Deaths by suspected suicide 2024 to 2025 technical guidance
This guidance is associated with:
Glossary
Associated factors
These factors are obtained from ‘historical risk factors’ as reported by Police. These include:
- Covid 19 – Direct
- Covid 19 – Indirect
- Long Covid
- Adverse Childhood Experience
- Affected by Suicide
- Alcohol Misuse/Abuse
- Awaiting CJ Outcome
- Bereaved by Suicide
- Bereavement
- Bullying (inc. Cyber)
- Carer Responsibilities/Issues
- Child Contact Issues/Child Removal
- Dental pain/dentist access issues
- Domestic Abuse
- Drug Misuse/Abuse
- Educational/Exam Stress
- Family Issues
- Financial Issues
- Hearing Issues/Tinnitus
- Housing/Shelter Concerns
- Learning Difficulties
- Long-term illness
- Mental Health Illness
- Military History
- Not Taking Medication
- Other (free-text field)
- Other Medical Issues
- Pain – Emotional
- Pain – Physical
- Peer Pressure (inc. Cyber)
- Peri/Menopause
- Physical Disability
- Post/Perinatal Concerns
- Previous Convictions Identified
- Previously detained S136 MH Act
- Problem Gambling
- Relationship Issues (i.e Partner/Friends)
- Severe Emotional/Stressful Life Event
- Sexual Orientation Concern
- Social Isolation
- Substance Misuse/Abuse
- Work Issues
The information gathered from the Police reporting form in the above categories, along with further information available in the free text section is used to populate the following RTSSS ‘associated factor’ categories and may be used to populate other RTSSS categories if relevant.
RTSSS ‘associated factor’ categories include:
Covid related: Current Covid-19 / History of Covid-19 infection / Long Covid
Personal circumstances: Financial issues / Work issues / Housing/shelter issues / Social isolation / Bereaved / Bullying (inc. cyber) / Racial abuse/discrimination / Sexual orientation concern / Educational/exam stress / Failure to access support / Peer pressure (inc. cyber) / Problem Gambling / Veteran status / Adverse Childhood Experiences
Domestic circumstances: Care leaver / Carer responsibilities/issues / Domestic abuse – victim / Domestic abuse – perpetrator / Domestic abuse – bystander / Sexual abuse – perpetrator / Sexual abuse – victim / Sexual abuse cyber
Impacted by suicide : Affected by or exposed to suicide / Bereaved by suicide – family / Bereaved by suicide – friend/other
Substance use: Drug use/misuse / Alcohol use/misuse / Drug and Alcohol use/misuse / Drug/alcohol misuse unspecified
Emotional issues: Relationship issues / Family issues / Neighbour issues / Severe emotional/stressful life event / Experience of trauma
Medical and health: Mental health condition (Depression; Anxiety; Dementia; Schizophrenia/ other delusional disorder; PTSD; bipolar disorder; personality disorder; other; unspecified) / Long term illness / Chronic pain / Learning disability / Physical disabilities / Prescribed medication / Adherence with prescribed medication / Post/peri-natal concerns / Perimenopause/menopause / Neurodivergence (e.g. Autism, ADHD) / Other medical issues / Previous self-harm
NB: some of the above categories may be combined for reporting purposes.
Confidence intervals
Confidence intervals are indications of the natural variation that would be expected around a rate and they should be considered when assessing or interpreting a rate. The size of the confidence interval is dependent on the number of events occurring and the size of the population from which the events came. In general, rates based on small numbers of events and small populations are likely to have wider confidence intervals. Conversely, rates based on large populations are likely to have narrower confidence intervals. A 95% confidence interval means that we are 95% confident that the true value of the estimate lies within the range.
Count
The count is the number of suspected suicides that occurred over a particular period of time.
Crude rate
A crude rate is the number of suspected suicides occurring in a population over a specific time period, expressed as the number of deaths per 100,000 of the population. These rates were used as they are most suitable to inform action, which is one of the aims of the RTSSS.
Known to Police (in previous 6 months) categories:
- Vulnerable person – this includes situations where a person is unable to take care of or protect themselves or others from harm or exploitation.
- Suspected/convicted of a crime – includes situations where a person was suspected of a crime or crimes; they may or may not have been convicted.
- Victim of a crime – includes situations where a person was a victim of a crime.
- Witness – includes situations where a person was a witness to a crime.
- Other – this includes situations not captured by existing categories, for example police attendance at an incident such as disputes. It does not include whether the person was a firearms license holder.
Mean
The average number of deaths.
Mental health condition
This is a broad term covering conditions that affect emotions, thinking and behaviour, and which substantially interfere with our life. Mental health conditions can significantly impact daily living, including our ability to work, care for ourselves and our family, and our ability to relate and interact with others. This is a term used to cover several conditions (e.g. depression, post-traumatic stress disorder, schizophrenia) with different symptoms and impacts for varying lengths of time, for each person. Mental health conditions can range from mild through to severe and enduring illness. People with mental health conditions are more likely to experience lower levels of physical and mental wellbeing, but this is not always or necessarily the case. Some mental health conditions like eating disorders and schizophrenia are associated with a higher risk of mortality (Understanding: a suicide prevention and self-harm strategy).
The sources of information for ‘mental health condition’ may have differed slightly in each Police force. All Police force reporters based the information reported to RTSSS on evidence that the person had a diagnosed mental health condition. This information may have been obtained from health records, police records or from a family statement.
Mode of death categories
Mode of death categories reflect categories used in Near to real-time suspected suicide surveillance (nRTSSS) for England, which are based on ONS categories. They are presented as follows:
- Hanging, strangulation and suffocation
- Drowning
- Fall and fracture
- Poisoning
- Jumping or lying in front of a moving object
- Sharp object
- Other or unknown (*Other includes the following modes of death: shotgun/firearm; self-harm by explosive material; self-harm by smoke, fire and flames; self-harm by steam, hot vapours and hot objects; self-harm by blunt object; self-harm by crashing of motor vehicle; self-harm by electrocution; other modes not included in any other category).
Rate
The rates in this report are crude rates (see above).
Regions
The three regions of North Wales, Mid and West Wales and South-East Wales are defined below and are consistent with the regional suicide prevention fora in Wales.
North Wales – Health boards: Betsi Cadwaladr University Health Board. Local authorities: Isle of Anglesey, Gwynedd, Conwy, Denbighshire, Flintshire, Wrexham.
Mid and West Wales – Health boards: Hywel Dda University Health Board, Swansea Bay University Health Board, Powys Teaching Health Board. Local authorities: Carmarthenshire, Ceredigion, Pembrokeshire, Swansea, Neath Port Talbot, Powys.
South-East Wales – Health Boards: Aneurin Bevan University Health Board, Cwm Taf Morgannwg University Health Board, Cardiff & Vale University Health Board. Local authorities: Blaenau Gwent, Caerphilly, Monmouthshire, Newport, Torfaen, Bridgend, Merthyr Tydfil, Rhondda Cynon Taf, Cardiff, Vale of Glamorgan.
Self-harm
Self-harm refers to an intentional act of self-poisoning or self-injury, irrespective of the motivation or apparent purpose of the act. Self-harm includes suicide attempts as well as acts involving little or no suicidal intent (Understanding: a suicide prevention and self-harm strategy).
The sources of information for ‘history of previous self-harm’ may have differed slightly in each Police force. All Police force reporters based the information reported to RTSSS on evidence that the person had a history of self-harm. This information may have been obtained from health records, police records or from a family statement.
Standard deviation
A measure of the amount of variation of a set of values in relation to the mean.
Statistical significance
Statistical significance when comparing local area estimates to the all-Wales value was determined using 95% confidence intervals. The local area estimate is statistically significantly different if its confidence interval lies outside the Wales value. If the confidence interval overlaps with the Wales value then the difference is not statistically significant.
When comparing local area estimates with another local area estimate, age groups by sex, and deprivation fifths, non-overlapping confidence intervals between values indicate that the difference is unlikely to have arisen from random fluctuation (i.e. statistically significant). However, when the confidence intervals overlap, we cannot determine if there is a statistically significant difference or not by comparing confidence intervals alone, so a more exact test is required. The pairwise comparison looked at the difference between the rates and the 95% confidence intervals of the difference. When the confidence interval of the rate difference is above zero, the two rates are considered significantly different with 95% confidence.
Substance misuse
Substance misuse is formally defined as the continued use of any psychoactive substance that substantially affects a person’s physical and mental health, social situation and responsibilities. The most severe forms of substance misuse are normally treated by specialist drug and alcohol rehabilitation services. Substance misuse covers misuse of a range of psychoactive substances including alcohol, illicit drugs and licit drugs including prescribed medications taken in a way not recommended by a GP or the manufacturer (Understanding: a suicide prevention and self-harm strategy).
Suspected suicide
When a person is suspected to have taken their own life intentionally (Understanding: a suicide prevention and self-harm strategy).
A death by suspected suicide as reported in most cases here has been determined by the Police. The College of Policing have outlined the classification of suspected suicide and state that:
“..There is often a requirement for an initial judgment to be made on whether a case is potentially suicide. … Officers should use their professional judgment – based on all the known facts – and supported by the national decision model (NDM), to record whether a fatality is a suspected suicide. Witness accounts, CCTV material, the presence of a suicide note and other available evidence will help in this determination.
Legal basis
The legal basis for the processing of data is Paragraph 3(b) of the Public Health Wales NHS Trust (Establishment) Order 2009 “to develop and maintain arrangements for making information about matters related to the protection and improvement of health in Wales available to the public in Wales; to undertake and commission research into such matters and to contribute to the provision and development of training in such matters” and Paragraph 3(c) of the Public Health Wales NHS Trust (Establishment) Order 2009 which states as one of its functions: ‘to undertake the systematic collection, analysis and dissemination of information about the health of the people of Wales in particular including cancer incidence, mortality and survival; and prevalence of congenital anomalies.’
RTSSS has Data Disclosure Agreements in place with the four Welsh Police forces and British Transport Police to receive information via the British Transport Police (BTP)/National Police Chief’s Council (NPCC) data collection template with the addition of fields for name and date of birth. Although the Data Protection Act 2018 and General Data Protection Regulations do not apply to the data collected for RTSSS, the exchange of personal data is conducted within the legal framework of the Data Protection Act 2018 and in compliance with the common law duty of confidence. We have conditional support from the Confidentiality Advisory Group to process confidential patient information without consent (Ref: 22/CAG/0163).
Data sources
Notification of deaths by suspected suicide: Data in this report were obtained from the RTSSS database. Information is provided to RTSSS mainly by the four Welsh police forces, using a template developed by the British Transport Police (BTP) for the National Police Chief’s Council (NPCC) Suicide Prevention Portfolio. Suspected suicides have been determined to be suspected suicides by the Police (see ‘suspected suicide’ in glossary).
In addition to data in the BTP/NPCC template, we obtain name and date of birth from the four Welsh Police forces so that we are able to link each record with other data sources to cross check information and add additional information. BTP also separately notify suspected suicides which are transport related. Other sources for initial notifications include ad-hoc reports from services outside of Wales and the NHS Wales Joint Commissioning Committee.
Cross checking and additional information: A number of sources of data are used to cross check the information received in the initial notification, and to obtain further information where there may be gaps. These include: Welsh Clinical Portal, Welsh Demographic Service, Child Death Review Programme, Network Rail and Nationally Reported Incidents held by NHS Performance and Improvement.
Welsh Index of Multiple Deprivation 2021 (WIMD) was used as the estimate of deprivation. It is Welsh Government’s official measure of relative deprivation for small areas in Wales. It is made up of eight separate domains/types of deprivation: Income, Employment, Health, Education, Access to Services, Housing, Community Safety and Physical Environment.
The ONS mid-year estimates (MYE) were used as the denominator when calculating rates. The ONS is the official source of population sizes, produced annually, covering populations of local authorities, counties, regions and countries of the UK by age and sex. Denominator for 2024/25 rates were based on lower super output areas MYE, 2022. For local authorities, the rates were based on local authority MYE 2022, 2023, 2024.
ONS Census 2011 data was used for estimating employment rates.
Location data were derived from postcodes, What Three Words and grid reference data provided by the data suppliers, on the British National Grid. If these were not available, name of health board or region of residence was supplied, if known.
Data quality
Data quality has been considered using the Data Management Association UK dimensions, i.e. completeness, accuracy, timeliness, uniqueness, consistency and validity.
It should be noted that ethnic group and gender identity are important indicators, but the data quality of these data items is not clear, so we are not able to report on them. See section on Limitations for a further explanation.
Completeness
| Indicator in RTSSS report | Data item in deaths reported by Police | Data item in deaths reported by non-Police source | Completeness |
|---|---|---|---|
| Month of death | Date of death | Date of death or month of death was available | 100% complete, <5 cases with minor date discrepancy (i.e. day of death) between different sources. |
| Region & health board of residence | Postcode of residence | Region & health board of residence was available | 100% complete in deaths reported on BTP/NPCC template. For deaths notified by other sources postcode was not available but region and health board of residence was available for all deaths. |
| Area deprivation (based on Welsh Index of Multiple Deprivation) | No data field – obtained from postcode of residence | Not available unless postcode provided | 98% complete. |
| Age range | Age | Age or age range was available | 100% complete. |
| Sex (assigned at birth) | Gender (includes information on sex and gender) | Data on sex was available | 100% complete. Only male and female data items used from the NPCC/BTP gender category and referred to as sex. For data items other than male/female, sex assigned at birth obtained from Welsh Demographic Service. |
| Employment status (Unemployed, Employed/self employed, Retired, Student/apprentice, Other) | Employment status/occupation (includes the following options: former police officer, serving police officer, former police staff, serving police staff, serving HM forces, NHS staff, social care worker, other public sector, other private sector, self-employed, unemployed, student, retired, other, unknown). | Employment status was not available | 76% complete. If occupation rather than employment type was listed in BTP/NPCC return, employment status was determined from this. Free text field was also interrogated. |
| Associated factors | Historical risk factor(s) | Information on associated factors was available for some deaths | Only positive findings were collected and reported. Unknown level of completeness therefore may be underestimated. Free text field was also interrogated to populate data fields in RTSSS database. |
| Known to mental health services in previous 6 months | Known to mental health services 6 months prior to death | Use of mental health services was available for some deaths | 75% complete overall. Free text field was also interrogated. Mainly police reported although some information from health systems (see Limitations section). |
| Known to Police in previous 6 months |
Known to Police in previous 6 months |
Known to Police not available | 98% complete. Free text field was also interrogated. |
| Mode of death | Suicide method | Mode of death usually available | 98% complete. |
| Location type | Suicide means location type | Location type usually available | 100% complete |
Accuracy
Data were cross checked with other sources where possible. Queries relating to the data on the BTP/NPCC return were checked directly with the Police Forces. A small number of cases were reported by sources other than the Police and there was less information available on these. Also, we did not obtain personally identifiable information on these cases so cross-checking or adding further information was not possible.
An evaluation of RTSSS data is planned which will compare suspected suicides against ONS reported suicides. This will give an indication of the accuracy of the system.
Timeliness
Data on the BTP/NPCC return were received by RTSSS within 10 working days of the following month from the month of death. One Police force submitted a weekly return. The addition of further information could take up to several weeks. A small number of deaths were not reported in the initial BTP/NPCC return, but were added retrospectively, with revised figures from the previous year being published. Deaths reported via other sources could take one to two months to be received by RTSSS.
Data on occurrences of suspected suicides is more timely than ONS official suicide statistics, so can provide a more timely indication of emerging patterns and trends.
Consistency
Each record was inputted by a member of the RTSSS core team and quality checked by another team member. If data values conflicted with other values the data source was checked and queried if necessary. A monthly data review meeting was held by the core RTSSS team to review any records where there were queries about the data identified from either the data entry or data quality checking stages.
Uniqueness
Duplicated entries would be identified during the data entry or quality checking stages.
Validity
Data were cross checked with other sources where possible. RTSSS has a minimum dataset with definitions and rules relating to the data items collected.
Data analysis
Rates
The rates referred to in this report are crude rates as they are most suitable to inform action, which is one of the aims of the RTSSS. A crude rate is the number of suspected suicides occurring in a population over a specific time period, expressed as the number of deaths per 100,000 of the population. For these analyses, the rates used a denominator based on lower super output areas 2022 ONS mid-year estimates. This is because later years were not available. For local authorities, the population used to derive the rates were based on 2022, 2023, 2024 ONS mid-year estimates as these were available for LA geographies.
Region, health board, sex, age/sex, and deprivation rates are estimated rates. 95% confidence intervals around these rates were calculated to give an indication of the precision of the estimate of the rate.
For comparisons between:
- regional estimates
- health board estimates
- local authority estimates
- deprivation estimates
- sex estimates
and the all-Wales rate, the all-Wales rate is treated as an exact reference (no confidence interval). This is a widely adopted method for national level estimates, with the random error deemed negligible for large populations. If the confidence interval of the estimate lies outside of the all-Wales rate, then the difference is statistically significant. If the confidence interval of the estimate overlaps the all-Wales rate, the difference is not statistically significant.
For comparisons between two estimates such as:
- regional estimates with another regional estimate
- health board estimates with another health board estimate
- local authority estimates with another local authority estimate
- deprivation estimates with another deprivation estimate
Non-overlapping confidence intervals between values indicate that the difference is unlikely to have arisen from random variation (i.e., statistically significant). However, when the confidence intervals overlap, we cannot determine if there is a statistically significant difference or not by comparing confidence intervals alone, so a more exact test is required. The pairwise comparison looked at the difference between the rates and the 95% confidence intervals of the difference. When the confidence interval of the rate difference is above zero, the two rates are considered significantly different with 95% confidence. In order for the pairwise estimates to be robust a minimum count of 10 was required, so this was not undertaken for comparisons between age groups.
Therefore, where estimated rates are compared with each other, a difference is statistically significant if either:
- the confidence intervals of the values do not overlap
- the confidence intervals overlap, but the confidence interval for the difference between the rates does not include zero.
Where appropriate, the mean (average) number of cases and standard deviation were estimated. It is expected that counts are within one standard deviation above or below the mean two thirds of the time. This gives a measure of whether there are counts or trends of concern.
Strengths
Figures are for deaths that occurred during the stated time period and provide a timely indication of suspected suicides. This compares with official statistics which are published by year of registration, so the actual occurrence of those deaths may have been months or years prior.
Data fields for month of death, age, sex, postcode (for deprivation quintile), known to police in previous 6 months, mode of death and location type were almost 100% complete.
Collection of personally identifiable information means that we were able to link the data with additional data sources, either to add further information or to cross check existing data therefore improving the quality of the dataset.
Limitations
The data collected are surveillance data so although we are able to provide more timely data than official statistics, the data are not of as high quality. There is limited trend data available, as only data from the previous two years are available as a comparison.
This report contains categories with small numbers which are prone to random variation and so caution should be used when interpreting the rate.
There are large confidence intervals around the rate estimates.
Deaths of all Welsh residents by suspected suicide may not be fully captured, because:
- We do not yet have fully established links with other RTSSS teams outside of Wales, so we may not have captured all deaths of Welsh residents that occurred outside Wales.
- We do not yet have an established process for capturing suspected suicides where the death occurred in hospital after the event.
Data on occupation was incomplete so were not included in this report.
Data on mental health conditions and whether the person was known to mental health services in the 6 months prior to death was mainly based on information available to the Police. Much of this information included data from health information systems but this may not have always been available. We were able to include additional information and cross check some, but not all, data on mental health conditions and use of mental health services with other sources (e.g. Welsh Clinical Portal, reports from NHS Wales Joint Commissioning Committee).
The extent of data capture may vary between Police forces as different systems are accessed to obtain data.
A list of data fields has been developed for the RTSSS, but we are not yet able to collect all of the data, e.g., religion, disability status, or to establish the level of data quality, e.g. ethnic group, gender identity, sexual orientation.
For ethnic group we will explore the options for improving data quality, but it remains difficult in the absence of access to GP data. Data on gender identity or sexual orientation is not currently readily available from other sources, but when this becomes available from a reliable source, we will explore the feasibility of collecting this information.
The Office for National Statistics have published analyses estimating the rates of suicide by ethnic group, disability status, religious group (Sociodemographic inequalities in suicides in England and Wales – Office for National Statistics) and sexual orientation (Self-harm and suicide by sexual orientation, England and Wales – Office for National Statistics).
Amendments since 2023 to 2024 annual surveillance report
Following feedback from users:
Additional analyses using data from 1 April 2022 to 31 March 2025 have been included. These are:
- Crude rates of death by suspected suicide, by local authority area of residence, 3 year period
- The most common associated factors by sex, 3 year period
- The most common mode of death by sex, 3 year period
- Domestic abuse perpetrators and victims, by sex, 3 year period.
- The ‘Known to police (ever)’ indicator has been removed and ‘Known to Police in previous 6 months’ has been added.
- The title of the ‘Known to mental health services’ indicator has been amended to ‘Known to mental health services in previous 6 months’ to clarify the time period in which the person had been known.