Summary of main findings
This study has shown that the incidence of presentations for attempted hanging more than doubled between 2007 and 2019, while the incidence of presentations for attempted drowning increased by almost 50%. Notably, the incidence of presentations for both methods was highest in young people aged 15–24 years and in males. Males, people experiencing homelessness and those living in the capital city, Dublin, had the highest odds of presenting for attempted hanging, compared with presenting for all other self-harm methods. However, women, non-private household residents and those living in the capital were more likely to present with repeated episodes of attempted hanging, following an index presentation to hospital. Males and people experiencing homelessness had the highest odds of presenting for attempted drowning, while people experiencing homelessness had the highest odds of a repeated episode of attempted drowning following an index presentation to hospital.
Comparison of findings with existing literature
The findings of this study were largely consistent with the existing international literature. Firstly, males had higher odds than females of presenting to hospital after attempted hanging and attempted drowning. In studies from Europe, the USA and Australia, men have been found to be at greater risk of suicide by hanging than women [
7,
18,
19]. This finding has also been observed in older adults, and after adjusting for confounding factors [
18,
19]. Similarly, a systematic review of the global literature on suicide and self-harm by drowning observed that there were a small preponderance of males engaging in suicide by drowning, which is consistent with the findings of this study [
20].
An almost inverse relationship was observed between age and odds of hospital presentation for attempted hanging. This is in line with previous Irish and international studies which reported that victims of hanging suicides were younger than victims of suicide from other methods, and that younger age groups were at higher risk of intentional hanging than older age groups [
21,
22].
Similarly, studies from Europe and Australia indicate that older adults may have the highest risk of suicidal drowning compared with other age groups, and this study reported the same finding in relation to age and self-harm by drowning [
23,
24].
In keeping with the existing literature, this study found that area deprivation was associated with self-harm by hanging and drowning, with individuals living in the most deprived areas having the highest odds of presenting to hospital after an episode of attempted hanging or drowning. Observational studies conducted in Canada and the United States (US) have found similar positive associations between higher levels of deprivation and risk of suicide by hanging in both males and females and among adolescents [
25,
26]. No studies were identified, however, which explored the relationship between deprivation and risk of suicide or self-harm by drowning.
Correspondingly, this study found that individuals experiencing homelessness had considerably higher odds of attempted hanging and drowning, as well as repeated attempts, than individuals living in a private residence or other setting. This is largely consistent with another study performed in Ireland that observed a distinctly higher incidence of all methods of self-harm among homeless individuals compared with individuals with a fixed residence [
27]. Moreover, a study of people experiencing homelessness presenting to three EDs in the United Kingdom (UK) found that homeless individuals had a higher risk of self-harm repetition than domiciled individuals [
28].
Strengths and limitations
This study had a number of strengths. Firstly, this study availed of data from the NSHRI, one of the few national self-harm surveillance systems worldwide [
1]. Several countries have self-harm surveillance systems, however, many lack national coverage, or only collect information on self-harm that results in admission to an inpatient hospital ward [
29]. Secondly, the NSHRI has had complete coverage of the EDs in all acute hospitals in Ireland since 2006, which afforded this study a comprehensive and national perspective of attempted hanging and drowning in Ireland [
30].
Thirdly, the quality of the data collected by the NSHRI and analysed in this study was high. The NSHRI’s DROs are independently trained and apply the self-harm case definition and inclusion/exclusion criteria in a standardised, systematic manner. There are numerous quality control measures in place to ensure the validity of the data, including continual checks for consistency and accuracy and regular measurements of interrater reliability between DROs, which are high (e.g. Kappa statistic of 0.90 in 2017). Finally, while there is extensive evidence available about the factors associated with death by suicide by hanging and drowning, there is limited evidence on the factors associated with self-harm by hanging and drowning. This study provides valuable insights into the characteristics of individuals who present to hospital after highly lethal self-harm methods.
This study also had a number of limitations. Firstly, the NSHRI is a routine data source which collects a limited number of variables. The registry does not collect information on marital status, ethnicity, or history of mental illness in the individuals presenting to EDs after self-harm. These and other unmeasured variables may contribute towards the risk of attempted hanging and drowning and without their inclusion, there may have been residual confounding of the associations observed in this study [
18,
31,
32]. Due to a number of factors, including the lack of an individual health identifier in Ireland, limited and inconsistent use of diverse electronic health record systems across all levels of the health system, and general practitioners largely existing as private practitioners who are not paid by the public health service, it is unfeasible to link the NSHRI data to other patient records to capture these potential confounders. Secondly, the Registry does not record specific details about the level of lethality of acts of attempted hanging or drowning. Thirdly, the NSHRI only collects information about self-harm, and not death by suicide. Deaths by suicide are recorded by the CSO in Ireland and these data are not linked to the NSHRI [
33]. Therefore, the risk of death by suicide among those who self-harm by hanging and drowning cannot be readily calculated, nor can the case fatality rate of different methods of self-harm. It is possible that some of the individuals included in this dataset died by suicide during the study period and this may have biased the analyses assessing risk of repeated self-harm by hanging and drowning. For example, in this study, men were found to have lower odds of repeat hospital presentations for attempted hanging and drowning compared with women; while it is possible women were more likely than men to engage in repeated self-harm attempts, it is also possible that men were as, if not more, likely than women to engage in repeated episodes of self-harm by hanging and drowning, but their episodes of self-harm ultimately resulted in death by suicide, which were not captured by this study [
31,
34].
Fourthly, the quality of data collected on acute alcohol consumption in self-harm presentations is less than it is for other variables in the NSHRI. All other variables are routinely recorded in the patient notes or in hospital administrative records. Acute alcohol consumption is not systematically recorded in the ED patient notes and there may be misclassification of acute alcohol consumption status. Therefore, associations between acute alcohol consumption and attempted hanging and drowning should be interpreted with caution. Furthermore, while the NSHRI has complete coverage of all 33 acute hospitals in Ireland, attempted hanging and drowning incidents treated in primary care or on psychiatric inpatient wards, or which are untreated by the health service, are not captured by the NSHRI. This may underestimate the incidence of self-harm by hanging and drowning, or may bias the findings of the analyses exploring the factors associated with these methods of self-harm. Finally, the NSHRI does not collect information on where self-harm acts take place. This information could be very useful, as it could reveal public locations where intentional hanging or drowning repeatedly occur and these sites could be targeted for interventions to prevent suicide.
Implications
Firstly, this study demonstrates that the rate of hospital presentations for both hanging and drowning increased steadily between 2007 and 2019. This occurred against a backdrop of fluctuating rates of hospital presentations for all methods of self-harm, suggesting that attempted hanging and attempted drowning comprise a growing problem compared with all other self-harm methods. Secondly, certain subsets of the population may be at increased risk of self-harm by hanging and drowning and these findings may help to inform self-harm and suicide prevention strategies. As adolescents and young adults may be at greatest risk of self-harm by hanging compared with other age groups, primary prevention of suicidal behaviour through school-based interventions that raise awareness about suicide or increase connectedness among students may be effective at addressing some of the risk factors for self-harm [
35,
36]. Similarly, as homeless individuals appear to be at particularly high risk of intentional hanging and drowning and repetition of self-harm by these means, this population may need to be prioritised for effective therapeutic interventions for mental health conditions both before and after episodes of self-harm. Finally, previous research has shown that self-harm by hanging and drowning are associated with a higher risk of suicide than many other self-harm methods [
9]; consequently, individuals who present to hospital after engaging in these highly lethal methods of self-harm should receive appropriate care, including a specialist, biopsychosocial assessment to identify the drivers of distress leading to self-harm, and follow-up care with inpatient or outpatient mental health services [
37].
Self-harm and suicide by hanging and drowning are uniquely challenging methods of self-harm to prevent. Both methods can be rapidly lethal, reducing the probability of a life-saving intervention, and the means for carrying out either method are almost universally available [
31,
34]. Additionally, compared with other self-harm methods, individuals who engage in these two highly lethal methods of self-harm already have a higher risk of death by suicide [
9,
38]. Therefore, to prevent self-harm and suicide by hanging and drowning, the more distal determinants of self-harm and suicide, such as substance abuse, poor mental health, and a reduced sense of connectedness, must be addressed [
39,
40]. Additionally, the impact of exposure to harmful media coverage of suicide and self-harm must be addressed. Sensationalist, detailed and extensive reporting of suicide in the media is associated with increased suicide and exposure to this harmful coverage can contribute to suicide contagion and copycat suicides, especially among vulnerable populations, such as young adults [
41,
42]. Conversely, responsible media coverage that emphasises how individuals can overcome suicidal ideation and address mental health issues may encourage vulnerable individuals to seek help (“the Papageno effect”) [
43]. Therefore, efforts must be made to ensure that not only media depictions of suicide are in accordance with media guidelines for reporting suicide and social media platforms rigorously enforce their policies on content related to suicide, but that media reporting of suicides has a positive preventive effect where possible.
As mentioned previously, hanging and drowning are uniquely difficult methods of self-harm to prevent. As most hanging suicides likely occur in private settings [
31,
44], intervention by others is challenging. However, self-harm by drowning may occur in more public settings, where a limited number of interventions can be implemented acutely during incidents of self-harm. Voluntary organisations which provide suicide prevention patrols and monitor major waterways can be established to identify and provide assistance to individuals attempting to self-harm. However, strategies to avoid the presence of these organisations leading to greater public awareness about locations or ‘hotspots’ where intentional drowning repeatedly occurs would need to be considered. For example, closed-circuit television (CCTV) surveillance systems could allow volunteers to inconspicuously monitor a location and improving the lighting around waterways where intentional drowning occurs may be useful for not only discouraging individuals from self-harming there, but also increasing their chances of detection and intervention [
45]. Finally, consideration can be given to erecting signposts which provide a contact number for a crisis helpline, or installing emergency telephones that connect the caller directly with a crisis helpline, at locations where intentional drowning is known to occur repeatedly [
45].