Official Statistics

Suicide in people with severe mental health problems: report

Published 2 September 2026

Applies to England

This report contains sensitive content which refers to details on deaths by suicide.

If you are struggling to cope, please call the Samaritans for free on 116 123 (UK and the Republic of Ireland) or contact other sources of support, such as those listed on the NHS help for suicidal thoughts webpage. Support is available around the clock, every day of the year, providing a safe place for you, whoever you are and however you are feeling.

Any onward reporting of this data should consider following the Samaritans’ media guidelines on the reporting of suicide because of the potentially damaging consequences of irresponsible reporting. In particular, the guidelines advise on terminology and include links to sources of support for anyone affected by the themes in this report.

Introduction

This report presents statistics on deaths by suicide in people with severe mental health problems in England. This group is defined as people aged 18 to 74 who had been referred to secondary mental health services in the 5 years before they died. In this report, this group is described as the ‘mental health (MH) population’. Some of these people had only been referred to services, but a larger proportion had contact with community or outpatient services, or had stayed in hospital as an inpatient.

These statistics also compare suicide in the MH population to suicide in all other people aged 18 to 74, who are described as the ‘non-mental health (non-MH) population’.

Suicide has been added to the existing mortality statistics on the MH population aged 18 to 74. This adds intelligence to support understanding of mortality in this group and informs suicide prevention planning. Given the complexity of the new suicide statistics, this report provides context that aids interpretation.

The report also includes information on use of secondary mental health services in the year before death. It highlights some of the differences in service experience between those who die by suicide and those whose death is due to another cause.

Death by suicide results from a complex interaction of biological, psychological, social and environmental factors^{[footnote 1]}. Evidence suggests that psychosocial factors, including trauma, adverse childhood experiences or social adversity^{[footnote 2]}, severe loneliness^{[footnote 3]}, poor physical health^{[footnote 4]} and substance misuse^{[footnote 5]}, can increase risk of both suicide and mental health problems.

People with mental health conditions are at a higher risk of suicide than those without^{[footnote 6]}. That risk varies by mental health condition with some less common conditions having a higher risk^{[footnote 7]}.

Services provide important support to people with mental health problems and those who are at an increased risk of suicide. Improved provision of mental health services and new initiatives aiming at suicide prevention have been associated with reduction in suicide rates ^{[footnote 8]}. 

The suicide prevention strategy for England: 2023 to 2028 sets out the ambition to reduce suicide rates. It identifies people in contact with mental health services as a priority group for targeted action and highlights the need to strengthen data and evidence. The statistics in this report can support the strategy by providing a baseline for monitoring and evaluation.

To support identification of opportunities for suicide prevention, this report and the accompanying data tables:

• describe trends and variation in suicide in the MH population
• compare suicide in the MH population and the non-MH population
• describe mental health service use before death in the MH population

This report is intended for:

• national organisations responsible for suicide prevention strategy, policy and guidance
• local organisations that plan, commission, deliver or evaluate suicide prevention and mental health services

This report is about people aged 18 to 74 because the statistics are part of a wider indicator set on premature mortality in people with severe mental health problems. Suicide is an avoidable cause of death, and suicide prevention is relevant across the life course.

This analysis includes all suicides where there was evidence of a referral to secondary mental health services. It does not include everyone who died by suicide and experienced a mental health problem before death. This is because some people may have accessed support from other mental health services, including NHS Talking Therapies, primary care or voluntary sector services that do not submit to the mental health services data set (MHSDS).

This report and statistics have been jointly produced by the Office for Health Improvement and Disparities (OHID), part of the Department of Health and Social Care (DHSC), and NHS England.

NHS England publishes data on excess mortality for people referred to secondary mental health services (aged 18 to 74).

Summary findings

This report considers suicide as a cause of death in people experiencing severe mental health problems. It shows that between 2022 and 2024, there was an annual average of 2,967 registered suicides in the MH population. This is nearly 60% of all suicides in England in people aged 18 to 74. The proportion was higher for women (71.0%) than men (54.9%). This analysis highlights that around 40% of people aged 18 to 74 who died by suicide were not in the MH population.

Suicide accounts for around 7% of deaths in the MH population. By contrast, more than half of the deaths were due to a combination of cancer, cardiovascular disease, liver disease and respiratory disease.

For suicides registered between 2022 and 2024:

• the rate was 83.4 per 100,000 MH population - it was 2.6 times higher for men (126.2 per 100,000) than women (48.7 per 100,000)
• the MH population was 14.6 times as likely to die by suicide than the non-MH population - this was significantly higher for women (24.0) than men (13.4)
• the difference in mortality between the MH and non-MH populations for suicide (14.6 times) was significantly higher than that reported for cancer (2.3), cardiovascular disease (3.8), respiratory disease (6.1) and liver disease (6.4)

Between 2015 to 2017 and 2022 to 2024, the number of suicides registered in the MH population increased by 43.4% (40.7% for women and 44.6% for men). However, much of this increase was due to the MH population growing by 45.6% over the same period (48.0% for women and 37.0% for men).

When interpreting trends, it is important to consider growth in the MH population together with the change in the standard of proof for suicide and increasing delay in death registration.

For registrations between 2015 to 2017 and 2022 to 2024 in England, the suicide rate in the MH population increased significantly for persons by 6.0% and for men by 11.4%. There was no statistically significant change for women.

For deaths registered between 2022 to 2024, regional suicide rates in the MH population were highest in the South West for persons (105.6 per 100,000) and women (63.0 per 100,000), and in the North East for men (157.8 per 100,000). The lowest rates were in London for persons (55.5 per 100,000), women (33.6 per 100,000) and men (81.9 per 100,000).

For England, women were more likely than men to have a referral (with or without contact or inpatient stay) to secondary MH services. This highlights that men, although at higher risk of suicide, were less likely to access services.

Analysis in this report shows that around 80% of the MH population who died by suicide had an open referral in the year before death and around 14% of this group had no contact or hospital spell. Crisis or acute services were the most recorded service type for this group, and this service type was more common than in those who died by other causes of death. The proportions of face-to-face contacts and hospital spells were also higher in those who died by suicide than other causes of death.

About these statistics

To enhance the use of these statistics, some important terms and aspects of the analysis are explained here. Further details are available in the ‘Data sources and methods’ section.

The report presents the latest data available, for 2022 to 2024, and trends starting from 2015 to 2017. Data is available for England and regions only.

Study population

This report includes analysis of people aged 18 to 74 who died by suicide, using year of registration. Suicide deaths recorded in the Office for National Statistics (ONS) mortality extracts were linked to the MHSDS. This identified whether people had an open referral to secondary mental health services in the 5 years before death. Some of these people had only been referred to services, but a larger proportion had contact with community or outpatient services, or had stayed in hospital as an inpatient.

Deaths were grouped into 2 populations. A death by suicide was included in the:

• MH population if there was evidence of an open referral to secondary mental health services in the 5 years before death, with or without recorded contact or an inpatient stay
• non-MH population if there was no evidence of an open referral to secondary mental health services in the 5 years before death

Please note that this report uses people with an open referral to secondary mental health services as a proxy for ‘people with a severe mental health problem’.

Statistics and their presentation

Deaths by suicide in the 2 groups were used to calculate population-based statistics for England and the 9 statistical regions. These statistics are:

• directly standardised rates (DSRs) - which is a rate adjusted to account for differences in the age structure of populations to allow suicide rates to be compared between geographical areas and across time periods
• mortality rate ratios - which measure the difference in DSRs between suicide in the MH and non-MH populations. This is presented as ‘times as likely’ and referred to as ‘excess’ in charts

The denominator populations were estimated using MHSDS and ONS mid-year population estimates. For the MH population, the denominator was calculated as the number of people referred to secondary mental health services in the financial year corresponding to the year of death registration or the 4 previous. This was expressed as ‘per 100,000 MH population’. For the non-MH population, the ONS mid-year population estimates for that year and the 4 previous were used, minus the estimated MH populations. This was expressed as ‘per 100,000 non-MH population’.

For comparison purposes, suicide in the MH population is also presented as a rate in all people aged 18 to 74 (described as the ‘general population’). ONS mid-year estimates were used, and rates expressed as ‘per 100,000 general population’.

Suicide statistics in this report use 3-year combined deaths and populations, which are shown as an annual average to make the statistics clearer.

Differences are indicated as statistically significant if the 95% confidence intervals (CIs) between values do not overlap. They are indicated as error bars (vertical lines) on each chart.

This report also presents analysis on service use prior to death by suicide, based on date of occurrence and is presented as counts or proportions.

ONS mortality extracts record sex, while MHSDS records gender. Breakdowns of suicide deaths are therefore presented by sex. Estimated MH population breakdowns are presented by gender.

Date of occurrence and registration for deaths

Date of death is used to measure the time between referral, contact or inpatient stay and death.

Date of registration is used for presentation of DSRs and statistics on differences in mortality between MH and non-MH population (‘excess’).

Date of death is used in the analysis presented in the ‘MH service use in the year before suicide’ section.

Related sources of intelligence

These statistics improve understanding of mortality in the MH population and work alongside the following related intelligence.

The DSRs and statistics on differences in mortality between MH and non-MH population for suicide come from the same source as a group of indicators published in the Adult mental health and wellbeing profile. The profile includes data on all deaths and 4 major physical health causes of death (cancer, cardiovascular disease, liver disease and respiratory disease) for the MH population. Adding suicide to this indicator set supports:

• suicide prevention planning
• design and delivery of preventive, clinical and support services to reduce early mortality experienced by people with mental health problems

This suicide analysis complements the national confidential inquiry into suicide and safety in mental health (NCISH), which reports detailed clinical data on people who died by suicide while under the care of mental health services up to one year before death. These new statistics sit alongside but do not replace NCISH reporting, which remains an important source of intelligence on this topic. The new analysis includes data on:

• suicide and service use in people in England who were referred to mental health services up to 5 years before their death
• suicide rates and statistics on differences in mortality between MH and non-MH population for English regions

Wider context to these statistics

To support the use and interpretation of statistics in this report, consider the following important factors.

Suicide in the general population: impact of changes in standard of proof and delay in registrations

ONS data based on year of registration shows an increase in suicide rates in England over the period of this report. However, this should be interpreted alongside 2 important factors:

1. In July 2018, the standard of proof for suicide conclusions changed from the criminal standard (‘beyond reasonable doubt’) to the civil standard (‘the balance of probabilities’). ONS analysis found that this change was likely to increase the number of deaths registered as suicide.
2. The COVID-19 pandemic affected coronial services. One impact of this is lengthening of registration delays post pandemic (from 2020). For suicides registered in 2024 39% occurred in the same year, reducing from 53% of suicides being registered and occurring in 2018. There are differences in registration delays by region.

Mental health service population: change over time

As evidenced by NHS England data, the MH population increased between 2015 and 2024. The increase may reflect:

• changes in the level of mental health need and demand
• clinical characteristics of population accessing services
• referral patterns
• service availability
• MHSDS coverage, as more providers submit data

At England level (see figure 1), the MH population increased by 45.6% between 2015 to 2017 and 2022 to 2024, from around 2.76 million to 4.02 million per year. The increase was 48.0% for women, from 1.46 million to 2.17 million, and 37.0% for men, from 1.29 million to 1.77 million. In 2022 to 2024 around 10% of people aged 18 to 74 were in the MH population.

Figure 1: estimated MH population aged 18 to 74 average per year by gender - England, 2015 to 2017, 2019 to 2021 and 2022 to 2024

Gender	2015 to 2017	2019 to 2021	2022 to 2024
Persons	2,758,172	3,455,509	4,016,518
Women	1,464,883	1,828,462	2,168,669
Men	1,288,478	1,604,028	1,765,454

Data source: MHSDS and predecessor data sets.

The size of the MH population varies by region (see figure 2). For the period 2022 to 2024 it ranged from 201,136 in the North East up to 588,401 in London. Between 2015 to 2017 and 2022 to 2024, the largest percentage increase was in the West Midlands (56.4%), followed by London (49.6%) and Yorkshire and the Humber (49.1%). The smallest increase was in the North West (28.2%).

Figure 2: estimated MH population aged 18 to 74 average per year by region - 2015 to 2017, 2019 to 2021 and 2022 to 2024

Region	2015 to 2017	2019 to 2021	2022 to 2024
North East	140,108	178,678	201,136
North West	441,551	525,519	566,184
Yorkshire and the Humber	276,786	355,999	412,770
East Midlands	231,846	278,918	334,414
West Midlands	296,619	388,956	463,952
East of England	292,993	363,494	401,716
London	393,321	497,922	588,401
South East	379,438	487,303	561,331
South West	250,783	297,269	336,320

Data source: MHSDS and predecessor data sets.

Full data on England’s and regional populations is available from tables 1 and 2 in the accompanying data tables on the Suicide in people with severe mental health problems page.

Suicide rates in the MH population

This section presents suicide in the MH population using year of registration. This is consistent with the wider indicator set on mortality and differences in mortality in this population.

Between 2022 and 2024, suicide accounted for 6.9% of all deaths in the MH population, whereas cancer (18.8%), cardiovascular disease (18.7%), respiratory disease (10.9%), and liver disease (8.1%) together accounted for nearly 60%.

Interpretation of suicides rates and regional comparison should consider:

• the change in the standard of proof for DSRs from July 2018. Use of 3-years combined may make this effect less visible
• registration delays, as the statistics are based on year of registration and:
  • recent trends should be interpreted with caution
  • regional differences in delays may impact on interpretation of rates
• the increase in the population accessing secondary mental health services as it affects:
  • population estimates for both the MH and non-MH populations
  • the number of suicides assigned to the MH and non-MH populations, so changes in the number of suicides alone do not necessarily indicate a change in risk

To aid interpretation, trend charts present 3 time periods rather than the full series. These are:

• 2015 to 2017 - start of the series and before the change in standard of proof
• 2019 to 2021 - the first period after the change in standard of proof
• 2022 to 2024 - the latest point in the time series

Further data is available from tables 3 to 6 in the accompanying data tables on the Suicide in people with severe mental health problems page. This includes full time series for number of deaths and rates for England and regions.

Proportion of suicides in the MH population

In England, there was an annual average of 5,552 deaths by suicide between 2022 and 2024, equivalent to 10.9 deaths per 100,000 general population. 5,027 of these deaths were among people aged 18 to 74, around 90%. In this age group, the suicide rate was 12.4 deaths per 100,000 general population. Of these, an annual average of 2,967 deaths were in the MH population, equivalent to 7.3 deaths per 100,000 general population. This means that 59.0% of suicides among people aged 18 to 74 occurred in the MH population, although this group accounts for about 10% of the general population. The proportion was higher for women (71.0%) than for men (54.9%) (see table 1).

Table 1: suicide rates (per 100,000 general population) by population group, and percentages of all people and all suicides in the MH population, in England by gender - ages 18 to 74, deaths registered 2022 to 2024 

Sex	DSR general population	DSR MH population	Percentage of general population that is MH population	Percentage of all suicides in MH population
Persons	12.4	7.3	9.9	59.0
Women	6.2	4.4	10.5	71.0
Men	18.9	10.4	8.9	54.9

Data source: ONS annual mortality extracts; ONS mortality extracts linked to MHSDS and predecessor data sets.

Note: DSRs for the MH population uses the general population as the denominator to allow direct comparison with the DSR for the general population.

The percentage of all suicides for people aged 18 to 74 that were in the MH population changed over the course of the reporting period. For:

• 2015 to 2017 it was 2,069 out of 4,176 suicides per year, 49.5%
• 2019 to 2021 it was 2,524 out of 4,684 suicides per year, 53.9%
• 2022 to 2024 it was 2,967 out of 5,027 suicides per year, 59.0%

The increase in the MH population over the reporting period, and therefore the proportion of suicides identified in this group, impacts on the above percentages.

For the 2022 to 2024 period, the regional variation in suicide rate for the MH population was consistent with people aged 18 to 74 overall (see table 2). The overall rates were highest in the North East with 16.6 per 100,000 general population and lowest in London with 8.2. For the MH population the figures were 10.1 per 100,000 general population in the North East and 4.6 in London.

At regional level, the proportion of all suicides in the MH population were:

• highest in the North West (62.6%) followed by the North East (60.6%) with each region showing the highest:
  • proportion of the general population aged 18 to 74 in the MH population (10.6% and 10.5% respectively)
  • overall suicide rate per 100,000 general population (15.8 and 16.6 respectively)
• lowest in London (55.5%) followed by East of England (56.3%) with:
  • each region showing 9.0% of the general population aged 18 to 74 in the MH population
  • London having the lowest overall suicide rate per 100,000 general population (8.2)

Regional variation in suicide rates in the MH population is likely to be influenced by the level and differences in MH need and access to secondary MH services.

Table 2: suicide rates (per 100,000 general population) by population group, and percentages of all people and all suicides in the MH population, in England by region - ages 18 to 74, deaths registered between 2022 to 2024

Region	DSR general population	DSR MH population	Percentage of general population that is MH population	Percentage of all suicides in MH population
North East	16.6	10.1	10.5	60.6
North West	15.8	9.9	10.6	62.6
Yorkshire and the Humber	14.6	8.4	10.5	57.4
East Midlands	13.3	7.8	9.6	58.6
West Midlands	12.0	7.2	11.1	59.6
East of England	11.3	6.4	9.0	56.3
London	8.2	4.6	9.0	55.5
South East	11.7	7.0	8.6	59.8
South West	13.3	7.7	8.3	58.2

Data source: ONS annual mortality extracts; ONS mortality extracts linked to MHSDS and predecessor data sets.

Note: DSRs for the MH population uses the general population as the denominator to allow direct comparison with the DSR for the general population.

Comparison of the proportion of suicides in the MH population with NCISH

The proportion of overall suicides assigned to the MH population in this analysis is higher than in the NCISH annual report. This mainly reflects differences in data sources, population definitions, time period and inclusion criteria (see table 3).

The ‘MH services use in the year before suicide’ section provides further comparison with the NCISH report and highlights where analyses differ or are similar.

Table 3: comparison of methodology used for analysis in this report and NCISH annual report

Specification	OHID	NCISH analysis
Population coverage	People aged 18 to 74	People aged 11 and over
Geographical coverage	England	UK and Jersey
Service coverage	All providers submitting to MHSDS or predecessor data sets	NHS Mental Health Trusts or other providers close to place of residence or death
Time basis	Year of registration	Year of occurrence
Population identification	Routine linkage	Electronic questionnaires returns completed by providers
Evidence of being in the MH population	Open referral to secondary MH services (with or without a contact or an inpatient stay)	MH service care contact
Retrospective follow up period from date of death	5 years	1 year
Important exclusions	People with learning disability and autism activity only	Psychiatric liaison services

Note: full list of providers submitting to MHSDS is available in the ‘Data sources and methods’ section.

Variation and change over time

This section presents rates using the MH population as the denominator, because changes in the size and composition of this population affect interpretation.

Suicide rates by sex

Figure 3 shows that between 2022 and 2024 the suicide rate per 100,000 MH population in England was:

• 83.4 for person
• 48.7 for women
• 126.2 for men

Within the MH population, the suicide rate for men was 2.6 times the rate for women, compared with around 3 times in the general population.

The overall rate for the MH population was higher than the NCISH annual report estimate, likely reflecting the methodological differences.

Figure 3: DSRs for suicide in the MH population for persons and by sex - England, deaths registered between 2022 to 2024

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets. 

Suicide rates by region

Regional patterns in suicide rates in the MH population (see figure 4) were broadly similar to suicide rates in the general population (see table 2), with the exception of the South West. Between 2022 and 2024, rates were:

• highest in the South West for persons and women
• highest in the North East for men
• lowest in London for persons, women and men

Regional comparisons should be interpreted with caution because denominator populations, their composition and suicide registration delay differ by region.

Figure 4: DSRs for suicide in the MH population for persons and by sex - regions, deaths registered between 2022 to 2024

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

Number and rates of suicides over time

The number of deaths by suicide in the MH population increased over time, while rates increased by a smaller amount. Figures 5 and 6 show:

• between 2015 to 2017 and 2022 to 2024:
  • the number of suicides increased by 43.4% for persons, 40.7% for women and 44.6% for men
  • rates of suicide increased by 6.0% for persons, 1.9% for women and 11.4% for men
• between 2019 to 2021 and 2022 to 2024:
  • the number of suicides increased by 17.5% for persons, 17.9% for women and 17.4% for men
  • rates of suicide increased by 5.0% for persons, 4.3% for women and 8.7% for men

The smaller increase in suicide rates than counts suggests interpretation should consider changes in the MH population and service use. The MH population increased by:

• 45.6% for persons, 48.0% for women and 37.0% for men between 2015 to 2017 and 2022 to 2024
• 16.2% for persons, 18.6% for women and 10.1% for men between 2019 to 2021 and 2022 to 2024

For men, the percentage increase in suicide exceeded the increase in the MH population, suggesting possible changes in risk, and frequency and type of service used.

Increase in suicide rates was also influenced by the 2018 change in the standard of proof and registration delay.

Figure 5: change (%) in the number of suicides in the MH population - England, between 2015 to 2017 and 2022 to 2024, and 2019 to 2021 and 2022 to 2024

Gender	2015 to 2017 and 2022 to 2024	2019 to 2021 and 2022 to 2024
Persons	43.4	17.5
Women	40.7	17.9
Men	44.6	17.4

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

Figure 6: change (%) in suicides rates in the MH population - England, between 2015 to 2017 and 2022 to 2024, and between 2019 to 2021 and 2022 to 2024

Gender	2015 to 2017 and 2022 to 2024	2019 to 2021 and 2022 to 2024
Persons	6.0	5.0
Women	1.9	4.3
Men	11.4	8.7

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

Suicide rates in all people aged 18 to 74 increased by 16.8% between 2015 to 2017 and 2022 to 2024, and by 5.6% between 2019 to 2021 and 2022 to 2024. This rate increase was higher than in the MH population. Although this report has not examined reasons for these differences, changes in denominator populations are important when interpreting rates. Between 2015 to 2017 and 2022 to 2024, the general population aged 18 to 74 increased by 3.4%, whereas in the MH population aged 18 to 74 the increase was 45.6%.

Figure 7 shows a statistically significant increases in suicide rates for persons and men in the MH population. This differs from the NCISH report, which reported a fall in the suicide rate among patients under mental health care. The difference is likely to reflect methodological differences, including population coverage and use of year of occurrence rather than registration. Growth in MH service population may also contribute.

Figure 7: DSRs for suicide in the MH population for persons and by sex - England, deaths registered 2015 to 2017, 2019 to 2021 and 2022 to 2024

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

Comparing mortality due to suicide in the MH and non-MH populations

This section compares suicide rates in the MH and non-MH populations, based on year of suicide registration. Changes in the populations and in the number of suicides assigned to each should be considered when interpreting these statistics.

Further data to the statistics presented in this section is available from tables 7 to 9 in the accompanying data tables on the Suicide in people with severe mental health problems page. This includes full time series and data on other causes of death for England and regions.

Difference in suicide rates between MH and non-MH populations by sex

Between 2022 and 2024, people in the MH population were 14.6 times as likely to die by suicide as those in the non-MH population (see figure 8). The difference was greater for women (24.0 times) than men (13.4 times).

Figure 8: difference in mortality (excess) due to suicide in the MH and non-MH population for persons and by sex - England, deaths registered between 2022 to 2024

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

Difference in suicide rates between MH and non-MH populations by causes of death

Although suicide accounted for a smaller proportion of deaths than major physical health causes, it showed the highest difference in mortality (14.6 as likely) between the MH and non-MH populations. Between 2022 and 2024, people in the MH population compared to the non-MH population were:

• 2.3 times as likely to die from cancer
• 3.8 times as likely to die from cardiovascular disease
• 6.4 times as likely to die from liver disease
• 6.1 times as likely to die from respiratory disease

Difference in suicide rates between MH and non-MH populations by region

In all regions, people in the MH population were more likely to die by suicide than those in the non-MH population (see figure 9). Between 2022 and 2024, the difference was:

• highest in the South East for persons, women and men - 18.4 times as likely for persons, 29.2 times for women and 16.7 times for men
• lowest for persons and men in Yorkshire and the Humber - 12.7 times as likely for persons and 12.5 times for men
• lowest for women in the North East - 18.0 times as likely

In 8 of the 9 regions, the difference in mortality between the MH and non-MH populations was significantly higher for women than for persons and men.

Regional differences in the mortality are influenced by variation in the size of the MH and non-MH population, and the suicide rate in the general population. Factors that influence this are explored in the premature mortality in adults with severe mental illness report.

Figure 9: difference in mortality (excess) due to suicide in the MH and non-MH population for persons and by sex - regions, deaths registered between 2022 to 2024

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

Difference in suicide rates between MH and non-MH populations over time

Figure 10 shows that across the reporting period, a statistically significant increase was identified for persons (9.0%) and men (16.5%). Although the difference in mortality due to suicide was consistently higher for women than for men and persons, no statistically significant change was identified.

No region showed a statistically significant change over the period.

Figure 10: difference in mortality (excess) due to suicide in the MH and non-MH population for persons and by sex - England, deaths registered 2015 to 2017, 2019 to 2021 and 2022 to 2024

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

MH services use in the year before suicide

This section describes MH service use in the year before death by suicide. It also contains comparisons between the MH population who died by suicide and who died by other causes.

Service use is defined as an open referral, with or without service contact, including inpatient stay. The term ‘open referral’ is used throughout this section. MH hospital spells within this analysis are a single continuous period of inpatient care, starting from the moment of admission to the final discharge - this will include people on home leave during their spell.

Around 80% of the MH population who had been referred to secondary mental health services in the 5 years before they died by suicide had an open referral in the year before death. This compares with around 60% of those who died from other causes (see table 4). Recent years may be less complete because of registration delay.

Table 4: number of suicide and not suicide deaths in the MH population with open referral in 5 years and 1 year before death - England, deaths occurring in 2018 to 2023

Year of death	Suicide: deaths (referral in 5 years)	Suicide: deaths (referral in 1 year)	Suicide: percentage of deaths (referral in 1 year)	Not suicide: deaths (referral in 5 years)	Not suicide: deaths (referral in 1 year)	Not suicide: percentage of deaths (referral in 1 year)
2018	2,655	2,205	83.1%	35,919	23,638	65.8%
2019	2,671	2,172	81.3%	36,407	23,856	65.5%
2020	2,742	2,217	80.9%	42,061	26,538	63.1%
2021	2,855	2,290	80.2%	42,143	26,143	62.0%
2022	2,894	2,320	80.2%	40,994	25,720	62.7%
2023	2,835	2,259	79.7%	39,603	25,102	63.4%

Data source: ONS mortality extracts linked to MHSDS and predecessor data sets.

The remaining analysis in this section focuses on people in the MH population with an open referral in the year before death (rather than the full 5 years). This supports a more direct comparison with other studies on suicide among the MH population.

When interpreting findings in this section, consider that:

• analyses are based on year of death rather than year of registration and therefore the number of suicides differs from earlier sections in the report
• using year of death reduces the effect of registration delays and supports clearer interpretation of recent patterns and service use - results are presented up to 2023 rather than 2024
• all tables in this section relate to the MH population only
• where relevant, findings are compared within the MH population to people who died from causes other than suicide
• where proportions are calculated, the denominator is the number of people in the MH population who died by suicide and had an open referral in the year before death - for example, in 2023 this was 2,259 rather than 2,835
• people may have more than one referral (including to different service types) and more than one contact within a referral - as percentages are based on unique people in the MH population, they may add up to more than 100%
• lower counts and proportions in 2020 and 2021 may reflect changes in access to and delivery of services during the COVID-19 pandemic and should be interpreted with caution

Full data supporting this section is available from tables 10 to 19 in the accompanying data tables on the Suicide in people with severe mental health problems page. This includes full time series and full data on not suicide deaths in the MH population.

Open referrals with no contact or hospital spell

In 2023, 309 people who died by suicide had no attended contact or mental health hospital spell recorded across referrals open in the year before death. This was 13.7% of people with an open referral in 2023, compared with 14.5% in 2018.

Of those 309 people with no attended contact or hospital spell, 202 had been discharged from all open referrals before death. This means that 107 people were not discharged but were not seen by services in the year before death.

This proportion who had no attended contact and no hospital spell in the year before death was higher among those who died from causes other than suicide (19.7% in 2023, down from 21.3% in 2018).

Patients were assumed to have been discharged if a provider did not submit their referral in the month of death, but their referral had been submitted in the preceding reporting period.

Having no attended contact or mental health hospital spell should not be interpreted as no mental health support. People may have used services in the years prior to the year before death or accessed other forms of support not captured here.

This analysis does not adjust for how long a referral was open within the year. This will differ between people.

Open referrals by provider type

Most of the MH population who died by suicide had an open referral to an NHS trust in the year before death (95.9% in 2023). Referrals to independent sector providers increased from 3.8% in 2018 to 6.2% in 2023, which may partially reflect improved reporting to MHSDS.

Open referrals by service group and type

Crisis or acute services were the most recorded service group, followed by community and general services. People who died by suicide often had referrals to more than one team or service type within and across groups. The crisis or acute service group was also commonly used by people in the MH population who died from causes other than suicide, but less frequently.

Within the crisis or acute services group, crisis resolution or home treatment services and psychiatric liaison services were the 2 most recorded service types. This is broadly consistent with the NCISH annual report. Although activity in psychiatric liaison services are not included in the figures, the report notes that mental health crises and people at risk of suicide are often managed in emergency departments, psychiatric liaison services and crisis resolution or home treatment services. For people in the MH population who died from causes other than suicide, psychiatric liaison services were also the most commonly recorded service type, while single point of access was the second most common.

Forensic, health and justice services were the third most common service type used by the MH population who died by suicide.

Attended contacts by consultation method

In the MH population who died by suicide, around 80% had an attended contact that was face-to-face. This stayed relatively constant between 2018 and 2023. Over the period the proportion who had a telephone contact increased from 44.8% to 63.4%.

Of the patients who died from other causes there was a lower, and falling, proportion of face-to-face contacts (from 70.5% to 66.0%).

Hospital spells in the MH population

The proportion of the MH population who died by suicide and had a mental health hospital spell in the year before death decreased from 23.6% in 2018 to 17.4% in 2023. This proportion was lower in the MH population who died from causes other than suicide, decreasing from 7.1% in 2018 to 5.4% in 2023.

In 2023 393 patients who died by suicide had a hospital spell, of that group 51 had a hospital spell that was still open at the time of death. This is between 2% and 3% of suicides in the MH population with an open referral in the year before death. This is lower than the 5% reported NCISH annual report likely reflecting methodological differences.

Proportion discharged from hospital spells before death

Suicide risk is known to be highest in the first week after discharge and remains elevated for up to 3 months ^{[footnote 9]}. This analysis shows that for people in the MH population who had an open referral in the year before death, 10.0% were discharged from their last hospital spell in the 12 weeks before their death in 2018 decreasing to 7.1% in 2023.

The proportion discharged in the week before death increased from 1.5% in 2018 to nearly 1.8% in 2023.

Most recent contact or missed appointment in the year before death

Attended face-to-face contact was the most frequently recorded latest service activity for people who died by suicide, although it fell from 46.8% in 2018 to 41.0% in 2023. Other community contact types, including email, messaging services or chat rooms, increased from 17.1% in 2018 to 27.8% in 2023 in this group.

The proportion whose latest recorded service activity was a missed appointment fell for the MH population who died by suicide, from 19.0% in 2018 to 15.4% in 2023. This proportion was lower in those who died from causes other than suicide (15.2% in 2018 and 12% in 2023).

Conclusions

The data in this report, covering people aged 18 to 74, provides important intelligence on the number and rate of suicides in the MH population. It suggests that monitoring this group can support suicide prevention.

The figures for England are useful and regional figures add further insight. However, regional patterns are harder to interpret because several factors vary by area, including:

• overall suicide rates
• suicide rates in the MH population
• the size and composition of the MH population
• suicide registration delays

Analysis in this report shows that there is a large difference in suicide mortality between the MH and non-MH populations, and the difference is greater for women than men. It also shows that this difference is greater for suicide than for 4 major physical causes of mortality.

The ‘MH services use in the year before suicide’ section shows that a large proportion of people in the MH population did have a service contact or hospital spell. It also provides useful insight into service use in those who died by suicide and how that differs to people who died by other causes. This analysis could assist service providers in developing their suicide prevention strategies.

Although this report makes an important contribution, further analysis would provide an understanding of:

• MH service users who died by suicide who have an open referral between 2 and 5 years prior to death
• the people accessing psychiatric liaison services, including their clinical pathway both before and after that contact

Further to the analysis presented in this report, there would be value in considering:

• the MH population and suicide for all ages
• the population (all ages) in contact with NHS Talking Therapies services
• combining MH and NHS Talking Therapies populations (all ages) and identifying where they overlap
• suicide by date of occurrence rather than registration, which would aid interpretation but require a longer wait for accurate data

Acknowledgements

We are particularly thankful for the contributions from Professor Louis Appleby, Professor Nav Kapur, Dr Alison Brabban, Dr Adrian Whittington, Hina Sharma and Debra Moore.

The responsible statistician is the Head of Intelligence, Mental Health Intelligence Team, OHID.

The product leads are the:

• Programme Lead, Mental Health Intelligence Team, OHID
• Deputy Director, Clinical Epidemiology, OHID

If you have any questions relating to this publication, contact mhit@dhsc.gov.uk.

Analysis on groups not covered by this report

Data on suicide in this report is an addition to existing mortality statistics on the MH population aged 18 to 74. This report does not include analysis on children and young people or adults aged 75 and over.

Suicide prevention activities are not limited to people in contact with secondary mental health services and generally consider people of all ages. Wider evidence indicates the importance of considering contact with primary care, other healthcare services, physical health services and community support. This report highlights some of these areas for future analysis. However, already existing evidence might be of value to organisations planning and delivering suicide prevention and mental health services.

NHS England’s Staying safe from suicide guidance sets out best practice for mental health practitioners in England. It recommends moving away from static approaches to suicide risk prediction and risk stratification. It supports collaborative and person-centred assessment, formulation, risk management and safety planning.

Children and young people

Research by NCISH considers people aged 10 to 19 who died by suicide. Evidence from these studies highlight that suicide prevention activities for children and young people should consider:

• self-harm presentation
• co-ordinated prevention across health, social care and education
• presentation of previous risks
• online activity

Self-harm presentation

Self-harm presentations in children and young people provide an important opportunity for comprehensive psychosocial assessment, treatment of underlying needs and continuing support to reduce the risk of suicide^{[footnote 10]}. Data shows that of the group studied:

• almost half had a recorded history of self-harm
• around one-quarter had self-harmed in the 3 months before death
• recent self-harm was more commonly recorded among girls than boys
• recent self-harm was more common in those with a diagnosed mental health condition, alcohol misuse, experience of physical, sexual or emotional abuse, recent life adversity, and contact with mental health services, emergency departments or general health services for a mental health condition in the 3 months before death

Co-ordinated prevention across health, social care and education

Co-ordinated prevention across education, health services, social care and the voluntary sector is an important part of suicide prevention. This response should reflect differences in young people’s experience^{[footnote 11]}. Data shows that:

• witnessing domestic violence, abuse, bereavement, bullying, self-harm and academic pressures were recorded before death and these were generally more common in girls

• drug misuse and workplace problems were recorded before death and these were generally more common in boys

• 60% of the group studied had been in contact with specialist children’s services before death

Presentation of previous risks

Suicide risk identification should not rely only on previous self-harm, disclosed suicidal thoughts or contact with specialist services, and timely access to crisis services is important ^{[footnote 12]}. Data shows that in the group studied:

• around one-third had no known history of suicidal thoughts or self-harm
• lower levels of other recognised risk factors and less contact with services was recorded

Online activity

Mental health professionals should consider online activity experience during assessment of children and young people at risk of suicide. Wider action to improve online safety and support for children and families is also important. Data^{[footnote 13]} shows:

• suicide-related online experience in 24% of the deaths studied, including searching for information about method, posting suicidal thoughts and experiencing online bullying
• self-harm, bereavement, social isolation and mental and physical ill-health were more common in those with suicide-related online experience
• online bullying often occurred alongside face-to-face bullying

Adults aged 75 and over

The NCISH annual report 2026 includes UK and Jersey data on suicide among people aged 75 and over between 2013 and 2023. This data shows that for people in this age group:

• there were an estimated 5,184 suicides, accounting for 7% of all suicides
• the number of suicides increased over the period, driven by increases among people aged 75 to 79 and men aged 90 and over
• one-fifth of people aged 75 and over who died by suicide had been in contact with mental health services in the year before death - this proportion was lower than in younger age groups
• among those who had recent contact with mental health services:
  • 55% had a major physical illness
  • 51% lived alone
  • 49% had a primary diagnosis of depression
  • 41% had a history of self-harm
  • 19% had self-harmed in the previous 3 months
  • 14% had a primary diagnosis of dementia
  • 12% had been bereaved in the previous 3 months
  • 7% lived in a nursing or care home

Data from the NCISH report suggests:

• because of the lower level of contact recorded suicide prevention for this age group should extend beyond specialist mental health services and include routes to support through primary care, physical healthcare, social care and community services
• suicide prevention should not rely only on risk factors more commonly associated with suicide in younger age groups

Follow up after discharge and people not in contact with mental health services

Timely follow-up after discharge from inpatient mental health care supports suicide prevention. The NHS Standard Contract requires follow-up within 72 hours. Data published by NCISH shows that the period immediately following discharge from mental health in-patient care is a time of maximum vulnerability - the risk of suicide was highest in the first 1 to 2 weeks post-discharge, peaking specifically on days 3 and 4. Another study shows ^{[footnote 14]} that adverse life events, short (less than one week) final admission, older age and comorbid psychiatric disorders were associated with post-discharge suicide.

Evidence shows that suicide prevention should not rely only on specialist mental health services, a diagnosed mental health condition or previous suicidal behaviour. Prevention approaches should also consider changes in primary care use, emergency and hospital contacts, physical health needs, pain and medicine use. Community-based support is also important for people who have little or no contact with healthcare. Data shows that:

• non-receipt of mental health services was more common in males, younger or older age groups and people living in a rural area, and these groups were less likely to have a recorded psychiatric diagnosis, previous suicidal behaviour or contact with general health services ^{[footnote 15]}
• many people who die by suicide have contact with other healthcare services. One study ^{[footnote 16]} found:
  • 31.4% had contact with a healthcare service in the week before death
  • the last recorded contact was most often in general practice and commonly related to mental health
  • in the month before death, 16.6% had attended an emergency department and 13.0% had been admitted to hospital
  • contacts relating to self-harm, mental health or substance misuse were particularly associated with suicide, especially when they occurred in emergency or urgent care
• patterns in primary care consultation may provide opportunities for suicide prevention. One study ^{[footnote 17]} found that:
  • consulting GP more than once a month during the year before deaths was associated with an increased of suicide
  • medication reviews, depression and pain were the most common reasons for consultation among people who died by suicide
  • increasing consultation frequency may be an important indication of changing need, including where the presenting concern is physical rather than mental health
• healthcare use before suicide also varies substantially. One study ^{[footnote 18]} shows that:
  • two-thirds of people who died by suicide had no or low healthcare use in the year before death
  • some groups had contact mainly for physical health needs, used mental health medicines or had frequent contact with mental health services

Limitations of the analysis

When using these statistics, consider that:

• the MH population is identified using evidence of an open referral to secondary mental health services in the 5 years before death in MHSDS - this is used as a proxy for severe mental health problems and does not capture everyone with mental health problems who died by suicide
• the analysis does not include NHS Talking Therapies, primary care or voluntary sector services that do not submit to MHSDS
• not all people with open referral had a contact and there might be different reasons for this
• the analysis is limited to people aged 18 to 74 because:
  • people younger than 18 years of age are not included as it has not been possible to search for the full 5 years before death because data for this age group was not submitted to MHSDS prior to April 2016
  • people aged 75 and over are not included as the statistics are part of a wider indicator set focusing on premature (under 75) mortality in the MH population
• the statistics start from 2015 because earlier mental health data did not support a consistent 5-year look-back period
• differences in access to secondary mental health services, referral patterns and MHSDS submission may affect who is included in the MH population and this should be considered when comparing areas
• the MH population increased and may have changed in composition over time and this affects the number of deaths assigned to the MH population and the interpretation of trends
• trends should be interpreted alongside changes in:
  • MHSDS coverage
  • service recording
  • coronial processes
  • the 2018 change in the standard of proof for suicide conclusions
  • registration delays (recent years are more likely to be affected by registration delays)
• records with missing age are excluded from MH population estimates because age is needed to calculate directly standardised rates - totals by sex or region may not add up to the England total
• this analysis describes differences between the MH and non-MH population. It cannot show whether referral, contact or inpatient stay in secondary mental health services is associated with a higher or lower risk of suicide
• people at risk of suicide might be accessing MH support from services other than secondary MH services (including NHS talking therapies, primary care or voluntary sector services not submitting to MHSDS) and therefore some of the statistics may underestimate the suicide rate in people with severe mental health problems
• some counts may differ from other published figures because of differences in mortality extracts, reporting periods, population definitions or analytical methods

Data sources and methods

These statistics use the following 3 data sources:

• ONS civil registration of deaths, referred to as ONS mortality extracts
• mental health services data set (MHSDS) and predecessor data sets
• ONS mid-year population estimates

The analysis uses record-level MHSDS data for people referred to secondary mental health services in England. It excludes activity relating only to learning disability or autism.

A person who died by suicide is included in the MH population if they had evidence of an open referral to secondary mental health services in the 5 years before death. This includes people with or without a recorded contact or inpatient stay.

Deaths by suicide

This analysis uses date of registration and starts with people who died by suicide and looks back to identify whether they had been referred to secondary mental health services before death.

Deaths were included if the underlying cause of death was recorded as suicide or injury of undetermined intent, classified using International Classification of Disease (ICD)-10 codes X60 to X84 and Y10 to Y34. Deaths by suicide were assigned to 2 groups:

• MH population for people referred to secondary mental health services in the 5 years before death
• non-MH population for people not referred to secondary mental health services in the 5 years before death

Estimated population for population-based statistics

Denominator populations for the MH and non-MH populations were estimated using MHSDS and predecessor data sets, together with ONS mid-year population estimates for people aged 18 to 74 in England.

For each year of death, the MH population was calculated as people with a referral to secondary mental health services in the financial year that began in that year, or in the 4 previous financial years. Each person was counted once, even if they had more than one referral. For example, the MH population for 2015 included people with a referral in financial years 2014 to 2015, 2013 to 2014, 2012 to 2013, 2011 to 2012 and 2010 to 2011.

The non-MH population was calculated by subtracting the estimated MH population from the ONS mid-year population estimate.

Population estimates were combined across 3-year periods to calculate directly standardised rates.

Rates and difference in mortality

Rates are directly standardised for age using the 2013 European Standard Population. They are calculated using pooled 3-year data and standard OHID methods.

Difference in mortality for suicide in people aged 18 to 74 in the MH and non-MH population (‘excess’ statistic) is calculated in this analysis using the mortality rate ratio approach. This is the difference between the mortality rate for the MH population (those referred to secondary mental health services) and the mortality rate for the non-MH population (those not referred to secondary mental health services), divided by the mortality rate for the non-MH population. It is calculated using DSRs as per the below formula.

Mortality rate ratio = ((DSR for MH population - DSR for non-MH population) ÷ DSR for non-MH population) + 1

In this report, the ‘excess’ statistic is used as ‘times as likely’ or ‘times more likely’. Data published in adult mental health and wellbeing profile (for all deaths and 4 major physical health causes of death) and by NHS England on excess mortality for people referred to secondary mental health services (aged 18 to 74) use percentage change when presenting statistics for differences in mortality in the MH and non-MH population.

The 95% CIs indicate the uncertainty around the statistics and the range within which the value of the mortality estimate is likely to lie. Wider intervals indicate greater uncertainty, while narrower intervals indicate greater precision. For DSR CIs used Dobson and Byar’s method^{[footnote 19]} and for ‘excess’ statistics standardised rate ratio based method was used^{[footnote 20][footnote 21]}. For further detail on the method used for ‘excess’ statistic, see the NHS outcomes framework methodology.

Services included

Statistics in this report use activity relating to patients who receive assessments and treatment from MH services submitting to MHSDS for:

• MH conditions
• support with mental wellbeing

For each patient attending a service located in England data is submitted to MHSDS if the care is:

• wholly funded by the NHS - the data submission for that patient is mandatory
• partially funded by the NHS - the data submission for that patient is mandatory
• wholly funded by any means that is not NHS - the data submission for that patient is optional

Activity relating only to learning disability or autism submitted to MHSDS is not included.

A range of health care providers and organisations submit MH services data to MHSDS, including:

• NHS MH trusts
• NHS acute trusts
• NHS care trusts
• independent sector health care providers offering services to NHS-funded and non-NHS-funded patients
• voluntary sector health care providers and any qualified providers offering MH services
• community services offering secondary care to children

NHS England publishes a list of organisations that currently submit data to MHSDS and are in scope.

Services excluded and their role

Not all people who die by suicide and who experience mental health problems are captured in this analysis.

Mental health services that did not submit data to MHSDS were not included in the analysis.

Service activity within MHSDS relating only to learning disability and autism was excluded.

The analysis also does not include support provided through NHS Talking Therapies, primary care or voluntary sector services not submitting to MHSDS. These services provide important support to people with mental health problems. For example, NHS England’s data shows that in the financial year ending in 2025 1.81 million referrals were made to NHS Talking Therapies.

References

1. Pirkis J, Dandona R, Silverman M and others. Preventing suicide: a public health approach to a global problem. The Lancet Public Health: volume 9, issue 10, pages 787 to 795 (2024).  ↩

2. Arias P, Arqueros M, Suárez–Soto E and others. Early life adversity and suicide: the role of childhood trauma and risk factors in adolescent suicidal behavior. Children and Youth Services Review: volume 179 (2025). ↩

3. Alothman D, Lewis S, Card T and others. The association between living alone, loneliness and suicide mortality and effect modification by age: a case control study. Journal of Affective Disorders: volume 352, pages 278 to 280 (2024). ↩

4. Nafilyana V, Morgana J, Mais D and others. Risk of suicide after diagnosis of severe physical health conditions: a retrospective cohort study of 47 million people. The Lancet Regional Health - Europe: volume 25 (2022) ↩

5. Athey A, Shaff J, Kahn G and others. Association of substance use with suicide mortality: an updated systematic review and meta-analysis. Drug and Alcohol Dependence Reports: volume 14 (2025). ↩

6. Too LS, Spittal MJ, Bugeja L and others. The association between mental disorders and suicide: a systematic review and meta-analysis of record linkage studies. Journal of Affective Disorders: volume 259, pages 302 to 313 (2019).  ↩

7. Chesney E, Goodwin GM and Fazel S. Risks of all-cause and suicide mortality in mental disorders: a meta-review. World Psychiatry: volume 13, issue 2, pages 153 to 160 (2014). ↩

8. While D, Bickley H, Roscoe A and others. Implementation of mental health service recommendations in England and Wales and suicide rates, 1997–2006: a cross-sectional and before-and-after observational study. The Lancet: volume 379, issue 9820, pages 1,005 to 1,012 (2012). ↩

9. Musgrove R, Carr MJ, Kapur N and others. Suicide and other causes of death among working-age and older adults in the year after discharge from in-patient mental healthcare in England: matched cohort study.  The British Journal of Psychiatry: volume 221, issue 2, pages 468 to 475 (2022). ↩

10. Aggarwal S, Tham S-G, Ibrahim S and others. Self-harm in children and young people who die by suicide: UK-wide consecutive case series. The British Journal of Psychiatry: volume 227, issue 3, pages 593 to 600 (2025). ↩

11. Rodway C, Tham S-G, Ibrahim S and others. Children and young people who die by suicide: childhood-related antecedents, gender differences and service contact. The British Journal of Psychiatry Open: volume 6, issue 3 (2020). ↩

12. Rodway C, Tham S-G, Turnbull P and others. Suicide in children and young people: can it happen without warning?, Journal of Affective Disorders: volume 275, pages 307 to 310 (2020).  ↩

13. Rodway C, Tham SG, Richards N and others. Online harms? Suicide-related online experience: a UK-wide case series study of young people who die by suicide. Psychological Medicine: volume 53, issue 10, pages 4,434 to 4,445 (2023). ↩

14. Bickley H, Hunt IM, Windfuhr K and others. Suicide within two weeks of discharge from psychiatric inpatient care: a case-control study. Psychiatric Services: volume 64, number 7, pages 653 to 659 (2013). ↩

15. Tang S, Reily NM, Arena AF and others. People who die by suicide without receiving mental health services: a systematic review. Frontiers in Public Health: volume 9 (2022). ↩

16. John A, DelPozo-Banos M, Gunnell D and others. Contacts with primary and secondary healthcare prior to suicide: case–control whole-population-based study using person-level linked routine data in Wales, UK, 2000–2017. The British Journal of Psychiatry: volume 217, issue 6, pages 717 to 724 (2020). ↩

17. Danah Alothman, Sarah Lewis, Andrew W Fogarty, Timothy Card and Edward Tyrrell. Primary care consultation patterns before suicide: a nationally representative case–control study. British Journal of General Practice: volume 74, issue 744, pages 426 to 433 (2024). ↩

18. Chitty K, Sperandei S, Carter G and others. Five healthcare trajectories in the year before suicide and what they tell us about opportunities for prevention: a population-level case series study. eClinicalMedicine: volume 63 (2023). ↩

19. Breslow NE, Day NE. Statistical methods in cancer research, volume 2: the design and analysis of cohort studies. Lyon: International Agency for Research on Cancer, World Health Organization (1987) ↩

20. Newman SC. Biostatistical methods in epidemiology . New York: John Wiley and Sons (2001)  ↩

21. Armitage P, Berry G, Matthews JNS. Statistical methods in medical research (4th edition). Oxford: Blackwell Science (2001) ↩