds an Assistant Professor of Psychiatry academic appointment at Harvard Medical School and a research appointment at VA Boston Healthcare System. Dr. Phillip Kleespies is a Clinical Psychologist at the VA Boston Healthcare System, MA. He has an appointment as Assistant Clinical Professor in the Department of Psychiatry, Boston University School of Medicine. He is also a Fellow of Divisions 12 and 18 of the American Psychological Association.

Phillip M. Kleespies VA Boston Healthcare System (116 B) 150 South Huntington Ave Boston, MA 02130 USA phillip.kleespies@va.gov

News, Announcements, and IASP John Terry Maltsberger – Suicidologist Extraordinaire Mark J. Goldblatt1, Elsa Ronningstam1, Benjamin Herbstman1, and Mark Schechter2 Department of Psychiatry at McLean Hospital, Harvard Medical School, Boston, MA, USA North Shore Mental Health Center, Harvard Medical School, Salem, MA, USA

1 2

John Terry Maltsberger (1933–2016) grew up in Cotulla, Texas, and attended college at Princeton and Harvard Medical School before doing his Psychiatric Residency at Massachusetts Mental Health Center in the early 1960s. He spent his adult life in Boston but contributed to suicide studies the world over. Terry stood at the intersection of psychoanalysis and suicide studies, which were central to his core interest. He valued relationships and held on to them throughout his life. He talked about things he learnt in residency as if they happened yesterday. He spoke with an immediacy of the people he admired and their ideas that had obviously stayed with him for many years and continued to influence him. He also spoke of the many relationships he had in suicide-related organizations, the journals where he was a reviewer, and the congresses where he presented his work during his long and illustrious career. In this paper, we want to describe some of the highlights of his contributions to suicide studies by noting his influential publications that significantly affected the field.

When Strong Negative Emotions Arise in Psychotherapy: Countertransference Challenges Terry’s first major publication was with Dan Buie in 1974 and dealt with the countertransference challenge of strong negative emotions that may arise in the treatment of suicidal patients. Maltsberger and Buie (1974) questioned how therapists may acknowledge and deal with unpleasant feelings that can arise in the therapist during the course of treatment of suicidal patients. They recognized that it is only human to feel helpless, frustrated, and even antagonistic when working with suicidal patients, and they wanted to bring this emotional response into the arena of academic discussion. At that time, it was considered © 2018 Hogrefe Publishing

problematic and shameful if therapists had anything but caring and loving feelings toward their patients. These authors recognized that coming to grips with these difficult feelings is essential for helping the patient. Disavowing feelings of frustration, and helplessness, or pretending that they do not exist, is dangerous, because it gives these feelings more power, with the likelihood that they will somehow leak out and be experienced by the patient as a manifestation of rage or emotional withdrawal, which can be suicide inviting. They particularly noted that even subtle emotional withdrawal on the part of the therapist could be experienced by the patient as an abandonment, which could increase the risk for suicide. Since that time, and to a large extent based on this article, discussion has opened up and it is now common to deal with these feelings in supervision and in consultation. Maltsberger and Buie went on to make several more important contributions to suicide studies, including: “The Devices of Suicide” (Maltsberger & Buie, 1980); “The Psychotherapist as an Accomplice in Suicide” (Maltsberger & Buie, 1994); and The Practical Formulation of Suicide Risk (Buie & Maltsberger, 1983).

Suicide Risk Assessment Terry offered a radical departure from the checklist of risk factors for evaluating an individual’s risk of suicide (Maltsberger, 1988). He saw disturbances of the self-representation as the key underlying commonality in suicide. He urged clinicians to recognize the individual in pain sitting with them in the room, and to listen at the right frequency for signs of breakdown of the self. He described how to listen for the disturbances and subtle signs of a self-representation’s fragmentation and outlined five components to evaluate the risk for suicide. First, assess a patient’s past responses to stress, especially losses, as these will likely predict future responses. Second, Crisis (2019), 40(1), 67–71 https://doi.org/10.1027/0227-5910/a000547

News, Announcements, and IASP

assess a patient’s vulnerability to three life-threatening affects – aloneness, self-contempt, and murderous rage. Third, determine the nature and availability of exterior sustaining resources. Fourth, explore the emergence and emotional importance of death fantasies – what does the patient imagine happens when he or she dies? And finally, fifth, assess the patient’s capacity for reality testing – signs of psychosis. All of these components indicate the patient’s response to stressors and the risk for suicidal action.

emotional experience, and at once contains and makes use of his or her countertransference experience. Out of this collaboration came journal articles (e.g., Michel et al., 2002), the book edited by Michel and Jobes (Maltsberger, 2011) as well as the website, and a conference that occurs every 2 years. The direct descendant of the Aeschi conferences in Switzerland now takes place in Vail, Colorado, and highlights the patient-centered approach that Terry strongly advocated.

American Foundation for Suicide Prevention

The Descent Into Suicide

In 1990, Terry joined the Board of Directors of the American Foundation for Suicide Prevention (AFSP) and became a member of their Scientific Council until 2006. He was co-director with Herb Hendin of the AFSP Suicide Data Bank Project. Together with others in this group, they studied in depth therapist survivors who had lost a patient to suicide. They produced at least eight papers that identified important aspects of these treatments. These papers dealt with: the therapists’ reactions to patients’ suicides (Hendin, Lipschitz, Maltsberger, Haas, & Wynecoop, 2000); recognizing and responding to a suicide crisis (Hendin, Maltsberger, Lipschitz, Haas, & Kyle, 2001); precipitating events in suicide (Maltsberger, Hendin, Haas, & Lipschitz, 2003); desperation and other affective states (Hendin, Maltsberger, Haas, Szanto, & Rabinowicz, 2004); factors contributing to therapists’ distress after the suicide of a patient (Hendin, Haas, Maltsberger, Szanto, & Rabinowicz, 2004); problems in psychotherapy with suicidal patients (Hendin, Haas, Maltsberger, Koestner, & Szanto, 2006); and the role of intense affective states in signaling a suicide crisis (Hendin, Maltsberger, & Szanto, 2007). These papers provided documented research in suicide studies and added significantly to evidence-based treatments for suicidal patients.

The Aeschi Approach Terry was one of the original group that met in Aeschi, Switzerland (in 2000), and described the Aeschi approach to suicidal patients. The central characteristics of this approach are respectful supportive listening to the patient’s narrative by the interviewer, who is empathic and promotes the patient’s active participation in telling their story. The interviewer fosters the establishment of a therapeutic alliance, pays particular attention to the patient’s subjective Crisis (2019), 40(1), 67–71

In this major publication (Maltsberger, 2004), Terry describes in vivid language the suicidal state as experienced by the rapidly deteriorating patient. Terry uses the literary works of Edgar Allan Poe (The Descent Into the Maelstrom; Poe, 1978/1842) to describe the catastrophic effects of being caught up in the hopelessness of suicidal currents. He identifies four stages associated with the break-up of the self. These are (1) affective deluge, (2) efforts to master the affective flooding, (3) loss of control and disintegration, and (4) grandiose schemes for self-preservation through jettison of the body. In a suicidal crisis, the self breaks apart. Intense desperation and other agonizing affects flood the individual. In the first stage, this torrent of anguish, or psychache, builds and becomes unendurable and indicates imminent suicidal collapse. And then down we came with a sweep, a slide and a plunge, that made me feel sick and dizzy, as I was falling from some lofty mountain-top in a dream. But while we were up I had thrown a quick glance around – and that one glance was all sufficient… The Moskoe-strom whirlpool was about a quarter of a mile dead ahead… I involuntarily closed my eyes in horror. (p. 588)

In the second stage the patient tries desperately to avoid spiraling down further. He/she may use mechanisms to cope with this impending crisis. Some turn to others for help. Others dissociate from their emotional distress, either through internally driven defenses or via external contributions from the use of drugs and alcohol. Those who make the decision to take their own life appear to experience a sense of calm and mastery as they go about their deadly plans. Some deliberately conceal their desperation because they do not want anyone to interfere in their suicidal strategies. It may appear strange, but now, when we were in the very jaws of the gulf, I felt more composed than when we were approaching it. Having made up my mind to hope no more, I got rid of a great deal of that terror which unmanned me at first… I began © 2018 Hogrefe Publishing

News, Announcements, and IASP

to reflect how magnificent a thing it was to die in such a manner, and how foolish it was in me to think of so paltry a consideration as my own individual life, in view of so wonderful a manifestation of God’s power. (p. 589)

In the third stage the patient experiences a sense of falling apart accompanied by intense horror and fear, becoming overwhelmed and giving way to despair. We careered round and round… flying rather than floating, getting gradually more and more into the middle of the surge and then nearer and nearer to its horrible inner edge…round and round we swept- not in any uniform movement- but in dizzying swings and jerks. (p. 591)

Finally, in the fourth stage, patients feel trapped in unmanageable pain, and experience themselves as disintegrating. Suicidal patients then resort to magical fantasies to save themselves from psychic annihilation. They commonly imagine they can split off their mental and physical selves by killing their bodies. In psychoanalytic terms, the ego functioning fails massively resulting in narcissistic collapse, loss of reality testing, self-fragmentation, and ego failure.

Suicidal Fantasy as a Life-Sustaining Recourse Terry recognized that sometimes people think about suicide and do not go on to kill themselves (Maltsberger, Ronningstam, Weinberg, Schechter, & Goldblatt, 2010). Often, thoughts of suicide precede action, but at other times day-dreaming about suicide can inhibit deadly action. He engaged the Boston Suicide Study Group (see next section) in a project to analyze how some people benefit from their suicidal fantasies and how for others it is a rehearsal before action. He suggested that it was essential to make a formal suicide risk assessment rather than rely on suicide ideation as a predictor of suicidal action in individual patients. Sustaining fantasies do not indicate that the patient is falling apart. At times, suicide fantasies may be used to mitigate hopelessness, provide self-soothing, relieve distress, and enhance self-cohesion, all of which contribute to making suicide less necessary. Suicidal ideation prior to a suicide attempt is marked by a suicide crisis (Hendin et al., 2001), in which there is a crescendo of psychic pain, a precipitating event, and often behavioral indicators of a crisis, usually accompanied by increased drug and alcohol use. However, occasionally, a vulnerable patient who has used suicide fantasies for their sustaining support may deteriorate into a suicide crisis and use the thoughts for actual planning for suicide action. © 2018 Hogrefe Publishing

Interestingly, literary authors have long known of this distinction. Nietzsche (1907, Chapter IV, # 157) said: “It is always consoling to think of suicide: in that way, one gets through many a bad night.” A Walker Percy character (Percy, 1961) says: They all think I’m going to commit suicide. What a joke. The truth of course is the exact opposite: suicide is the only thing that keeps me alive. Whenever everything else fails, all I have to do is consider suicide and in two seconds I’m as cheerful as a nitwit. But if I could not kill myself – ah then, I would. I can do without Nembutal or murder mysteries but not without Suicide. (pp. 194–195)

Boston Suicide Study Group In 1993, Terry founded the McLean Suicide Study Group, which continued in 2009 as the Boston Suicide Study Group. Terry led this group in a groundbreaking study that evaluated five manualized treatments – dialectical behavioral therapy (DBT), mentalization-based treatment (MBT), transference-focused psychotherapy (TFP), schema-focused therapy (SFT), and cognitive behavioral therapy (CBT) – that claimed to be effective in the treatment of suicidal patients with borderline personality disorder (Weinberg et al., 2010). The study has been lauded by many in the suicide research community. This study found that there were several points in agreement among the manualized therapies. They all had a treatment framework, or an agreed-upon strategy to manage suicidal crisis; they paid attention to affect; they endorsed an active therapist; they encouraged exploratory interventions, and suggested change-oriented interventions for the suicidal patient. These modalities also had significant differences. They differed on the use of multimodal and team collaborative treatments; on whether or not suicidality should be an explicit target of treatment; on how they focused on the treatment relationship; on the use of interpretation; on the use of supportive interventions; and on providing support for the therapist. Terry also led the Boston Suicide Study Group in a study of trauma and its effects on the suicidal patient (Maltsberger, Goldblatt, Ronningstam, Weinberg, & Schechter, 2011). Suicide attempts are considered to be traumatizing and affect patients just as if they had been subject to attempted murder. Repeated affective traumatization erodes the capacity for hope and the ability to make sustaining relationships.

Crisis (2019), 40(1), 67–71

News, Announcements, and IASP

Suicide Studies Terry lamented the current state of affairs in suicide studies and, by extension, what is valued in contemporary suicide risk assessment (Maltsberger, Schechter, Herbstman, Ronningstam, & Goldblatt, 2015). The emphasis on counting and statistical analysis of larger bodies of data continues to increase. While 50 years ago there was a fuller emphasis given to individual idiosyncrasies of patients, this is no longer true. Detail, not overall gestalt, rules the day. The era of checklists, questionnaires, and tables is full upon us. (p. 388)

He quoted W. B. Cameron (1963), “not everything that counts can be counted, and not everything that can be counted counts.” Counting is of course indispensable, but sometimes it can get in the way of deeper and more productive thinking. The human condition cannot be reduced to a series of risk factors and correlations; the drive to empiricism, as helpful as it is on the one hand, risks drowning out other ways of understanding people on the other… We need to free ourselves from the constriction of general, homogenizing diagnosis. We need more reports that reflect the deeper experiences of our patients, including more qualitative research. (p. 389)

Conclusion Terry Maltsberger’s contribution to the study of suicide and the treatment of self-destructive patients is incalculable. His influence on our lives and many of our careers is similarly inestimable. He holds a special place in our hearts and minds. His work will sustain us through many a dark night.

References Buie, D. H., & Maltsberger, J. T. (1983). The practical formulation of suicide risk. Cambridge, MA: The Firefly Press. Cameron, W. B. (1963). Informal sociology, a casual introduction to sociological thinking. New York, NY: Random House. Hendin, H., Haas, A. P., Maltsberger, J. T., Koestner, B., & Szanto, K. (2006). Problems in psychotherapy with suicidal patients. American Journal of Psychiatry, 163, 67–72. https://doi.org/10.1176/ appi.ajp.163.1.67 Hendin, H., Haas, A. P., Maltsberger, J. T., Szanto, K., & Rabinowicz, B. S. (2004). Factors contributing to therapists’ distress after the suicide of a patient. American Journal of Psychiatry, 161, 1442– 1446. https://doi.org/10.1176/appi.ajp.161.8.1442 Hendin, H., Lipschitz, A., Maltsberger, J. T., Haas, A. P., & Wynecoop, S. (2000). Therapists’ reactions to patients’ suicides. American Journal of Psychiatry, 157(12), 2022–2027. https://doi. org/10.1176/appi.ajp.157.12.2022 Crisis (2019), 40(1), 67–71

Hendin, H., Maltsberger, J. T., Haas, A. P., Szanto, K., & Rabinowicz, B. S. (2004). Desperation and other affective states in suicidal patients. Suicide and Life-Threatening Behavior, 34, 386–394. https://doi.org/10.1521/suli.34.4.386.53734 Hendin, H., Maltsberger, J. T., Lipschitz, A., Haas, A. P., & Kyle, J. (2001). Recognizing and responding to a suicide crisis. Suicide Life-Threatening Behavior, 31(2), 115–128. Hendin, H., Maltsberger, J. T., & Szanto, K. (2007). The role of intense affective states in signaling a suicide crisis. Journal of Nervous and Mental Disease, 195, 363–368. https://doi.org/10.1097/ NMD.0b013e318052264d Maltsberger, J. T. (1988). Suicide danger: clinical estimation and decision. Suicide and Life-Threatening Behavior, 18(1), 47–54. Maltsberger, J. T. (2004). The descent into suicide. International Journal of Psychoanalysis, 85(Pt. 3), 653–667. https://doi. org/10.1516/002075704774200799 Maltsberger, J. T. (2011). Empathy and the historical context, or how we learned to listen to patients. In K. Michel & D. A. Jobes (Eds.), Building a therapeutic alliance with the suicidal patient (pp. 93–107). Washington, DC: American Psychological Association. Retrieved from http://www.aeschiconference.unibe.ch Maltsberger, J. T., & Buie, D. H. (1974). Countertransference hate in the treatment of suicidal patients. Archives of General Psychiatry, 30(5), 625–633. Maltsberger, J. T., & Buie, D. H. (1980). The devices of suicide. International Review of Psychoanalysis, 7, 61–72. Maltsberger, J. T., & Buie, D. H. (1994). The psychotherapist as an accomplice in suicide. Italian Journal of Suicidology, 4, 75–81. Maltsberger, J. T., Goldblatt, M. J., Ronningstam, E., Weinberg, I., & Schechter, M. (2011). Traumatic subjective experiences invite suicide. The Journal of the American Academy of Psychoanalysis and Dynamic Psychiatry, 39, 671–694. https://doi.org/10.1521/ jaap.2011.39.4.671 Maltsberger, J. T., Hendin, H., Haas, A. P., & Lipschitz, A. (2003). Determination of precipitating events in the suicide of psychiatric patients. Suicide Life-Threatening Behavior, 33(2), 111–119. Maltsberger, J. T., Ronningstam, E., Weinberg, I., Schechter, M., & Goldblatt, M. J. (2010). Suicidal fantasy as a life sustaining recourse. Journal of the American Academy of Psychoanalysis and Dynamic Psychiatry, 38(4), 611–624. https://doi.org/10.1521/ jaap.2010.38.4.611 Maltsberger, J. T., Schechter, M., Herbstman, B., Ronningstam, E., & Goldblatt, M. J. (2015). Suicide studies today: Where do we come from? What are we? Where are we going? Crisis, 36, 387–389. https://doi.org/10.1027/0227-5910/a000338 Michel, K., Maltsberger, J. T., Jobes, D. A., Leenaars, A. A., Orbach, I., Stadler, K., … Valach, L. (2002). Discovering the truth in attempted suicide. American Journal of Psychotherapy, 56(3), 424–437. Nietzsche, F. (1907). Beyond good and evil. New York, NY: The Macmillan Company. Percy, W. (1961). The moviegoer. New York, NY: Alfred A. Knopf. Poe, E. A. (1978). The descent into the maelstrom. In T. O. Mabbott (Ed.), The collected works of Edgar, Allan Poe, Vol. 2: Tales & sketches I (pp. 574–597). Cambridge, MA: Belknap Press of Harvard University Press. (Original work published 1841) Weinberg, I., Ronningstam, E., Goldblatt, M. J., Schechter, M., Wheelis, J., & Maltsberger, J. T. (2010). Strategies in treatment of suicidality: Identification of common and treatment-specific interventions in empirically supported treatment manuals. Journal of Clinical Psychiatry, 71(6), 699–706. https://doi. org/10.4088/JCP.08m04840blu

News, Announcements, and IASP

Accepted March 14, 2018 Published online October 12, 2018 Mark J. Goldblatt, MD, is Associate Clinical Professor of Psychiatry at Harvard Medical School, Clinical Associate at McLean Hospital, faculty member of Boston Psychoanalytic Society and Institute and member of the Boston Suicide Study Group. His long-standing interest in suicide studies and psychoanalysis is reflected in his publications and teaching. Mark Schechter, MD, is Chair of Psychiatry at North Shore Medical Center, and part-time instructor in Psychiatry at Harvard Medical School. He is a member of the Boston Psychoanalytic Society and Institute, and a member of the Boston Suicide Study Group. He teaches and writes about aspects of psychotherapy with suicidal patients. Benjamin Herbstman, MD, MHS, is a psychiatrist in private practice in Cambridge, MA, and an advanced candidate at the Boston Psychoanalytic Society and Institute. He is an assistant psychiatrist at McLean

ÂŠ 2018 Hogrefe Publishing

Hospital, a part-time lecturer in Psychiatry at Harvard Medical School, and a member of the Boston Suicide Study Group. Elsa Ronningstam, PhD, is Associate Professor of Psychology at Harvard Medical School and Clinical Psychologist at McLean Hospital. She is a faculty member of the Boston Psychoanalytic Society and Institute and a member of the Boston Suicide Study Group. She has published and presented extensively on studies and treatment of suicide and personality disorders.

Dr. Mark J. Goldblatt Associate Professor of Psychiatry Harvard Medical School Clinical Associate, McLean Hospital 1105 Massachusetts Avenue, Suite 2C Cambridge, MA 02138 USA mark_goldblatt@hms.harvard.edu

Crisis (2019), 40(1), 67â&#x20AC;&#x201C;71

News, Announcements, and IASP

Announcements The 52nd Annual Conference of the American Association of Suicidology (AAS) is taking place April 24–27, 2019, in Denver, CO, USA. For more information go to http://www.suicidology.org/Annual-Conference/51stAnnual-Conference The 31st Annual Healing After Suicide Loss Conference of the American Association of Suicidology (AAS), entitled “Converging Fields, Expanding Perspectives,” is taking place on April 27, 2019, in Denver, CO. For more information go to http://www.suicidology.org/ Annual-Conference/31st-Annual-Healing-Conference

The 2019 IASR/AFSP International Summit on Suicide Research is taking place October 27–30, 2019, in Miami, FL, USA. For more information go to http://suicide researchsummit.org/ The 3rd Annual Conference on Suicide by Kevin’s Song, entitled “Confronting a Public Health Crisis,” is taking place November 8–10, 2018, in Plymouth, MI. For more information go to https://kevinssong.org/wp-content/ uploads/2018/08/KS-Brochure-FINAL-8_5x11.pdf

The 30th IASP World Congress, entitled “Breaking Down Walls and Building Bridges,” is taking place September 17–21, 2019, in Derry-Londonderry, Northern Ireland. For more information go to https://www.iasp2019. com/

Crisis (2019), 40(1), 72–73 https://doi.org/10.1027/0227-5910/a000586

News, Announcements, and IASP

International Association for Suicide Prevention Would you like to join IASP?

−− to carry out programs of research, especially those which can be pursued through international joint cooperation.

IASP is a nonprofit organization for those interested and working in suicide prevention. The membership consists of individuals (clinicians, scientists, crisis workers, volunteers and persons who have lost a family member by suicide) and community, national and international organizations. At this moment the membership extends over 50 countries and is affiliated with the World Health Organization as the key organization concerned with suicide prevention.

The activities of the association focus on the importance of sharing information, research and knowledge in order to address the issues of suicide and its prevention. The biennial congress, the bimonthly journal Crisis, biannual newsletter and the website serves as a medium for international exchange and a directory of members involved in suicidology is made available. Candidates for IASP membership may apply directly to the Central Administration Office membership@iasp.info or join IASP online at http://www.iasp.info Membership consists of: −− Regular membership for individuals regularly engaged in suicide studies, clinical management of suicidal patients, and/or suicide prevention activities, and who are interested in furthering the work of the Association. −− Organizational membership is for national or international voluntary or other nonprofit organizations (incorporated or unincorporated) working in the field of suicide studies, clinical management of suicidal patients and/or suicide prevention activities under the terms of their governing document. −− Associate membership is for individuals and organizations who are interested in working toward the goals of the Association, but who are not qualified for regular membership. Please contact the IASP Central Administration Office at membership@iasp.info.; on the web at http://www.iasp.info

The main objectives of IASP are: −− to provide a common platform for all representatives of the many professions who are engaged in the field of suicide prevention and crisis intervention; −− to allow interchange of acquired experience in this area in various countries, especially through the exchange of literature; −− to promote the establishment of national organizations for suicide prevention; −− to facilitate the wider dissemination of the fundamentals of effective suicide prevention to professional groups and to the general public; −− to arrange for specialized training of selected persons in the area of suicide prevention in selected training centers;

IASP membership fees (include a subscription to Crisis) Zone

Crisis

Individual Membership

Organizational Membership

Hard copy

US $190.00 (early bird US $180.50, 2 yrs US $361.00, 3 yrs US $515.00)

US $235.00 (early bird US $225.00, 2 yrs US $446.50, 3 yrs US $635.00)

Hard copy

Online only US $162.00 (early bird US $153.00, 2 yrs US $308.00, 3 yrs US $438.00) US $200.00 (early bird US $190.00, 2 yrs US $380.00, 3 yrs US $543.00) US $160.00 (early bird US $152.00, 2 yrs US $304.00, 3 yrs US $430.00) US $180.00 (early bird US $170.00, 2 yrs US $342.00, 3 yrs US $485.00)

Online only US $136.00 (early bird US $129.00, 2 yrs US $258.50, 3 yrs US $368.00) US $153.00 (early bird US $145.00, 2 yrs US $290.50, 3 yrs US $414.00) 3

Hard copy

US $135.00 (early bird US $128.00, 2 yrs US $256.50, 3 yrs US $365.00) US $160.00 (early bird US $150.00, 2 yrs US $304.00, 3 yrs US $430.00)

Online only US $115.00 (early bird US $109.00, 2 yrs US $218.50, 3 yrs US $310.00) US $136.00 (early bird US $129.00, 2 yrs US $258.50, 3 yrs US $368.00) 4

Hard copy

US $115.00 (early bird US $109.00, 2 yrs US $218.50, 3 yrs US $310.00) US $125.00 (early bird US $120.00, 2 yrs US $237.50, 3 yrs US $340.00)

Online only US $98.00 (early bird US $93.00, 2 yrs US $186.00, 3 yrs US $239.00)

US $106.00 (early bird US $101.00, 2 yrs US $201.50, 3 yrs US $288.00)

All zones: Online only US $115.00 (early bird US $109.00, 2 yrs US $218.50, 3 yrs US $310.00) For Advanced Organization Membership fees go to students, http://www.iasp.info/application.php volunteers, & retirees

IASP Executive Committee 2017–2019 President: Vice President 1: Vice President 2: Vice President 3:

Dr. Murad Khan Prof. Jane Pirkis Prof. Rory O’Connor Prof. Ping Qin

General Secretary: Dr. Daniel Reidenberg Treasurer: Dr. Thomas Niederkrotenthaler Representative of the College of Presidents: Prof. Ella Arensman

IASP Council of National Representatives 2017–2019 Co-Chairs: Prof. Madelyn Gould and Prof. Jie Zhang For a current listing of the IASP Council of National Representatives please refer to https://www.iasp.info/council_of_national_represent atives.php © 2019 Hogrefe Publishing

Crisis (2019), 40(1), 72–73

Instructions to Authors – Crisis: The Journal of Crisis Intervention and Suicide Prevention Crisis – The Journal of Crisis Intervention and Suicide Prevention is an international periodical that publishes original articles on suicidology and crisis intervention. Papers presenting basic research as well as practical experience in the field are welcome. Crisis also publishes potentially life-saving information for all those involved in crisis intervention and suicide prevention, making it important reading for clinicians, counselors, hotlines, and crisis intervention centers. Crisis: The Journal of Crisis Intervention and Suicide Prevention publishes the following types of articles: Research Trends, Short Reports, and Clinical Insights. Manuscript Submission: All manuscripts should be submitted electronically at http://www.editorialmanager.com/cri Detailed instructions to authors are provided at http://www. hogrefe.com/j/cri Copyright Agreement: By submitting an article, the author confirms and guarantees on behalf of him-/herself and any coauthors that he or she holds all copyright in and titles to the submitted contribution, including any figures, photographs, line drawings, plans, maps, sketches and tables, and that the article and its contents do not infringe in any way on the rights of third parties. The author indemnifies and holds harmless the publisher from any third-party claims. The author agrees, upon acceptance of the article for publication, to transfer to the publisher on behalf of him-/herself and any coauthors the exclusive right to reproduce and distribute the article and its contents, both physically and in nonphysical, electronic, and other form, in the journal to which it has been submitted and in other independent publications, with no limits on the number of copies or on the form or the extent of the distribution. These rights are transferred for the duration of copyright as defined by international law. Further-

Crisis (2019), 40(1)

more, the author transfers to the publisher the following exclusive rights to the article and its contents: 1. The rights to produce advance copies, reprints, or offprints of the article, in full or in part, to undertake or allow translations into other languages, to distribute other forms or modified versions of the article, and to produce and distribute summaries or abstracts. 2. The rights to microfilm and microfiche editions or similar, to the use of the article and its contents in videotext, teletext, and similar systems, to recordings or reproduction using other media, digital or analog, including electronic, magnetic, and optical media, and in multimedia form, as well as for public broadcasting in radio, television, or other forms of broadcast. 3. The rights to store the article and its content in machine-readable or electronic form on all media (such as computer disks, compact disks, magnetic tape), to store the article and its contents in online databases belonging to the publisher or third parties for viewing or downloading by third parties, and to present or reproduce the article or its contents on visual display screens, monitors, and similar devices, either directly or via data transmission. 4. The rights to reproduce and distribute the article and its contents by all other means, including photomechanical and similar processes (such as photocopying or facsimile), and as part of so-called document delivery services. 5. The right to transfer any or all rights mentioned in this agreement, as well as rights retained by the relevant copyright clearing centers, including royalty rights to third parties. Online Rights for Journal Articles: Guidelines on authors’ rights to archive electronic versions of their manuscripts online are given in the document “Guidelines on sharing and use of articles in Hogrefe journals” on the journal’s web page at www.hogrefe. com/j/cri. February 2016

A unique and comprehensive handbook presenting the state-of-the-art in suicide bereavement support

Karl Andriessen / Karolina Krysinska / Onja T. Grad (Editors)

Postvention in Action

The International Handbook of Suicide Bereavement Support 2017, xviii + 424 pp. US $94.00 / â&#x201A;Ź 74.95 ISBN 978-0-88937-493-5 Suicide is not merely the act of an individual; it always has an effect on others and can even increase the risk of suicide in the bereaved. The International Association for Suicide Prevention, the World Health Organisation, and others have recognized postvention as an important strategy for suicide prevention. This unique and comprehensive handbook, authored by nearly 100 international experts, including researchers, clinicians, support group facilitators, and survivors, presents the state-of-theart in suicide bereavement support. The first part examines the key concepts and the processes that the bereaved experience and illustrates them with illuminating clinical vignettes. The second and third parts

www.hogrefe.com

look in detail at suicide bereavement support in all the most relevant settings (including general practices, the workplace, online and many others) as well as in specific groups (such as health care workers). In the concluding section, the support provided for those bereaved by suicide in no less than 23 countries is explored in detail, showing that postvention is becoming worldwide strategy for suicide prevention. These chapters provide useful lessons and inspiration for extending and improving postvention in new and existing areas. This unique handbook is thus essential reading for anyone involved in suicide prevention or postvention research and practice.

Suicide and Culture Diana van Bergen / Amanda Heredia Montesinos / Meryam Schouler-Ocak (Editors)

Suicidal Behavior of Immigrants and Ethnic Minorities in Europe 2015, viii + 190 pp., hardcover US $54.00 / € 38.95 ISBN 978-0-88937-453-9 Also available as eBook

Nearly half of the inhabitants of several large European cities, such as London, Berlin, or Amsterdam, and a rising proportion of many countries’ overall population, are immigrants or from an ethnic minority. However, this fact has been understudied in research and prevention of suicidal behavior. This volume addresses this gap. Leading experts describe rates

and risk factors of suicidal behavior among immigrants and ethnic minorities, looking at high-risk groups such as female immigrants and refugees, as well as examining the role of cultural factors. They also show how epidemiology, theory, and other research findings can be translated into solid prevention and treatment programs.

Erminia Colucci / David Lester (Editors) with Heidi Hjelmeland / B. C. Ben Park

Suicide and Culture Understanding the Context 2013, xiv + 270 pp. US $49.00 / € 34.95 ISBN 978-0-88937-436-2 Also available as eBook

The increasing domination of biological approaches in suicide research and prevention, at the expense of social and cultural understanding, is severely harming our ability to stop people dying – so run the clearly set out arguments and evidence in this lucid book by leading social scientists and

www.hogrefe.com

suicide researchers. In the concluding section, the editors highlight both the necessity and the challenges of conducting good culturally sensitive studies, as well as suggesting solutions to these challenges. This volume is thus essential reading for anyone involved in suicide research and prevention.

Hogrefe OpenMind Open Access Publishing? It’s Your Choice! Your Road to Open Access Authors of papers accepted for publication in any Hogrefe journal can now choose to have their paper published as an open access article as part of the Hogrefe OpenMind program. This means that anyone, anywhere in the world will – without charge – be able to read, search, link, send, and use the article for noncommercial purposes, in accordance with the internationally recognized Creative Commons licensing standards.

The Choice Is Yours 1. Open Access Publication: The final “version of record” of the article is published online with full open access. It is freely available online to anyone in electronic form. (It will also be published in the print version of the journal.) 2. Traditional Publishing Model: Your article is published in the traditional manner, available worldwide to journal subscribers online and in print and to anyone by “pay per view.” Whichever you choose, your article will be peer-reviewed, professionally produced, and published both in print and in electronic versions of the journal. Every article will be given a DOI and registered with CrossRef.

www.hogrefe.com

How Does Hogrefe’s Open Access Program Work? After submission to the journal, your article will undergo exactly the same steps, no matter which publishing option you choose: peer-review, copy-editing, typesetting, data preparation, online reference linking, printing, hosting, and archiving. In the traditional publishing model, the publication process (including all the services that ensure the scientific and formal quality of your paper) is financed via subscriptions to the journal. Open access publication, by contrast, is financed by means of a one-time article fee (€ 2,500 or US $3,000) payable by you the author, or by your research institute or funding body. Once the article has been accepted for publication, it’s your choice – open access publication or the traditional model. We have an open mind!

@iasp2019

30th World Congress of the International Association for Suicide Prevention Theme “Breaking Down Walls and Building Bridges” Derry-Londonderry, Northern Ireland

17-21 September, 2019 Key Themes & Topics at IASP 2019 Include: • • • • • • •

Innovation in suicide prevention and treatment Suicide in context Suicide, trauma and conflict Suicide in the ageing population Suicide and self harm in young people Clinicians as survivors of suicide Psychological, psychosocial and pharmacological treatments of suicide risk • The role of “big data” in suicide prevention • Biological and genetic risk factors for suicide • Stigma, mental health and suicide • Postvention and lived experience • Psychological and social risk factors for suicidal behaviour • Public health approaches to suicide prevention • Symposium Submission deadline: 31st January 2019 • Abstract Submission deadline: 17th March 2019 • Early Bird Registration deadline: 31st March 2019

Visit www.iasp2019.com for more information.