Urgent need for better care after self-harm

1 Patton GC, Sawyer SM, Santelli JS, et al. Our future: a Lancet commission on adolescent health and wellbeing. Lancet 2016; 387: 2423–78. 2 Whitley R. Global mental health: concepts, conflicts and controversies. Epidemiol Psych Sci 2015; 24: 285–91. 3 Patel V, Saxena S, Lund C, et al. The Lancet Commission on global mental health and sustainable development. Lancet 2018; 392: 1553–98. 4 Caldwell MC, Davies SR, Hetrick SE, et al. School-based interventions to prevent anxiety and depression in children and young people: a systematic review and network meta-analysis. Lancet Psychiatry 2019; published online Nov 13. https://doi.org/10.1016/S2215-0366(19)30403-1. 5 US National Research Council, Institute of Medicine. Preventing mental, emotional, and behavioral disorders among young people: progress and possibilities. Washington, DC: National Academies Press, 2009. 6 Betancourt T, McBain R, Newnham E, et al. A behavioral intervention for war-affected youth in Sierra Leone: a randomized controlled trial. J Am Acad Child Adolesc Psychiatry 2014; 53: 1288–97. 7 Fazel M, Hoagwood K, Stephan S, Ford T. Mental health interventions in schools in high-income countries. Lancet Psychiatry 2014; 1: 377–87. 8 Weist MD, Hoover S, Lever N, et al. Testing a package of evidence-based practices in school mental health. School Mental Health 2019: 1–15.


Urgent need for better care after self-harm
A presentation in the emergency department with self-harm is one of the strongest predictors of death by suicide that has been identified. There is, therefore, good reason to carefully assess people who present with self-harm in order to provide the best aftercare option. It is well known that one or two people out of 100 who present with self-harm will die within the next year. Such dire perspectives underscore the need for providing adequate support for people after self-harm.
Questionnaires and risk screens may give us an indication of who might be in the high-risk group. However, several well-researched studies have shown that these tools are poor at predicting which people we should be attentive towards. One study found that the clinician's estimate was as good as the patient's own evaluation; both of these evaluations fared just as well-or badlyas established clinical scales. 1 Similarly, a meta-analysis revealed that the accuracy of the evaluated scales was not good enough to make informed decisions regarding what type of treatment allocation should be recommended. 2 In The Lancet Psychiatry, Galit Geulayov and colleagues 3 compared the risk of suicide following hospital presentation for self-harm according to patient characteristics, method of self-harm, and variations in area-level socioeconomic deprivation, and estimated the incidence of suicide by time after hospital attendance. By following 90 614 self-harm presentations by 49 783 individuals to hospitals in five English cities, 3 the authors identified 703 people who died by suicide within the subsequent 16 years. The incidence of suicide in the first 12 months following the index presentation to hospital for self-harm was 55·5 times (95% CI 49·2-62·8) higher than the incidence in the general population in England (2000-13). Relative to the general population, the highest inflation in suicide rate after 12 months of follow-up was observed in adults aged at least 55 years. The first month after discharge, this group had a particularly high incidence of 1787·1 (1423·0-2244·4) per 100 000 person-years, which is close to 200-fold higher than in the general population. This high incidence of suicide emphasises the importance of care immediately after presentation of self-harm. Additionally, men were three times more likely than women to die by suicide after self-harm (OR 3·36 [95% CI 2·77-4·08], p<0·0001). Other predictors that were linked to higher risks of suicide were having presented several times with self-harm, self-harm with a more lethal method, and use of several methods of selfharm. Counterintuitively, people who lived in the least deprived areas were more likely to die by subsequent fatal episodes of self-harm compared with people in most deprived areas. The reasons for this are unclear, but Geulayov and colleagues 3 made some suggestions.
What are the options of treatment allocation in the emergency department? First, the clinician might choose to refer the patient to a psychiatric ward for admission, after a psychosocial assessment. However, this option is limited by availability of hospital beds. Second, the patient might be referred to outpatient treatment, which can be a better solution if the patient is not under immediate risk of self-harming. Few countries offer the option of referral to specialised suicide prevention clinics where the patient can engage in sessions of psychosocial therapy to address the causes that led to the self-harm and possibly develop better coping strategies for future critical situations. 4 Some countries offer follow-up through home visits. 5 Unfortunately, in quite a few countries patients presenting with self-harm are likely to Comment www.thelancet.com/psychiatry Vol 6 December 2019 be sent home with a referral to their general practitioner, or with no referral at all. Other instruments, such as green cards allowing for an easy admission if a crisis should arise, check-up calls, or postcards, could help the patient respond better to future crises. These have instruments have been tested, with varying success. 6-8 National Institute for Health and Care Excellence guidelines provide a good overview of current recommendations and underscore the urgency of early intervention. An increasing range of eHealth self-help tools are emerging, including safety plans and self-help online therapy. 9,10 Although the body of evidence is still small, these could be viable options for a person in crisis who is searching the internet. Their appeal is enhanced by being inexpensive and not geographically limited.
The bottom line is that while the body of evidence of effective intervention is growing, we need to help people who present with self-harm. Operating in such a scenario is challenging but the numbers are clear; we need to ensure that patients receive support immediately when presenting and implement a continuation of care after discharge.

Common risk factors for abortion and suicide attempts
In the field of abortion and mental health research, few studies have focused on suicide attempts. The study of Julia R Steinberg and colleagues 1 in The Lancet Psychiatry is based on an impressively large, bias-free, real-world registry dataset, and is a welcome addition to the academic literature. Steinberg and colleagues correctly limit their analyses to first-time abortions and firsttime suicide attempts; although this decreases external validity, it is a necessary step to establish causal effects and ensure temporal order. However, the fact that the data are based on registries also poses some limitations to the interpretation of the results; furthermore, their findings do raise an important question. The results suggest that the abortion itself was not the reason for the increased risk of suicide attempts, but that one or more unmeasured variables contributed to it. Obviously, Steinberg and colleagues 1 could not capture what these co-occurring variables might have been. If the increased risk was not due to the abortion itself, then why was the risk higher in women who had an abortion the year before and after the abortion, relative to more than one year after the abortion?
One of the biggest challenges is separating potential effects of abortion from pre-existing risk factors that might predispose women to mental health problems or suicide attempts. First, the elevated risk in women who have had an abortion might be related to unmeasured mental health problems that are undiagnosed or untreated. Steinberg and colleagues 1 take pre-abortion mental health into account by controlling for pre-abortion psychiatric contact and psychiatric medications. The authors state that mental