Policy responses

The elevated risk of suicide after the suicide of a spouse was identified over a decade ago,Reference Agerbo² but limited progress has been made in addressing it. The suicide prevention strategies of most high-income countries advise comprehensive support programmes for people bereaved by suicide, also known as postvention, but are vague on their components. The dearth of trial evidence is to blame. A 2008 systematic review in this Journal highlighted the lack of interventions found to reduce suicide risk in people bereaved by suicide, and the few shown to improve their mental health or social functioning.Reference McDaid, Trowman, Golder, Hawton and Sowden⁵ Whereas accumulating research studies support a range of interventions to address pathological grief in people bereaved by any cause,Reference Boelen and Smid⁶ applying broad guidelines based on evidence from heterogeneous populations ignores issues specific to suicide loss. Approaches that fail to acknowledge the shame and self-stigma that tend to occur after a suicideReference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ are unlikely to be acceptable. Interventions that fail to address the specific mediators of suicide risk are unlikely to mitigate this risk. As yet, no epidemiological studies have identified the factors mediating suicide risk in people bereaved by suicide. Trial activity has stalled and needs invigorating. Gaining impetus relies on developing interventions addressing specific mediators of suicide risk while also meeting the need for support.

Likely pathways

Pathways to suicide in people bereaved by suicide can be hypothesised from available evidence. A recent Danish study of adults bereaved by a partner's suicide identified specific psychiatric disorders associated with suicide bereavementReference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ that are likely contributors to suicide risk. Suicide-bereaved partners, defined by marriage, civil union (legal in Denmark since 1989) or cohabitation, were at increased risk of incident mental disorder (specifically mood disorders and post-traumatic stress disorder).Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ They had higher rates of psychiatric admission after the death, and an increased risk of suicide and all-cause mortality.Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ Unlike suicide-bereaved men, women were at increased risk of self-harm and homicide.Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ Although rates of substance misuse exceeded those for the general population, they were similar to those for bereaved controls.Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³

A Canadian longitudinal study of parents bereaved by suicide identified a substantial burden of mental and physical health problems even before their loss.Reference Bolton, Au, Leslie, Martens, Enns and Roos⁷ Likely explanations include genetic traits shared with offspring for suicidality and mental illness, assortative mating based on the same traits, shared social adversity and the strain of being on suicide watch. These findings suggest a need to design acceptable systems for screening and treating incident or worsening psychiatric illness in partners and relatives bereaved by suicide, delivered in a manner that avoids pathologising grief.

Despite these vulnerabilities, survey evidence shows that the suicide-bereaved are poorly supported by family and friends and may feel too ashamed or unworthy to ask for support.Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸ The influence of stigma on help-seeking and on others’ propensity to offer help may also contribute to suicide risk. In a British survey, people bereaved by suicide were less likely to report receiving informal support than bereaved controls, or to have received formal or informal support in the immediate aftermath.Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸ Differences became non-significant when adjusting for perceived stigma.Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸

Danish men bereaved by a partner's suicide were less likely to consult their general practitioner than bereaved controls, despite their greater likelihood of mental illness.Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ These men were no more likely than the general population to be prescribed antidepressants,Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ albeit in a society with generally high rates of antidepressant prescribing. Bereaved at a mean age of 54, they had lost a key confidant(e) and the main social influence on their healthcare help-seeking. Their help-seeking behaviour might explain the much higher magnitude of suicide risk in bereaved male than female partners.Reference Agerbo² Together this evidence suggests that systems of proactive outreach are indicated after suicide so that access to support does not rely on bereaved individuals self-referring.

Studies describing the timing of onset of psychiatric illness and suicidality are critical in identifying windows of preventive opportunity. After a child's death, parents’ risk of suicide is particularly marked in the first month.Reference Qin and Mortensen⁴ These early weeks are a period during which people bereaved by suicide report being particularly poorly supported,Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸ reinforcing their sense of shame and self-stigma. Early intervention during this vulnerable period is likely to be indicated given the potential for deteriorating mental illness to contribute to suicide risk. This also promotes help-seeking for psychosocial difficulties by countering early on the stigmatising experience of being avoided by one's own network.Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸

Service evaluation

Services provided after suicide bereavement have been largely driven by the voluntary sector, responding to the demand for individual bereavement counselling and group peer support. Geographical provision is variable, particularly in rural areas, where suicide rates may be higher. Survey evidence indicates that a quarter of British people bereaved by suicide receive no formal or informal support after the loss, as distinct from those who prefer to handle it alone.Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸ A 2017 revision of the English suicide prevention strategy acknowledged poor progress in supporting families bereaved by suicide,⁹ and pledged to strengthen its focus by issuing local commissioning guidance. This emphasises collaborations between first responders (often emergency services), coroners’ officers, the voluntary sector and healthcare services, such as the systems of early proactive outreach currently being piloted in the UK⁹ and elsewhere. Local service evaluations provide valuable information on acceptability and uptake, identifying whether early intervention is likely to be welcomed, but we also need trial evidence of effectiveness to demonstrate the clinical impact and maximise the sustainability of funding.

With so much attention currently focused on commissioning services for the suicide-bereaved, there is a need to fill evidence gaps swiftly to inform implementation. Priorities include identifying the mediators of suicide risk in partners and relatives bereaved by suicide, including exploring the role of suicide suggestion (or contagion), and identifying acceptable screening systems to address incident or deteriorating mental illness. These are necessary first steps before designing controlled trials to assess whether services are effective in reducing suicide attempt rates. Assessments of their effects on proximal outcomes, such as mental illness, help-seeking behaviour, pathological grief and social functioning, are also important, as is measurement of potential adverse effects.Reference de Groot and Kollen¹⁰ The results of these studies can be used to inform commissioning of services for suicide-bereaved partners, relatives and friends: a group prioritised within suicide-prevention strategies but ostracised by those in their social networks.Reference Pitman, Rantell, Moran, Sireling, Marston and King⁸ Suicide prevention interventions are notoriously hard to evaluate, but only with trial evidence of clinical and cost-effectiveness in addressing the observed risks of mental illnessReference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ and suicideReference Agerbo^2, Reference Erlangsen, Runeson, Bolton, Wilcox, Forman and Krogh³ can we be more directive with the recommendations made in suicide prevention policies, and more certain that they might have their intended effect.