Background
Suicide is a significant social and public health problem, yet it is largely preventable [
1]. Every year there are more than 800,000 suicides around the world [
1,
2]. By the year 2020, suicide is estimated to contribute more than 2 % to the global burden of disease (1). South East Asia and the Western Pacific regions account for more than half of the suicides that occur globally every year [
3]. The world’s most vulnerable populations, including the young, the elderly and the socially isolated, are at particular risk of suicide. A high burden of disease is attributable to suicide in low- and middle-income countries (LMICs), which are often ill-equipped to meet the mental health needs of their populations, with services that are inadequately developed, often difficult to access and of poor quality [
1].
Among the many suicide prevention strategies that have been developed some have shown excellent results. However, prevention efforts are more limited in LMICs than in high-income countries despite the fact that most suicides (75 %) occur in LMICs [
4,
5]. Studies evaluating prevention efforts in LMICs are few. Most studies of prevention programs have been conducted in high-income countries and their relevance to LMICs is unknown (5). Although evidence on which suicide prevention strategies are most effective is limited, a few promising approaches have been identified. In a 2005 study, a panel of 15 suicide experts identified five major areas of prevention: education and awareness programs for the general public and professionals, screening methods for persons at high risk, treatment of psychiatric disorders, restricting access to lethal means, and changing media reporting of suicide [
6]. Available evidence suggests that physician education, restriction of means and gatekeeper education are promising interventions [
6].
Sri Lanka has one of the highest suicide rates in the world, 25 per 100,000 in 2005 [
7]. These rates, although still high, are a substantial improvement from 1995 where the suicide rate peaked at 47 per 100,000 [
7]. The reduction that has been achieved is reflective of the fact that Sri Lanka is one of the few LMICs that have formulated a specific national suicide prevention plan and taken action on suicide prevention [
4]. Among the actions implemented was the decriminalisation of suicide in 1998 [
7]. The decline in rates closely followed the banning of highly toxic pesticides in the country [
7]. Most of Sri Lanka’s suicide prevention strategies have been focussed on restricting access to lethal means, such as pesticides, as acute poisoning is the most common method of suicide in the country [
7,
8]. While withdrawing or restricting a number of pesticides, and initiating plans for safe use of pesticides through programs such as integrated pest management [
9], have been effective [
7,
10,
11], further areas of suicide prevention/intervention must be developed, as the current suicide rate in the country is still high. Suicide prevention interventions such as education and awareness programs for the general public and professionals could be beneficial in Sri Lanka, especially programs referred to as gatekeeper training [
12].
Gatekeeper training involves the training of people in the community to recognise and identify those who are at risk of suicide and assist them in receiving appropriate care [
13,
14]. In essence, gatekeepers ‘open the gate to help’ for those at risk of suicide [
13]. The best suited to act as gatekeepers are family members and friends due to their close relationship with the person at risk [
13]. However, gatekeepers may also include professionals who are in frequent contact with potentially vulnerable populations, such as public health officers (public health inspectors and midwives in Sri Lanka), religious leaders, first responders, persons employed in schools, prisons and military, and caregivers [
6].
Several gatekeeper training methods are available as train-the-trainer programs [
13]. Examples of such programs include the applied suicide intervention skills training (ASIST) by Living Works; Question, Persuade and Respond (QPR) and Yellow Ribbon International (YRI) for suicide prevention [
13,
15]. Although these programs are being implemented in many countries, there is a lack of conclusive evidence of effectiveness [
16]. Most of the studies and evaluations that have been conducted have been with school staff, in workplaces and the military [
6,
13,
15]. Studies of the effectiveness of gatekeeper training have also been conducted with indigenous and Aboriginal communities [
13,
14]. Uncontrolled evaluations of ASIST and QPR have found a positive impact on self-reported preparedness and observer-rated skills in suicide intervention [
13]. A randomised controlled trial among school staff in the United States showed an increased self-reported knowledge of suicide risk but had little impact on gatekeeper behaviour during follow up [
13]. A similar outcome was seen in a study done in a remote First Nations community [
13]. However, studies done on gatekeeper training in the Norwegian Army and the US Air force have reported success in lowering suicide rates [
6]. Further research is necessary to evaluate effectiveness of gatekeeper training in preventing suicide. Nonetheless, there is enough evidence to support a conclusion that gatekeeper training is effective in changing attitudes and behaviours, with increased knowledge about suicide and reduction in negative attitudes towards people who are at risk of suicide [
13,
15]. Mental health first aid (MHFA) training courses, which have been delivered in all states in Australia and 23 other countries, have shown similar results [
17,
18]. Evaluations of these programs have consistently shown that MHFA training is associated with improved knowledge of mental illnesses and their treatment, knowledge of appropriate first aid strategies, and confidence in providing first aid to individuals with mental illness [
19], that these benefits are sustained over time [
20] and that it is possible to reach very large numbers of the general population [
21]. Some studies have also shown improved mental health in those who attended the training, decrease in stigmatising attitudes and increase in the amount and types of support provided to others [
20,
22]. Such change in knowledge, attitudes, confidence and behaviour could prove beneficial in communities where mental illness and suicide are stigmatised, for example among communities in Sri Lanka [
23].
The availability of guidelines that would assist members of the public in how they should provide first aid to a person at risk of suicide could form the basis for the development and evaluation of gatekeeper training programs or other mental health first aid training courses. An obstacle to the development of such guidelines is the limited systematic evidence to guide the content of suicide gatekeeper training. In such circumstances consensus methods such as the Delphi process can be useful to develop appropriate guidelines [
22]. This method embodies the principle of practice-based evidence [
22]. Results relevant to the local population and culture can be attained by harnessing the expertise of professionals or people working in the area [
22]. This method is relatively inexpensive and, as Delphi group members do not need to meet, the study can be done using the internet [
24‐
26]. The Delphi method has been widely used in health research, [
27,
28] and has been used to develop suicide first aid guidelines for English-speaking countries [
29,
30]. This method has also been used to develop suicide first aid guidelines for Asian countries such as India, Japan and the Philippines and for people from immigrant and refugee background [
31‐
34]. There were similarities and differences among the guidelines produced for each of the Asian countries, therefore this project was undertaken to develop country-specific suicide first aid guidelines for Sri Lanka.
The aim of the project was to produce guidelines for members of the public providing first aid for a person in Sri Lanka who is having suicidal thoughts or displaying suicidal behaviours. No study of this kind has been conducted previously in Sri Lanka. In this study, the role of a mental health first aider was defined as “someone who helps a person who is developing a mental health problem or is in a mental health crisis”.
Methods
The study had three phases: a literature search, questionnaire development and the Delphi process.
Systematic search for possible suicide first aid actions in the literature
As part of a project to revise previous work [
29] done to develop suicide first aid guidelines for developed English-speaking countries a systematic search of relevant literature was carried out by the MHFA team [
29,
30]. Briefly, this search identified information about how to determine whether someone is having thoughts of suicide and what possible first aid actions can be undertaken. The items developed by Ross et al. [
29,
30] were used as the basis for development of the first round survey used in this project.
Questionnaire development
The questionnaire for English-speaking countries [
30] was developed from analysing the literature mentioned above and creating statements that suggest a potential first aid action (e.g. what the first aider should do or should not do) or statements that suggest what a first aider should know. These statements were grouped into common categories: identification of suicide risk, assessing seriousness of risk, initial assistance, talking with the suicidal person, no-suicide contracts, ensuring safety, passing time during the crisis, what the first aider should know, confidentiality and adolescent-specific statements [
30]. A working group was convened to ensure that the questionnaire did not include items that might be difficult to understand, repetitive or ambiguous.
The questionnaire developed for English-speaking countries had 436 items, each describing a potential action by a first aider that could be presented to the panel for rating [
30]. This initial questionnaire was modified for the current study with items generated in recent studies on suicide prevention for people from immigrant and refugee background and in Asian countries, in order to improve its cultural relevance [
31,
33,
34]. The initial questionnaire for Sri Lanka thus contained 473 first aid action items, plus six questions on participants’ socio-demographics and experience/training. Open-ended questions were included after each category, to give participants the opportunity to suggest culturally specific actions that could be used to generate further items for subsequent rounds of the Delphi process. An open-ended question was also included to gather gender-specific items regarding suicide in Sri Lanka. The questionnaire was translated to both Sinhala and Tamil, the two most common languages spoken in Sri Lanka.
To qualify as an ‘expert’, participants were required to have knowledge about suicide through their experience as a mental health professional or as a suicide prevention lived experience advocate. Participants were recruited in several ways. Clinical experts in medicine, psychiatry, public health, social work and psychology working in Sri Lanka were identified by the authors and invited to participate. Professionals who had published on suicide in Sri Lanka were also invited to be part of the panel. Attempts were made to recruit people with lived experience (i.e. people who have experienced suicidal ideation or made a suicide attempt in the past) through advocacy organisations in the country, but this proved to be unsuccessful. The main difficulty conveyed by the advocacy organisations were concerns about confidentiality and the stigma attached to suicide in Sri Lanka.
Invitation letters were sent using the web-based SurveyMonkey [
35], and further information about the study was provided through a web page link, which contained the plain language statement (PLS) in English, Sinhala and Tamil. The invitation letter also asked the participants to recommend or nominate colleagues who could be eligible as panel members. The email also contained the link for the first round of the English version of the questionnaire and participants were advised to contact the researchers if they preferred to receive the Sinhala or Tamil versions of the questionnaire, or printed copies of the survey. No participants opted for these alternatives.
In Round 1, the questionnaire derived from the process described above was sent to 77 potential panel members. It provided instructions and definitions to guide the panel members. Panel members were asked to rate each statement by indicating their level of agreement that the statements should be included in guidelines for a first aider who is helping a suicidal person. The 5-point Likert scale was: essential, important, do not know/depends, unimportant and should not be included.
Participants were asked to indicate whether they considered any item in the first round questionnaire to be culturally irrelevant or unacceptable, or not feasible because of limitations in the health system and resources in the country. Furthermore, after each category participants were asked to suggest any additional item that was not included in the initial questionnaire. This encouraged culturally and gender specific material to be introduced. Suggestions made by the panel members in response to the open-ended questions were analysed and used to construct new items for the Round 2 questionnaire.
Round 1 responses were analysed to calculate the percentage of the panel who rated an item as either “essential” or “important”. Items that were endorsed by 80 % or more of the panel were accepted for inclusion in the guidelines. Statements rated by 70–79 % of the members of the panel as essential or important were re-rated in the subsequent round of the Delphi process. Statements that were rated by less than 70 % of the panel members as essential or important were rejected. These are the same consensus thresholds for inclusion and re-rating used in previous Delphi studies [
32‐
34].
The Round 2 questionnaire consisted of new items that were generated from the open-ended questions in Round 1 and items from Round 1 to be re-rated. Participants were asked to rate the items in the Round 2 questionnaire. At the end of the round, percentages were analysed and items that reached the 80 % consensus criterion were accepted for the guidelines. Items that were re-rated in Round 2 and were below 80 % consensus were rejected. Items that were generated from open-ended questions in Round 1 and achieved only 70–79 % consensus in Round 2, were included to be re-rated in Round 3. As in Round 2, items that reached consensus were included in the guidelines and the rest were rejected.
Discussion
The aim of this project was to identify consensus among professionals on first aid actions that members of the public can take to help a suicidal person in Sri Lanka. A total of 304 first aid statements achieved consensus among the panel of experts.
In response to the open-ended questions comments were made by panel members regarding the lack of mental health resources in some rural areas in Sri Lanka, especially suicide hotlines. There is also a lack of awareness in the community of the existence of such hotlines where they do exist [
12]. Suicide hotlines are mostly run by non-government organisations such as Sumithrayo [
36], and CCCline, founded in 2009 by the CCC Foundation. A great deal of work has been undertaken to expand such services to rural areas [
4,
8,
12]. Compared to other countries where similar studies have been undertaken [
32‐
34], advocacy and support organizations for suicide in Sri Lanka are limited. These guidelines could be used as an additional training tool by mental health and suicide prevention organisations for training of non-mental health professional staff and volunteers.
As there is evidence in the literature that suicide in Sri Lanka is more common among adolescents, educating potential gatekeepers in schools and universities on how to help a young person at risk could be most useful. A focus group conducted with students about their perspectives on suicidal behaviours identified that teachers could be potential gatekeepers as they have significant and positive contact with young people and most students viewed them as accessible [
12]. Religious leaders could also have an important role in suicide prevention in the country, as religion plays an important part in Sri Lanka society [
23]. Some panel members commented that changes in a person’s religious beliefs, with either an increased or decreased interest in their faith, may indicate increased risk of suicide. Educating religious leaders about warning signs of suicide and how to identify and help a person at risk of suicide could prove helpful. Gatekeeper training programs for teachers [
12] and religious leaders based on these guidelines may prove to be an effective component of a broader suicide prevention strategy in Sri Lanka. While the guidelines constitute a basis for the development of training they may need to be further tailored to the specific religious and cultural factors in different regions in Sri Lanka, as studies have shown a difference in attitudes to and beliefs about suicide, and suicide rates, among different religious and cultural groups [
37,
38].
Domestic violence, alcohol abuse by a partner and relationship issues are important risk factors for suicide among females in Sri Lanka [
39]. Therefore, when developing suicide prevention training programs, these specific issues must also be addressed, within the specific socio- cultural context. For example, gatekeepers must know of services available to assist a person experiencing domestic violence and refer the person to such services. They could also be trained to help women develop safety plans to keep themselves and their children safe during violent episodes [
39] which may provide women with viable alternatives to suicidal behaviour and also support them to take effective action in the context of domestic violence.
Future directions
The guidelines should be translated to Sinhala and Tamil and a visually simplified version for low-literacy population (e.g. an infographic as developed for the guidelines for people from immigrant and refugee background see [
31]) should be developed. The various versions of the guidelines should then be distributed freely among the community and institutional settings to increase awareness about suicide and possible prevention strategies. The guidelines can be used as a basis for the development of training programs and can be implemented for specifically vulnerable groups such as adolescents and women experiencing domestic violence. To ensure that suicide prevention training is relevant to specific communities training programs should be developed in consultation with community members within the area so that ethnic, cultural and religious differences regarding suicide, for example within Sinhala and Tamil communities, are taken into consideration. In areas that are less developed, and that have limited resources, these guidelines could be used also by primary health workers, midwives and nurses. Any training programs developed with the use of these guidelines should be rigorously evaluated to determine impact on suicide prevention.
Authors’ contributions
HM and EC conceived the study. EM, CMK, AFJ modified relevant English questionnaire for the Sri Lanka study. HM and JM identified potential members of the expert panel in Sri Lanka. SDS carried out the data collection under the supervision of EC and HM. SDS and EC were primarily responsible for analysis. SDS wrote the first draft, all authors contributed to subsequent drafts and HM wrote the final draft of the manuscript. All authors read and approved the final manuscript.