'Don't push me aside, Doctor': Suicide attempters talk about their support needs, service delivery and suicide prevention in South Africa.

Few studies have explored the expressed support needs of suicide attempters in developing countries. Data, collected via in-depth interviews with suicide attempters admitted to a South African hospital, were analysed using thematic content analysis. Participants explicitly asked for integrated psycho-social services at a primary health care level and say they require assistance with alleviating psychiatric symptoms, establishing connectedness, interpersonal conflict and solving socio-economic problems. Findings highlight the importance for suicide prevention of (1) considering interpersonal and contextual socio-economic factors in addition to the psychiatric causes of suicidal behaviour; and (2) multilevel strategies, intersectoral collaboration and integrated person-centred primary health care.

Introduction

Suicide prevention remains a challenge in low- and middle-income countries (LMICs), where 75 per cent of the world’s suicides occur and where psychiatric resources are scarce (World Health Organization (WHO), 2014). Integral to suicide prevention is the provision of appropriate care to suicide attempters; nonfatal suicidal behaviour carries a high risk of repetition and is associated with suicide. The dominant biomedical view asserts that suicidal behaviour is a symptom of psychopathology and that the treatment of psychiatric illness is the cornerstone of suicide prevention (Marsh, 2010). This psychiatric view of suicide has been critiqued within the field of critical suicidology for its failure to take account of the contextual, socio-cultural, economic and historic factors which contribute to the aetiology of suicidal behaviour (White, 2015; White et al., 2015). Furthermore, mainstream suicidology has attracted criticism for failing to document the insider perspectives of suicide attempters (Hjelmeland and Knizek, 2010, 2011). Few studies have investigated the expressed support needs of suicide attempters, particularly in LMICs. It is within this context that I set out to document the expressed support needs of a group of patients admitted to an urban hospital in South Africa (SA) following a medically serious suicide attempt. My intention was to understand from the patient’s point of view what might be done on discharge from hospital to reduce the risk of repetition of suicidal behaviour. The findings are discussed within the context of calls from critical suicidologists to focus on the socio-cultural context of suicide, and the challenges of suicide prevention in LMICs, like SA.[1] The findings provide insight into possible suicide prevention strategies and highlight structural problems with the provision of mental health care in low-resource environments.

Literature review

Annually, approximately 800,000 people die by suicide globally (WHO, 2014) and for every completed suicide there are an estimated 20–30 suicide attempts (Wasserman, 2016). A suicide attempt is defined as, ‘A nonhabitual act with nonfatal outcome that the individual, expecting to, or taking the risk to die or to inflict bodily harm, initiated and carried out with the purpose of bringing about wanted changes’ (De Leo et al., 2006: 14). Individuals who attempt suicide are at high risk of repetition and are 20–30 times more likely than the general population to die by suicide (Hawton et al., 2003; Kapur et al., 2005; Owens et al., 2002). Suicide attempters who present at hospitals are a well-delineated group of patients who may be amenable to targeted interventions to reduce their suicide risk (Arensman et al., 2011). It is, however, not always clear what interventions are effective to prevent suicides (Brown and Jager-Hyman, 2014; Hetrick et al., 2016). Critics have noted that the ineffectiveness of suicide prevention strategies is in large part a result of the fact that suicide has predominantly been conceptualised within a biomedical risk-factor paradigm; consequently, much of the research has been quantitative, de-contextual and narrowly focused on a handful of individual variables and psychiatric factors (Hjelmeland and Knizek, 2010, 2011; Kral, 2012; White et al., 2015). Evidence of this is seen in a recently published meta-analysis which concluded that in spite of decades of quantitative risk factor suicide research, experts’ ability to identify patients who will engage in suicidal behaviour is no better than chance and has not improved significantly in the last five decades (Franklin et al., 2016).

Typically, suicide prevention strategies have focused on identifying psychiatric risk factors and promoting access to psychiatric care for at-risk individuals. For example, the four-level approach to suicide prevention, advocated by the European Alliance Against Depression, focuses on educating primary health care professionals and the general population about psychiatric risk factors for suicide, and increasing access to appropriate psychiatric care for at-risk individuals (Székely et al., 2013). A systematic review of psychological autopsy studies concluded that ‘Suicide prevention strategies may be most effective if focused on the treatment of mental disorders’ (Cavanagh et al., 2003: 395).

Critical suicidologists have pointed out the problems of focusing solely on psychiatric risk factors while failing to consider the socio-cultural context within which suicidal behaviours occur (Hjelmeland et al., 2014). These critiques have spurned a number of studies which have attempted to examine more broadly the role of contextual factors in the aetiology of suicide. Examples include the systematic review on the relationship between poverty and suicidal behaviour in LMICs (Iemmi et al., 2016) and a study among perinatal women living in a low-resource environment in SA which found that socio-economic adversity was a better predictor of suicidal behaviour than psychiatric conditions (Onah et al., 2016).

Authors have drawn attention to the role of the socio-cultural context in suicidal behaviour (Chan and Thambu, 2016). There are, for example, studies which have described how suicidal behaviour in different parts of the world is shaped by models of hegemonic masculinity and prescriptive gender roles (Olive et al., 2012), religion (Cook, 2014; Dervic et al., 2004), and attitudes towards suicide and help seeking (Reynders et al., 2014). This renewed interest in socio-cultural context has led some authors to question the role of mental illness in precipitating suicidal behaviour. Hjelmeland et al. (2014) have gone so far as to assert that ‘suicide is about far more than mental disorders, and may be about something quite different’ (p. 1370). There are good reasons to consider a wide range of contextual, socio-cultural and economic factors when planning suicide prevention interventions, particularly in low-resource environments. In many LMICs, psychiatric resources are scarce and consequently suicide prevention efforts that rely primarily on mental health professionals are unlikely to have the requisite reach to be effective.

Cahill (2007) and Cleary (2012) have described the potential benefits of investigating first-person accounts of the lived experiences of suicidal persons. White (2016) has affirmed that ‘those with lived experience of suicidality have invaluable insights to contribute to the current evidence base’ (p. 349). Documenting the voices of suicide attempters may help to highlight the social context in which suicidal behaviour occurs and to draw attention to factors, such as culture, which are not easily measured and included in quantitative studies. Qualitative studies in this field not only hold the promise of providing insight into the suicidal mind but, perhaps more importantly, can tell us what it is like to be a suicidal person negotiating one’s way through the health care system (Bantjes and Swartz, 2017). Furthermore, qualitative studies with suicide attempters may provide insight into the support needs of this group of patients and highlight problems with the organisation of care within the health care system which impede suicide prevention. Although this is a largely unexplored area of research in LMICs, in the last 10 years, there has been a burgeoning of research from high-income western countries which describes the lived experiences of suicide attempters (White, 2016). One of the earliest studies exploring the lived experience of suicidal individuals was undertaken by Rosen (1975), who interviewed seven individuals who had jumped from the Golden Gate Bridge in San Francisco in suicide attempts. On the basis of these qualitative interviews, Rosen (1975) suggested concrete strategies for suicide prevention. This has established a tradition of employing qualitative research with suicide attempters in order to generate evidence for suicide prevention. On the basis of her review of qualitative studies on suicide ideation, attempts, and suicide prevention, White (2016) concluded, ‘Far from providing a definitive statement about the nature and meaning of suicidality, these studies have shown the complex, dynamic, context-dependent, multiple, and contradictory character of suicide ideation and attempts’ (p. 349).

Methodology

Semi-structured interviews were conducted with 80 adult patients (46 females and 34 males) admitted to an urban hospital, following a suicide attempt. Participants were between the ages of 18 and 67 years.

Interviews lasted between 45 and 90 minutes and were conducted by the author, who is a psychologist. Participants were asked to recount how they had come to be in hospital and to describe what they thought they might need to prevent future episodes of suicidal behaviour. Participants were prompted to talk about the socio-cultural context in which their suicide attempt occurred and to express their own understanding of their behaviour. Interviews were conducted in a private space within the hospital prior to patients’ discharge. Interviews were audio-recorded and transcribed. Data were analysed independently by two researchers using thematic content analysis (Braun and Clarke, 2006). An inductive approach (Thomas, 2006) was adopted to code the data using Atlas-ti software. Theoretical sense was made of the findings after a process of triangulation of codes. This data-driven process led to the identification of five superordinate themes which describe patients’ expressed support needs, namely (1) alleviation of psychiatric and somatic symptoms, (2) access to integrated psychological care at a primary health care level, (3) help to establish connectedness and belonging, (4) assistance with interpersonal and family conflict and (5) practical help to solve situational problems. The analysis of data for this article was confined to identify expressed support needs, a more detailed account of the patients’ first-person understanding of the socio-cultural context in which their suicide attempt occurred and the contextual factors that contributed to this behaviour will be presented in subsequent publications.

Ethics

Permission to conduct this study was obtained from the health science research ethics committees at the University of Cape Town and Stellenbosch University. Institutional permission was obtained from the relevant authorities at the hospital and participants gave written informed consent to participate in the study. Pseudonyms have been used to protect the identity of the participants.

Findings

Alleviation of psychiatric and somatic symptoms

Many participants asked for assistance with the alleviation of psychiatric and somatic symptoms. They ascribed their suicide attempts to behavioural, cognitive and emotional symptoms which they described as unbearable. Cleo said, ‘I just need to be back on my medication. That is very important’. Nancy said that she needed professional help to control her feelings and thoughts because ‘… there is no way out of it and I feel like I’m drowning in everything that is happening in my mind’. Similarly, Jaco said,

I just don’t want to feel this way. I will get any help I can to take it (these feelings) off my mind … I don’t know why I can’t get healed of this inner pain. It draws me to suicide every time.

While some patients described symptoms of depressive and anxiety disorders which they said contributed to their attempted suicide, there were other patients who spoke about the role of trauma and loss. These patients asked for help with symptoms such as re-experiencing the trauma, high levels of fear, hyper vigilance and panic. Tim explained how he had been repeatedly sexually abused when he was growing up and how the trauma of this had left him with unbearable memories and high levels of anxiety and mistrust, with which he needed help. Similarly, participants described complicated grief reactions to recent losses. Jane, an 18-year-old school girl, described how her mother had recently died, leaving her alone and without financial or family support to look after her young sister. She said she was overcome with sadness at her mother’s death but had no way to process this grief because she was left unsupported and responsible for her sister.

Some participants described symptoms of psychotic illness, including persecutory and command hallucinations, which they said contributed to their suicidal behaviour. These participants said they required help with ‘the voices’ and ‘the crazy thoughts inside my head’. Vuyo expressed this by saying, ‘… maybe I can put this pain to bed and maybe just stop hearing everything that they (the voices) are saying about me, because I am tired of it’.

Participants also described the role of chronic pain in their suicide attempt and said they would require help with this. David said, ‘I need to be comfortable (free of pain) at least’.

Participants expressed a need for sustained psychiatric care within a supportive environment. Some patients described their previous experience of receiving care, saying that they had felt rushed and pushed aside by mental health service providers. Nina described the kind of care she required saying,

Don’t rush me when I tell you, Doctor, this is what I’ve been through – I was raped or whatever. Don’t push me aside, Doctor, because then I don’t feel open enough to speak to you. Then I just close up.

Access to integrated psychological care at a primary health care level

Participants articulated a need for psychotherapy and emotional support. They said they did not have access to ongoing psychological care in their communities. They requested psycho-social support to be integrated into existing primary health care settings. Ivan explained, ‘… they should actually have community day-hospitals where they can have counsellors to help us if you feel down, besides sisters. We see sisters maybe once a month. But we need to have counsellors on standby for us’. Similarly, Jannie said, ‘… if I can just go into therapy where it can help for at least a few months. The support and the fact of professional help, it will be much easier for me …’

Some participants asked for assistance to develop insight into their emotions, change their behaviour and regulate their feelings. Anita said, ‘But the thing is, how do I change my behaviour from having the same negative feelings go around and around and around in my head? How does that change? How can I stop that from happening?’ Likewise, Bongani said, ‘I’d like to understand what triggers my mood. Like specially when my mood goes down – what triggers it? And to learn methods of better managing when I lose focus and when my mind starts going ballistic’.

Help to establish relationships, connection and belonging

Many participants were emphatic about their need for help to establish connection and emotional closeness to others. They expressed a desire for emotionally supportive relationships. Anne said she needed ‘people to be there for me’ in order to curb future suicidal behaviour. Anton said it would help him if he was ‘not so alone’ and if ‘people actually listen to you and talk to you and pay attention to you’. Anton went on to describe the power of physical and emotional closeness to curb his suicidal behaviour by saying, ‘That’s what I always tell my girlfriend, the only thing that can help me is if she just comes and holds me and tells me that she loves me. I know for sure that that will help me’.

Some participants described how their suicide attempts had been precipitated by a sense of disconnection and ‘not belonging’. Jane said, ‘I’m just like faulty, I just don’t belong here. I don’t belong here’.

Some participants said they needed help to establish supportive and trusting relationships. Andile articulated this by saying he needed to ‘meet people that are in my situation. Maybe we can support one another and help and talk to one another openly, instead of getting that pushing down from people’. Colin echoed this saying,

Doctor, like I say, (I need) people I can talk to, just for one minute or two minutes or whatever for a day. People I can talk to who say, listen here, we are interested in helping you. We are genuinely interested to help you.

Similarly, Sally said, ‘I need a person that I can trust to speak to, and help me through the difficult times that I’m facing’.

Participants acknowledged that they needed help with establishing intimacy, and finding connection. Liz said,

It’s like nobody is around. I can’t get close to somebody. I tried so much. I even took a husband, got married and had a child, but I can’t be close to somebody. I’m so far away from my child as well. I can’t get closer to somebody.

Participants said that they needed to learn how to communicate their distress to others and how to ask for help in adaptive ways. Anthea described her experience saying, ‘… it feels like you’re alone. No one worries about your problems. If you speak, and try to explain, people do not realise that you are calling out for help’. She said she would need help to communicate her distress in order to ensure that she did not hurt herself again as a way of accessing support and care.

Assistance with resolving interpersonal and family conflict

Participants expressed a need for help to resolve interpersonal difficulties and conflicted family relationships. Tanya said, ‘I need to sort out my relationship with my family’. Neil said,

The support I will need, if Doctor can phone my family and explain to them that it’s not to do with money or anything like that. I just want them to support me where if I’ve got something to tell them, you know, don’t turn their backs towards me …

Some participants also asked for help with communicating their needs to their family. Nigel said,

My family also supports me, but I would like them to also support me where – it’s like if I go, at least come with me once a month to the psychiatric sister or to the doctor and ask how am I doing. You know, things like that, which they don’t do.

Participants described difficult family circumstances and identified family relationships as significant stressors which contributed to their suicide attempt. Nancy said, ‘Sometimes I wish I wasn’t born into this family’. Participants said that they needed help to achieve independence from their families and expressed a desire for liberation. Chris said, ‘I’m getting tired of living with my family. That’s what I want, I want to get my own place’. Fatima said, ‘And I really do want to be independent – I don’t want to have to ask my family and people for stuff’.

Help to solve situational problems

Participants expressed a need for practical help to solve concrete situational problems such as finding employment, resolving a financial or legal crisis and finding a place to stay. Economic hardship and unemployment were identified as significant stressors which contributed to participants’ suicidal behaviour; they said they would need relief from this burden. Jack said, ‘But a lot of it boils down to money’, and Frans said, ‘I also want to have my own business. I need to make some money’. Participants expressed a belief that if they had access to financial resources they would feel liberated from and more in control of the situation that had precipitated their suicidal crisis. Sabu explained, ‘I think what will help me is a better job to get out of that harshness. I’m applying for jobs. I think that’s one big obstacle …’. Similarly, Mark said,

Like if I had won the jackpot, I could just have taken care of all those worries … I’m just saying if I was in a position where I could at least have been in control, I don’t believe that I would have gotten to that stage (of wanting to die).

Many participants said they needed help finding a place to live. Participants who were homeless described the hardship of having no fixed abode and living on the street or in a shelter. Other participants expressed a desire to find alternative accommodation because their current living arrangements were stressful. Karin said, ‘If I had a home as well, an actual house where I could go …’, and Fred said, ‘I want to move out of the house (where I am staying now)’.

Discussion

Participants in this study are explicit in their request for ongoing psychiatric and medical care, as part of a strategy to curb their own suicidal behaviour. This assertion seems to point to an implicit understanding that their behaviour was at least in part a function of psychiatric and medical conditions which they believe doctors are equipped to alleviate. It is possible that the primacy participants give to what is essentially a biomedical formulation of their own suicidal behaviour, reflects something of the context in which the data for this study were collected. Interviews for this study were conducted in a medical setting by a researcher identified as a psychologist and with patients who had been admitted to hospital and who had all received medical and psychiatric care following their suicide attempt. It is plausible that the demand characteristics of the situation in which the data were collected, the positioning of the data collector as a psychologist and the participants recent experience of receiving psychiatric and medical care, all shaped how the participants narrated their suicide attempt and framed what they expressed as their support needs. It is also possible that the participants in this study are simply reproducing the dominant biomedical view of suicidal behaviour as a symptom of mental illness because they believe this is what is expected of them and they do not want to appear unsophisticated in their understanding of their own behaviour. Nonetheless, if we do as authors such as Hjelmeland (2012) suggest, and we take seriously the insights and voices of the ‘true experts’ on suicide (i.e. suicide attempters) when seeking to understand suicidal behaviour and plan interventions, then we have to acknowledge that many participants in this study explicitly attribute their suicide attempts to symptoms of mental illness and they ask for ongoing professional care to alleviate these psychiatric symptoms.

From the perspective of the participants in this study, appropriate, affordable, accessible, ongoing psychiatric care is integral to suicide prevention. This finding is entirely consistent with the conclusions drawn in a review of systematic reviews on effective interventions for suicidal behaviour; the review concluded that best practice in suicide prevention entails detecting and treating psychopathology, principally depressive disorders (Van der Feltz-Cornelis et al., 2010). The focus on depression in the international suicide prevention literature is noteworthy, particularly since the participants in this study identify needing help with symptoms of anxiety, psychotic illnesses, posttraumatic stress disorder, grief and pain disorders. Participants’ request for ongoing post-discharge psychiatric care is congruent with international best practice literature. There is evidence that ‘professionals’ maintenance of long-term contact with patients at risk of suicide can exert a suicide-prevention influence’ (Motto and Bostrom, 2001: 828).

Participants’ explicit request for ongoing psychiatric and psychological care for symptoms of psychopathology and psychological distress calls into question the assertion by Hjelmeland et al. (2014) that mental illness and suicidal behaviour are unrelated and that suicide may be ‘about something quite different’ from mental illness (p. 1370). It is significant that suicide attempters in this study, all of whom live in resource-constrained environments, explicitly assert the importance of access to psychiatric care and ask for access to ongoing psychiatric treatment and integrated psychological support. There is a real risk that claims made by authors such as Hjelmeland et al. (2014) that suicidal behaviour is not a symptom of mental illness will be taken up by policy makers to undervalue or discredit the importance of providing appropriate, accessible and effective psychiatric and psychological care as part of national suicide prevention strategies. There is a particular danger of this in low-resource environments, like SA, where the health care system is already under pressure and where there are a high number of competing demands for scarce psychiatric resources (Chopra et al., 2009; Mayosi et al., 2012). It is one thing to argue as Hjelmeland et al. (2014) do that suicide is not a psychiatric condition, in high-income countries where there is easy access to effective psychiatric care for those who seek it, but it is quite another thing to argue that psychiatric responses to suicide prevention are not important in contexts where there are inadequate and inaccessible psychiatric services.

It is significant that participants describe their struggle to access mental health care at a community level and feeling rushed by the psychiatric service providers they encounter. This finding draws attention to the health care context within which suicide attempts occur in SA and highlights the significant mental health treatment gap, the shortage of public mental health resources, and the fact that primary health care in the country is still predominantly biomedical in its orientation (Lund et al., 2010; Petersen, 2000; Petersen and Lund, 2011). The participants in this study request ongoing integrated person-centred care which includes psycho-social support at a community level. The findings suggest that integrating psycho-social services in primary health care settings is important for suicide prevention in SA. It is noteworthy that the policy framework to achieve this is already in place in the country (Department of Health, 1997) in the form of the White Paper for the Transformation of the Health System in South Africa, but that implementation of this policy has been hampered by a range of factors, including resource constraints and a lack of political will. The constraints on clinic staff in primary health care and community clinics have resulted in a situation where health care workers do little more than dispense medication without providing any psycho-social interventions (Bond et al., 2001; Petersen, 2000).

While participants attribute their suicidal behaviour to psychological distress and mental illness, they also acknowledge the significance of interpersonal factors and explicitly ask for help in establishing connectedness, building relationships and resolving conflict with family members. This finding highlights the interpersonal context of suicidal behaviour and draws attention to the need for family interventions and interpersonally orientated psychotherapy with individuals who have made suicide attempts. Interpersonal psychotherapy (IPT) is a time-limited psychotherapy that focuses on current problems with patients’ relationships (Weissman et al., 2000). A systematic review by Cuijpers et al. (2011) concluded that there is good evidence to support the use of IPT for depression. Similarly, evidence supports the use of IPT to treat mood disorders in primary care settings (Linde et al., 2015). Studies have also showed that IPT is effective in significantly reducing suicide ideation, death ideation, repetition of self-harm and depressive symptom severity, and resulted in significant improvements in perceived meaning in life, social adjustment, perceived social support and psychological well-being (Guthrie et al., 2001; Heisel et al., 2015). A systematic review, meta-analysis and meta-regression of psychological and psycho-social approaches to reduce repetition of self-harm, concluded that on the basis of current evidence it is not yet clear what interventions should be delivered to reduce the risk of repetition of self-harm although ‘interventions with an interpersonal focus and targeted on the precipitants of self-harm may be the best candidates’ (Hetrick et al., 2016: 1). Integrating psychological interventions alongside psychiatric and pharmaceutical interventions is consistent with best practice in suicide prevention which advocates the use of multilevel strategies, that is, strategies that target several populations or several levels within health care systems or include interventions with more than one focus, such as pharmacotherapy and psychotherapy (Van der Feltz-Cornelis et al., 2010).

The findings suggest that suicide attempters believe in the importance of establishing connectedness as an antidote to suicide. The interpersonal-psychological theory specifically identifies the role of ‘thwarted belonging’ in the aetiology of suicide (Hagan et al., 2016). Lack of relationships, social isolation, loneliness, living alone and being single are all recognised risk factors for suicidal behaviour (Ferreira et al., 2015; Holt-Lunstad et al., 2015). Participants’ request for help to establish connection and belonging is thus consistent with theory and empirical evidence. This finding highlights the importance of addressing the interpersonal context in which suicidal behaviour occurs as part of the post-discharge care of suicide attempters.

It is noteworthy that participants articulate a need for practical help to solve socio-economic and situational problems, such as financial difficulties, unemployment, homelessness and legal difficulties. These situational stressors are outside the bounds of medicine to remedy, and highlight the social and economic factors that contribute to suicidal behaviour. This finding suggests that teaching suicide attempters problem-solving strategies, helping them to identify alternative solutions and connecting them to appropriate social and legal services might reduce their suicide risk. There is evidence of the effectiveness of brief problem-solving cognitive therapy for suicide prevention (Gustavson et al., 2016; Hatcher et al., 2011; Stewart et al., 2009). Data suggest that brief cognitive therapy produces rapid changes on dimensions such as negative problem orientation and impulsivity/carelessness problem-solving style (Ghahramanlou-Holloway et al., 2012). Of course, there is a danger here that advocating problem-solving interventions delivered to individuals as part of a suicide prevention strategy reinforces the idea that suicidal behaviour is a problem that is contained within the individual and that building the resilience of the individual is central to preventing suicide. Interventions, such as teaching problem-solving skills to individuals and building individual resilience, deny the reality that many individuals in LMICs live under conditions of endemic violence, structural poverty, enduring inequality and continuous trauma. It is noteworthy that participants in this study ask for help with problems such as poverty, unemployment and homelessness. It is also significant that in the context of their suicide attempts, participants talk about their exposure to trauma and violence, particularly sexual- and gender-based violence. While it is no doubt important to provide problem-solving interventions and build individual resilience in order to prevent suicides, it is equally important to take note of what the data reveal about socio-economic and contextual problems that are faced by the suicide attempters in this study. This social and economic context needs to be addressed at a political and economic level in order to reduce suicidal behaviour in LMICs. Fox (2003) has noted that when research directly engages with the ‘politics of the setting which it explores’ (p. 96, emphasis in original), there is real potential for the research to become a source of social justice and to transform social relationships. The insights of the participants in this study certainly make visible sites of potential social and political transformation.

The kind of support requested by the suicide attempters in this study is unlikely to be achieved without intersectoral collaboration. The WHO has defined intersectoral collaboration for health as the establishment of working relationships between the health sector and other sectors (such as social welfare, criminal justice and housing services) in order to improve health outcomes more effectively, efficiently or sustainably than would otherwise be achieved (WHO, 1997). The effectiveness of intersectoral collaboration for the community-based treatment of mental illnesses has been documented and is considered best practice in a number of high-income countries (Darlington and Feeney, 2008; WHO, 1997). The importance of intersectoral collaboration for suicide prevention, particularly in LMICs, has been described (Bertolote et al., 2006; Reynolds and Conroy, 1999; Wang et al., 2008; WHO, 2012). There are a number of obstacles to achieving intersectoral collaboration in LMICs (Brooke-Sumner et al., 2016). Intersectoral approaches to the treatment of mental illnesses in SA have not been adequately achieved, especially in rural areas, because of the lack of dedicated funding for such approaches (Petersen and Lund, 2011; Thornicroft et al., 2010), failures of management and leadership to implement relevant policy (Jenkins et al., 2011) and a lack of political will (Brooke-Sumner et al., 2016). Brooke-Sumner et al. (2016) have further noted that the attainment of effective intersectoral collaboration is a complex challenge in LMICs due to the scarcity of resources, problems pertaining to the allocation and use of resources, inadequate organisational structures, mistrust, poor communication and a lack of distinct roles.

Limitations

Data were collected from one urban hospital in SA. It is possible that the experiences described by these participants are highly context specific and representative of public services within a particular region of the country. Suicide attempters in other centres and rural areas may have different support needs and other experiences of the health care system.

Conclusion

The findings of this study draw attention to the need for suicide prevention efforts in SA to include access to psychiatric care and integrated person-centred psycho-social support at a primary health care level. Furthermore, the findings highlight the interpersonal, social and economic context within which suicide attempts occur as well as systemic problems with the organisation of care within the health system which impede suicide prevention in SA. From the perspective of the participants in this study, multilevel approaches and intersectoral collaboration would seem to be important components of a national suicide prevention programme in SA. There would, however, appear to be a number of problems with the structure and delivery of health care in SA which impede the attainment of these ideals. Crucially, the findings of this study draw attention to the need in LMICs like SA, for suicide prevention interventions to focus on providing psychiatric and psychological care, and building individuals’ resilience while also addressing systemic social and economic issues.