Statistics from Altmetric.com
Question: How many people with first-episode psychosis have a history of deliberate self-harm (DSH), or commit DSH during the early course of treatment; and what demographic and clinical factors are associated with DSH?
Outcomes: DSH—A history of DSH was defined as a report of DSH at the time of or before initial contact for treatment of psychosis; DSH after treatment was defined as report of DSH in people previously treated for psychosis, regardless of treatment duration.
Design: Systematic review and meta-analysis.
Data sources: MEDLINE, EMBASE and PsycINFO were searched from inception to March 2012, supplemented by hand search of reference lists of included studies and search of the first 1000 hits of Google Scholar.
Study selection and analysis: Studies reporting the association between first-episode psychosis or schizophrenia or early schizophrenia and suicide attempt or DSH. As identified studies did not distinguish between lethal and non-lethal attempts, DSH was the preferred term of the review. Any mental health condition with psychosis as a feature was eligible. Studies of suicidal ideation without attempt were excluded. Demographic and clinical factors potentially associated were not defined by the review a priori but factors investigated in five or more studies were included in the meta-analysis. A random effects model was used to calculate standardised mean differences (SMD) for continuous data and ORs for dichotomous data.
Twenty-three studies met inclusion criteria, 18 (n=6962) reporting on DSH prior to psychosis treatment, and 13 (n=4846) reporting on DSH after treatment initiation. The prevalence of DSH at any time prior to psychosis treatment was 18.4% (95% CI 14.4 to 23.2; 18 studies, I2=93.8%). In 7 years after treatment initiation the prevalence of DSH was 11.4% (95% CI 8.3 to 15.5; 13 studies, I2=89.2%). Factors associated with increased risk of DSH were younger age at time of first-episode psychosis and treatment onset (for age at onset, SMD −0.28, 95% CI −0.50 to −0.60; for age at treatment SMD −0.18, 95% CI −0.31 to −0.04); longer duration of untreated psychosis (SMD 0.20, 95% CI 0.05 to 0.35); history of alcohol or substance use (alcohol: OR 1.68, 95% CI 1.16 to 2.44; substance: OR 1.46, 95% CI 1.14 to 1.88); depressed mood (SMD 0.49, 95% CI 0.35 to 0.62); expressed suicide ideas (OR 2.34, 95% CI 1.66 to 3.30) and greater insight into the presence of psychotic illness (OR 1.64, 95% CI 1.23 to 2.56). DSH prior to treatment initiation was associated with increased risk after starting treatment (OR 3.94, 95% CI 2.72 to 5.71). Heterogeneity was significant for all of these analyses with the exception of those for alcohol and substance use. Features of psychosis, including function and presence of positive or negative symptoms were not associated with DSH. Alcohol and substance use and depressed mood were the only factors associated with DSH both before and after treatment initiation.
In people with first-episode psychosis, DSH is common and associated with a number of factors including longer duration of illness, depression, and alcohol and substance use.
It was not clear from the reporting in the review whether the factors tested for association with DSH were assessed before the onset of DSH, which limits the conclusions that can be drawn regarding potential causality.
There is a lot to learn about suicide prevention in early psychosis. Suicide risk remains high not only in young individuals already attending mental health services but also in those with emerging symptoms of psychosis, who are still undetected in the community. Therefore, the importance of acquiring more knowledge and skills in the prediction of suicide must not be belittled.1
In this context, Challis and colleagues provide a useful meta-analysis. An inevitable challenge for works in this field is the terminology employed and related clinical meanings. Following a thorough literature search, the authors decided to group suicide attempts, self-injury and self-directed aggression under the term ‘deliberate self-harm’. Although they prove that between-study heterogeneity could not be explained by differences in the definition or choice of terminology, it should also be acknowledged that these self-destructive behaviours could follow different avenues in a continuum ranging from self-harm ideation to completed suicide. Indeed, different factors, such as comorbid personality disorders could play a role in the differentiation of these events, especially with regard to lethal purposes.
The results emphasise the importance of an accurate evaluation and treatment of depressed mood and substance misuse to reduce suicide risk in early psychosis. Interestingly, the authors also found an association between longer duration of untreated psychosis and self-harm before and after treatment initiation. These findings are relevant and appropriately contextualised as ‘additional aims’ for early intervention, without underestimating other important aspects of care, such as impaired functioning, which may not be strongly associated with suicidality, but still represent fundamental targets for the management of young people with early psychosis.
Hence, risk management plans should ideally consider mood-related clinical measures that, paradoxically, may be worsened by improved insight and information regarding substance misuse. This should be part of wider care programmes tailored to the specific needs of each patient in order to avoid suicide by enlivening their sense of self.
Competing interests None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.