Key points are not available for this paper at this time.
BACKGROUND: Homicide followed by suicide is rare, devastating and perpetrated worldwide. It is commonly assumed that the perpetrator had a mental disorder, raising concomitant questions about prevention. Though events have been reported, there has been no previous systematic review of the mental health of perpetrators. AIMS: Our aims were twofold. First, to identify whether there are recognisable subgroups of homicide-suicides in published literature and, secondly, to investigate the relationship between perpetrator mental state and aspects of the incident. METHODS: We conducted a systematic review of published literature on studies of homicide followed within 24 h by suicide or serious suicide attempt that included measures of perpetrator mental state. RESULTS: Sixty studies were identified, most from North America or Europe. Methodologically, studies were too heterogeneous for meta-analysis. They fell into three main groups: family, mass shooter, and terrorist with an additional small mixed group. There was evidence of mental illness in a minority of perpetrators; its absence in the remainder was only partially evidenced. There was no clear association between any specific mental illness and homicide-suicide type, although depression was most cited. Social role disjunction, motive, substance misuse and relevant risk or threat behaviours were themes identified across all groups. Pre-established ideology was relevant in the mass shooter and terrorism groups. Prior trauma history was notable in the terrorist group. CONCLUSION: Research data were necessarily collected post-incident and in most cases without a standardised approach, so findings must be interpreted cautiously. Nevertheless, they suggest at least some preventive role for mental health professionals. Those presenting to services with depression, suicidal ideation, relationship difficulties and actual, or perceived, changes in social position or role would merit detailed, supportive assessment over time.
Theodorou et al. (Sun,) studied this question.