|| Checking for direct PDF access through Ovid
The contribution of people with psychosis to homicide statistics is small, but there is a statistically significantly higher rate of homicide among them than in the general population. Legal authorities and the wider public call for more information. Our aim was a narrative synthesis of empirical literature generated by systematic searches for the five complete years 2013–2017.One article showed that people with psychosis are five times more likely to be homicide victims than those without, but focus remains on perpetrators. Consensus is that although psychotic symptoms contribute to homicidal acts, so do many of the variables that increase homicide risk more generally – including substance use, trauma histories and access to weapons; thus, both the epidemiology of homicide and some details, like method, may be country specific. In 2013–2017, variation in service provision as a risk factor for homicide has emerged more clearly but, overall, research made few homicide-specific advances.Criminal homicide is, fortunately, uncommon in most countries. Homicides by people with psychosis are so rare that they do not rate mention in the 2013 United Nations Global Survey of Homicide. This may account for the limitations to much research for this group. Despite some homicide-specific evidence of advantage for early intervention for psychosis, impact of improving treatment – the most promising way forward – will generally have to be inferred from the not entirely satisfactory surrogate of effects of treatment on violent offending more generally.