Highlights of this issue

Risky business

Self-harm is common, suicide relatively rare. The prevention of both is a cornerstone of mental healthcare but is hindered by the fact that most associated factors are unremarkable and non-specific. Several papers in this month's BJPsych explore crisis presentations and risk predictors. Young people presenting to accident and emergency (A&E) is a daily challenge for many of us; little work has looked at what it feels to be that individual in the middle, though it is an obvious opportunity to learn and improve that critical first face of mental health. Owens et al (pp. ) analysed an online forum for young people (aged 16–25) who self-harmed; sadly, a large theme was how they try to avoid A&E, partly out of a sense of shame. This runs interestingly counter to a background NHS narrative of crisis ‘frequent fliers’; however, such professional perceptions might link with a second theme of individuals avoiding A&E because of past negative experiences with staff. Positively, the work found that a non-judgemental and kind approach encouraged appropriate attendance. We are mental health, surely these are our core qualities? We don't see what we don't see; data suggest that only a minority of young self-harmers seek help – this work certainly re-emphasises the ‘liaison’ in liaison psychiatry.

Baer et al (pp. ) and Galway et al (pp. ) respectively evaluate self-poisoning severity and substance misuse as risk factors. Intentional overdose accounts for about four-fifths of suicide attempts, though curiously the literature suggests that its severity is not routinely taken into account by psychiatrists in subsequent risk profiling. Baer et al found that, compared with matched controls, individuals who had required mechanical ventilation had more frequent subsequent depressive disorders – with greater symptom severity – and showed lower survival rates 1 year later. Substance misuse often goes hand-in-glove with low mood and as a precipitant for impulsive self-harm, but less is known about the relationship between lifetime use and rates at the time of suicide. Galway and colleagues evaluated 2 years of coroners' reports and primary care records in a cohort of over 400 instances of suicide: almost a quarter had sought help for alcohol use in the previous year, and over a third had done so at some point in their lives; over half had alcohol in their bloodstream at the time of death, at levels of intoxication in about 40%.

Primum non nocere

Electroconvulsive therapy (ECT) suffers bad press. It remains publically controversial despite an excellent evidence base, and there are lingering concerns about longer-term iatrogenic harms. Preconceptions are challenged this month: Kaleidoscope (pp. ) reviews prospective data showing that ECT increased grey matter volumes across the temporal lobes and hippocampi, whereas Kirov et al (pp. ), analysing 10 years of cognitive performance data, show that repeated courses of ECT did not lead to cumulative cognitive deficits. Memory impairments, which are common, appear limited to the days following treatment; the findings should be reassuring for our patients.

I once had a patient tell me that he always thought he'd a happy childhood until he underwent psychoanalysis. It's anecdotal, but points towards a deeper and generally unexplored issue; we quickly think of side-effects with ECT and pharmacological interventions, but have been surprisingly slow to consider this with psychological ones. Although studies of psychological therapies report non-response to therapy, unlike the pharmacology literature they typically do not report rates of actual harm. Crawford and colleagues (pp. ) assessed almost 15 000 patients across England and Wales. One in twenty reported lasting negative effects from psychological therapy. There are undoubtedly challenges to this type of cross-sectional survey and there is a need to have a deeper understanding of this issue, but a gauntlet is thrown to future studies.

Few mental health issues are as pressingly urgent as the needs of traumatised refugees. We are in the midst of the greatest refugee migration since the Second World War; daily news reports discuss the multifaceted issues of accommodation and assimilation, race and racism, and perilous journeys of escape. In this context, many of us will have wondered about the efficacy of our standardised interventions in the multi-ethnic and linguistic groups of traumatised individuals who attain safety in Western nations. Buhmann et al's paper (pp. ) is timely, randomising refugees with war-related traumas to cognitive–behavioural therapy, antidepressant medication, their combination, or placebo. In this, the largest study of its kind, neither active intervention had any effect on symptoms of post-traumatic stress disorder (PTSD), even in combination, although medication had a modest impact on depressive symptoms. What are the factors hindering improvement? Is it the imposition of a Western-centric model of managing trauma, the confounder of individuals still adjusting to foreign and potentially unwelcoming conditions, the presence of comorbidities, or, despite best efforts, the lack of sufficiently culturally informed care? The beta draft of ICD-11 contains the construct of ‘complex PTSD’ to separate those individuals who have undergone exposure to extreme and prolonged or repetitive threatening or horrific acts from which escape is difficult or impossible, and further notes that their subsequent symptom profile may differ to include severe and pervasive problems in affect regulation and persistent self-beliefs. However, most PTSD work has focused on non-war traumas, and in Western populations. These current findings are disheartening, but the growing issue mandates urgent work on the topic.

Parity of esteem

This month's Kaleidoscope (sadly) shows that happiness doesn't make one healthy. Batelaan et al (pp. ) add to the bad news, their meta-analysis finding that anxiety is associated with a greater than 50% increase in the onset of cardiovascular disease (CVD) and crucially, the link appears to be a causal one. The underlying physiological or behavioural mechanisms are yet to be elucidated, but it produces a CVD risk profile similar to that engendered by obesity. When physical illness does develop, those with mental health difficulties often do worse, for a variety of reasons. Using a national database, Ishikawa and colleagues (pp. ) evaluated healthcare access and outcomes in individuals with schizophrenia who developed gastrointestinal cancer. They found that these patients had a higher proportion of late-stage cancer and lower rates of endoscopic and invasive surgical treatments than those without any psychiatric disorder. Compounding this, they had greater in-hospital mortality rates even after adjusting for cancer stage and treatment. From pathogenesis, through illness, to death: Hirvikoski et al (pp. ) took a population-based cohort to explore relative mortality rates in autism spectrum disorders (ASD). Compared with the general population, overall, those with ASD were two-and-a-half times more likely to die prematurely, with significantly greater risks in those with lower-functioning ASD. Previous studies have suggested that an excess of comorbid neurological difficulties such as epilepsy might underlie this, but in the current work there were elevated mortality rates in almost all physical diagnostic categories. Parity of esteem? Mind the gap.