Main findings

The number of individuals presenting with self-harm or ideation (i.e. thoughts of self-harm or suicide) had been increasing steadily for the 7 years pre-pandemic in Northern Ireland. Predictive modelling shows that the number of individuals presenting with self-harm or ideation fell during the first two to three months of the COVID-19 pandemic but was generally back within the expected range (as calculated from trends in the 7 years prior to the COVID-19 pandemic/restrictions) by June 2020. This is similar to data reported in a smaller study in England, which was limited in its comparison data, comparing numbers of presentations in 2019 with 2020 only.^{Reference Hawton, Casey, Bale, Brand, Ness and Waters8} A drop in secondary care service utilisation for other conditions during the first few months of the COVID-19 pandemic/restrictions has been evidenced in a range of studies.^{Reference Maguire, Kent, O'Neill, O'Hagan and O'Reilly14–Reference Steeg, Carr, Trefan, Ashcroft, Kapur and Nielsen17} However, a Welsh study examining both primary care and secondary care contacts for self-harm during the pandemic found primary care contacts with self-harm reduced disproportionately compared with non-self-harm contacts, suggesting a reduction in help-seeking behaviour within this cohort, which may be reflected in the decrease in presentations observed in this study.^{Reference DelPozo-Banos, Lee, Friedmann, Akbari, Torabi and Lloyd7}

However, after stratification, there were differences observed in the demographic profile of those presenting with self-harm or ideation, with slightly fewer men than women presenting, presentations remaining unaffected in individuals aged over 65 years and those who live alone or in affluent areas and presentations appearing to increase in those aged under 16 years. Groups that saw little change had small numbers of presentations. The expected tsunami of mental ill health is not evidenced in an increase in hospital-presenting self-harm or ideation. It is not known whether these differences will persist into long-term effects of the pandemic/restrictions.

Interpretation

In five out of the first seven months of the pandemic, male presentations to emergency departments with self-harm or ideation were lower than expected. This may be reflective of the increased ‘shielding’ among men, who are known to be at a higher risk of poor COVID-19 outcomes compared with women.^{Reference Baker, White and Morgan18} A US study found that individuals who overestimated the risk of COVID-19 were more likely to practise household isolation.^{Reference Ladapo, Rothwell and Ramirez19} Another explanation may be that more men who previously would have presented with self-harm or ideation died on their first suicide attempt, although this does not appear to be reflected in national statistics on death by suicide.^{Reference Appleby, Richards, Ibrahim, Turnbull, Rodway and Kapur20} The full extent of the impact of the pandemic on suicide is not yet fully understood owing to delays in suicide reporting and coronial investigations.^{Reference John, Eyles, Webb, Okolie, Schmidt and Arensman4} Lastly, the pandemic may have had a positive effect on men's mental health, at least during the initial seven months, where the collective nature of the stressor, coupled with increased time with family, and ‘furlough’ schemes that provided financial support for those who were unable to work, may have temporarily addressed the stressors associated with suicidal thoughts and supported good mental health.^{Reference Zortea, Brenna, Joyce, McClelland, Tippett and Tran21,Reference Wels, Booth, Wielgoszewska, Green, Di Gessa and Huggins22}

Age differences in the presentation profile of self-harm or ideation were also observed, with an increasing trend in the number of individuals presenting aged under 16 years. This supports reports from UK doctors at the time suggesting that more young people presented to emergency departments with self-harm.²³ A recent systematic review has suggested that the COVID-19 pandemic has had a particularly negative effect on child and adolescent mental health, with increased anxiety, depression and loneliness.^{Reference Panchal, Salazar de Pablo, Franco, Moreno, Parellada and Arango24} During the first seven months of the pandemic covered in this study, schools in Northern Ireland remained closed and children were educated using a mixture of home-schooling and distance learning, with schools and teachers utilising video call technology and remote learning applications. This separation from their peers and disruption to usual structure may have had a negative impact on their mental health. This may have become manifest in self-harm or ideation. In addition, children with additional needs, such as those with neuropsychiatric and neurodevelopmental disorders, may have found it difficult to cope with the initial pandemic/restrictions, which led to school closures, a reduction in usual care and reduced social care support, and this may have resulted in an increase in self-harm. The pandemic and associated restrictions have also been shown to have been particularly difficult for young people with existing mental health conditions.^{Reference Guessoum, Lachal, Radjack, Carretier, Minassian and Benoit25} This cohort may require additional attention post-pandemic and this should be taken into consideration when planning mental health services.^{Reference Moreno, Wykes, Galderisi, Nordentoft, Crossley and Jones26} Differences by method of self-harm are not overwhelmingly significant, although the ARIMA models are suggestive of a larger decrease in presentations involving ‘overdose’. This is supported by the small English study, which suggests a larger decrease in presentations involving ‘self-poisoning’ compared with other methods.^{Reference Hawton, Casey, Bale, Brand, Ness and Waters8}

Strengths and limitations

This study utilised eight years of data from the NIRSH, the world's only population-wide registry of self-harm and ideation, linked to individual-level health card registration data, to examine the impact of the COVID-19 pandemic and associated restrictions on hospital-presenting self-harm and self-harm/suicide ideation in Northern Ireland. The use of individual-level data allowed for an accurate analysis of numbers of individuals presenting to emergency departments across the country each month. The NIRSH uses standardised data collection practices and covers all emergency departments in Northern Ireland, providing a verified rich source information on presentations for self-harm and ideation at a population level. The information collected also facilitated the investigation of the impact of the pandemic/restrictions on subgroups of the population, and thus identification of populations at risk, such as males and children under 16 years, that could not be perceived from the full cohort analysis. Analysis was limited to the data included within the NIRSH, which is only those individuals who presented to an emergency department. Many individuals who self-harm do not present to secondary care. Therefore, the decline in presentations in the first few months of the pandemic is not necessarily reflective of a decrease in self-harm or ideation, but may be reflective of a decrease in service utilisation during that time. The use of longitudinal data and time series analysis (ARIMA) allowed for consideration of long-term trends and seasonality in self-harm behaviours. This design provides more reliable evidence of a change in expected patterns than simpler before-and-after comparisons, which may neglect to account for the pre-existing increase in many indicators for mental ill health.^{Reference Hawton, Casey, Bale, Brand, Ness and Waters8}