Long-term outcomes following self-poisoning in adolescents: a population-based cohort study

Summary

Background

Suicide is the third most common cause of death among adolescents worldwide, and poisoning is the leading method of attempted suicide. Unlike more violent methods, survival after self-poisoning is common, providing an opportunity for secondary prevention. We determined the risk and time course of completed suicide after adolescent self-poisoning, and explored potential risk factors.

Methods

We did a population-based cohort study using multiple linked health-care databases in Ontario, Canada, from Jan 1, 2001, to Dec 31, 2012. We identified all adolescents aged 10–19 years presenting to hospital after a first self-poisoning episode. Each was matched with 50 population-based reference individuals with no such history, matching on age, sex, and year of cohort entry. The primary outcome was the risk of suicide after a first self-poisoning episode. Secondary analyses explored factors associated with suicide and self-poisoning repetition.

Findings

We identified 20 471 adolescents discharged from hospital after a first self-poisoning episode and 1 023 487 matched reference individuals. Over a median follow-up of 7·2 years (IQR 4·2–9·7), 248 (1%) adolescents discharged after self-poisoning died, 126 (51%) of whom died by suicide. The risk of suicide at 1 year after self-poisoning was greatly increased relative to reference individuals (hazard ratio [HR] 32·1, 95% CI 23·6–43·6), corresponding to a suicide rate of 89·6 (95% CI 75·2–106·7) per 100 000 person-years over the course of follow-up. The median time from hospital discharge to suicide was 3·0 years (IQR 1·1–5·3). Factors associated with suicide included recurrent self-poisoning (adjusted HR 3·5, 95% CI 2·4–5·0), male sex (2·5, 1·8–3·6) and psychiatric care in the preceding year (1·7, 1·1–2·5). Adolescents admitted to hospital for self-poisoning were also more likely to die from accidents (5·2, 4·1–6·6) and from all causes (3·9, 2·8–5·4) during follow-up.

Interpretation

Self-poisoning in adolescence is a strong predictor of suicide and premature death in the ensuing decade, and identifies a high-risk group for targeted secondary prevention. Suicide risk is increased for many years after the index hospital admission, emphasising the importance of sustained prevention efforts.

Funding

The Canadian Drug Safety and Effectiveness Research Network, Ontario Ministry of Health and Long-Term Care, Paediatric Consultants Partnership.

Introduction

Suicide is the third leading cause of death in adolescents worldwide,¹ and many cases might be preventable.² About half of American high school students report suicidal ideation, one in six has seriously contemplated suicide, and 8% have attempted suicide in the preceding year.3, 4 The WHO considers adolescent suicide a major global health problem, and has called for focused prevention efforts.2, 5 However, little progress has been made in this regard over the past 60 years, leading some experts to characterise this goal as elusive.⁶ Adolescent suicide and self-harm also elicit strong public sentiment and concern from policy makers and health-care providers.1, 7 Aside from the loss of life, adolescent suicide has devastating effects for the family, friends, and community of the deceased.⁸

Of the various suicide methods used by adolescents, self-poisoning is the most common.1, 9 Unlike more violent methods, such as suicide by firearm, hanging, or jumping from a height, which are often fatal,¹ survival after self-poisoning is very common, providing an important opportunity for secondary prevention. Other forms of self-harm in adolescents, such as self-cutting, typically have minor clinical effects and rarely come to the attention of medical services, and are, therefore, more challenging to study systematically.10, 11 The identification of adolescents at risk of suicide is challenging. A case-control study¹² reported that two-thirds of adolescents who completed suicide did not access medical care in the month preceding death, and 56% had no record of mental illness before death.¹² Importantly, self-poisoning might not indicate genuine suicidal intent, particularly in young people.10, 13

Research regarding outcomes following self-poisoning in adolescents is scarce, particularly with regard to the long-term risk of suicide.¹⁴ Previous studies rarely identify an inception cohort, and they tend to assess short follow-up periods or selected populations, such as adolescents with established psychiatric disorders. We examined long-term outcomes after a first episode of self-poisoning in a large population of adolescents. Our primary hypothesis was that adolescent self-poisoning would be accompanied by a clear and durable increase in the subsequent risk of suicide.

Research in context

Evidence before this study

We undertook a review of PubMed, Medline, Embase, and other standard research repositories to March 10, 2015, seeking English-language publications related to adolescent self-poisoning and suicide. We identified several studies pertaining to self-poisoning, recidivism, and suicide in adolescents. However, we found no published studies characterising long-term outcomes (repetition of self-poisoning or death by suicide) in a population of adolescents following a first self-poisoning episode. We concluded that little is known about long-term outcomes following a first self-poisoning in adolescence.

Added value of this study

With more than 20 000 cases, this study represents the largest evaluation of long-term outcomes after adolescent self-poisoning published to date. We showed that following a first self-poisoning episode, adolescents faced a suicide risk more than 30 times higher than that of peers drawn from the general population. We also found that the risk of premature death in this population lingers for years, as most suicides occurred several years after the first hospital visit for self-poisoning. We identified several risk factors for suicide, including recurrent self-poisoning, male sex, and previous psychiatric care (presumably a marker of more severe psychiatric disease). Finally, we noted that adolescents discharged from the emergency department without hospital admission remained at increased risk of suicide, underscoring the risk and importance of long-term interventions even in adolescents whose initial presentation was not sufficiently severe to warrant admission.

Implications of all the available evidence

Secondary prevention initiatives need to be sustained for longer than previously appreciated in the vulnerable adolescent population. Front-line clinicians should be cognizant of the durable risk of suicide in this population and ensure that an appropriate follow-up plan is in place for all adolescents discharged from hospital after self-poisoning.