Hospitalisation from serious violence

Serious Violence is a public health problem that can impact on the provision of health services. The focus on Serious Violence has been highlighted since the Serious Violence Strategy, where it was shown that reducing Serious Violence is one the key aims of the Home Office as outlined in the Beating Crime Plan published in 2021. The requested admissions data is the key metric through which the Home Office measures its performance against this ambition.

The aggregate analyses published by NHS England are useful in establishing an initial sense of what is in the data. However, the Home Office requires the unsuppressed tabulated data to allow the data to be examined in greater, more accurate detail, helping the Home Office demonstrate the extent to which it is achieving this ambition. The data is not identifiable and cannot be combined with any other datasets to make it identifiable.

The data is widely used to inform analysis, the suppressed results of which are disseminated to No 10 and other departments to track progress towards reducing Serious Violence. Reducing Serious Violence is one of a small number of priorities (‘critical outcomes’) that No10 Delivery unit monitor and regularly update the Prime Minister (PM) on progress. There is strong interest in subnational trends and the work the Home Office carries out informs senior Ministerial decision making.

The aim and purpose of this Data Sharing Agreement is to receive unsuppressed tabulated Hospital Episode Statistics (HES) data. The HES Admitted Patient Care (APC) dataset is the national dataset for admitted patients treated in England. This data requested covers any patient admitted to hospital under Finished Admissions Episodes or Finished Consultant Episodes in England for assault injuries listed in the dataset. Police recorded crime data can be affected by police recording changes, so an independent measure is needed to ensure that other factors are not influencing funding decisions.

Analysis of the APC data will be used to do four main things:

1) Add to the evidence on the extent of assaults with injury and help to better understand the nature of the problem

2) Contribute to Home Office evaluations of serious violence policies. The evaluation of policies will make sure impact of Serious Violence policies is fully understood, helping direct public money to the areas to where there it can have the most impact. The data will be used to evaluate whether the Violence Reduction Units (VRUs) fund achieves its stated aims throughout the next three years and ensure that circa £105 million is best spent.

3) Allocate funding, nationally, at police force area level, and by local authority. The use of the data to allocate funding will ensure that policies are directed at the areas with most need.

4) Monitor serious violence trends nationally and by police force area. This will help the Policing Minister hold police forces to account for their performance through their monthly Crime Programmes meeting and help drive up standards.

As shown in the Home Office’s Serious Violence Strategy (2018), Serious Violence is a high harm form of crime, and has wider impacts on society, including health, poverty, education, and mental health. It is therefore in the public interest to provide policies to reduce Serious Violence, and to ensure that public funds spent are targeted to the areas that need it most, and where it can have the most impact.

The Home Office having access to the admissions data is important for healthcare and social care, as Serious Violence has both direct and indirect effects on them. Serious Violence leads to injury, which puts a pressure on healthcare, especially emergency services and hospitals. This is reflected in the hospital admissions data that the Home Office are applying for. It also indirectly affects healthcare, as high levels of Serious Violence are linked to negative health outcomes, and poor mental health of those affected by Serious Violence. Reducing Serious Violence could therefore reduce the strain that these incidents put on the NHS.

The data received will be aggregated with small numbers unsuppressed. The number of tabulations requested are necessary for the following reasons: monthly data by Police Force Area and under 25/25+ ages to monitor effectively the programme impacts and trends in serious violence, to feed back to the Prime Minister, Crime and Policing Minister and Home Secretary about serious violence for their monthly meetings. More detailed biannual breakdowns are required for investigating the drivers of trends (for example changes in the gender or age ratios) to understand better how to direct policies, and for evaluation of existing policies.

The data years are minimised by requiring a data back to capture the long term trends, but then to only require recent data updates in future months for those data which remain provisional. The long term trends are needed for placing levels in context, tracking back to historic highs and lows in trends, and for robust evaluations. The monthly data also only requires two age bands rather than multiple age bands or actual ages, which reduces the amount of information collected, and still fulfils the aims of reducing serious violence and in particular youth serious violence by being able to understand trends in the younger victims as well as for all ages. By only requesting more detailed demographics (10-year age bands, gender and ethnicity) in the biannual returns, this minimises the amount of data at each point.

The more detailed breakdowns are more infrequent, and are required for evaluations of serious violence programmes, to ensure all impacts of funding and programmes are collected and assessed. Serious violence can be captured by several measures, and whilst monthly, one measure is enough to understand trends and is high volume enough to be reliable; the other metrics are more low volume and likely to fluctuate, so annual data will suffice and minimises the data required. It is still necessary to capture all forms of serious violence, including those more likely to affect women and children, to ensure holistic public safety.

Data will be narrowed by aggregated geography (police force area for monthly data, and only as low as local authority for biannual data). It covers all patients for the admission types in question, both to ensure funding is directed accurately, requiring a national understanding of serious violence, and also to ensure higher volumes for reliable context. The admissions type is a wide category to ensure the data request is minimised.

Patient records are not linked so repeated patient records are not identifiable. There are no alternative, less intrusive ways of achieving the purpose.

The Home Office are the sole data controllers, and are also data processors of the data requested in this Agreement. There is also an evaluation partner for the Violence Reduction Unit (VRU) evaluation that will be using the data, the University of Hull. They have a storage partner (details below). The University of Hull do not determine the purpose or the means of data processing. This is evidenced via the Data Protection Impact Assessment completed by the Data Protection Office at the Home Office. The Home Office determine how the University of Hull will use the data and provide the legal bases for this use. All processing / analysis of the data is performed by substantive staff at both the Home Office and the University of Hull.

No other organisations are involved using the unsuppressed data. If future evaluation partners are commissioned to evaluate other policies, the Data Sharing Agreement will be updated and amended to reflect this. The Home Office solely determine the purposes and the means of the processing of the unsuppressed data. The Home Office share suppressed results with the Prime Minister via the Cabinet Office to feed into serious violence understanding. The Cabinet Office are not joint data controllers. They do not determine the purpose or the means of the processing.

AIMES supply IT infrastructure for the University of Hull and are therefore listed as data processors. They supply support to the system, but do not access data. They will play no part in the analysis. Therefore, any access to the data held under this agreement would be considered a breach of the agreement. This includes granting of access to the database[s] containing the data. AIMES have system administrative access to the system but act on the instructions of University of Hull DSH admin. AIMES are processor for the purposes of provisioning, securing and back-up of the service.

The Home Office rely on the below legal bases for the processing of data under this Agreement:

GDPR Article 6(1)(e) Public Task. The need for the data is a statutory and Government purpose - the purpose is to reduce serious violence, which is in the public interest for safety and security of society, to ensure that any funding is targeted to where it is needed most. This includes protecting the public funds, evaluating any policies on serious violence to ensure their effectiveness, and protecting the public effectively whilst ensuring public funds are used effectively. It is necessary to achieve this purpose because the police recorded crime data do not cover all elements of serious violence and can be affected by other factors, and the data the Home Office are requesting is not record level so risk to individuals is small.

This also leads to the public interest element of the data access. Under GDPR Article 9(2)(g), processing health data can be met by substantial public interest, which under Part 2 of Schedule 1 of the Data Protection Act 2018 includes processing for statutory and government processes, preventing or detecting unlawful acts, safeguarding of children and individuals at risk (among other factors, but these are the elements met by the way in which the Home Office will use the data). The work is on behalf of the Home Office as a Government department, and is necessary for the purposes above. It relates to preventing serious violent crime, which is an unlawful act, used to identify areas in which serious violence is more prevalent, target funding and policies used for preventing serious violence. Many of these policies are specifically targeted at young people using cases against those aged under 25, and are therefore more vulnerable. Policies include safeguarding young people and people at risk of serious violence, by targeting at the areas with the highest levels of serious violence. Whilst exact individuals will not be identifiable using the data, by using the areas as targeted by the levels of serious violence, it will reach more of those at risk of victimisation from serious violence.

Hospital Episode Statistics (HES) Admitted Patient Care (APC) data will be released to the Home Office in the form of tabulations with small numbers unsuppressed.

The dataset provided will be an ongoing supply and consists of 3 separate files as follows:

a) Bi-Annual Assault with a Sharp Object

The tabulation is the count of inpatient finished admission episodes and finished consultant episodes with a cause code of ‘assault with a sharp object’ broken down geographically by Government Office Region (GOR) of residence by month, Local Authority of residence, Police Force Area of residence and by gender, age and ethnicity, for the years 2012-13 onwards.

b) Monthly Police Force Area Assault with a Sharp Object

This tabulation shows counts of inpatient finished admission episodes with a cause code of ‘assault with a sharp object’, by month and by a) Police Force Area and b) Police Force Area and age, for the years 2012-13 onwards.

c) NHS hospital admissions from across all ICD-10 Assault categories (X92-Y09), Assault by sharp object being the predominant outcome, by month, by Police Force Area and Local Authority, by age categories of all ages, under 25s, and 25+, for the years 2012-13 onwards.

Data flow: NHS England sends unsuppressed data (as described above) to the Serious Violence team at the Home Office. The Home Office sends this on to the University of Hull approved team for them to use in the evaluation. All of their outputs will be suppressed in line with the HES Analysis guide. The Serious Violence team at the Home Office also sends suppressed analysis and suppressed data, once combined with NHS Wales data, to other teams within the Home Office, as well as to other government departments and police forces. All outputs will be suppressed if sent outside the immediate team (listed on the access log) in line with the HES Analysis guide. The data are not being used to target individuals, just areas which are wide.

Data are to be combined with the Welsh version of the dataset, using Police Force Area and month as linking points. This does not introduce any additional categories or information about cases, as the data are aggregated. The datasets are the same format; one collected by NHS services based in England, and the other is the same for Wales. This is to ensure a holistic national understanding of serious violence across England and Wales (as crime is not devolved in the way that health is). Additional risks of identification are mitigated by also suppressing the national figures where 'all ages' are not used, as this would risk revealing the Wales low volume figures. The data are not identifiable for individuals as volumes are aggregated and anonymised, and the linking with the Wales dataset increases the volumes so reduces risk.

Only those in the Serious Violence team would have access to the data, and the list of those people would be kept on a data access log. No unsuppressed data would leave this team, and would be kept in secure folders and tagged as not for sharing outside the team, with suppression instructions to hand. No attempt will be made to re-identify individuals. All processing will be done by substantive employees of the Home Office and University of Hull as processors who are trained in data protection and confidentiality.

The Home Office will store this data on a secure server which only the serious violence team at the Home Office can access. The folders are only accessible by someone already with access granting access to new people when they join the team and access log. No data will be stored on any cloud based systems. A record of all users of the unsuppressed data has been created. This list must be maintained regularly and anyone accessing the data that is not included on the list of users will be breaching the Agreement.

The University of Hull are the Violence Reduction Unit (VRU) evaluation team. They have been commissioned by the HO to evaluate the VRU funding, and so need access to the unsuppressed data as well, in order to effectively assess impact at an area level and for the age groups targeted by the funding. They will use a secure system to hold and transfer the data - data related to this project will be stored and processed exclusively within the University of Hull Data Safe Haven. The Data Safe Haven is managed by the Hull Health Trials Unit based at the University of Hull. The infrastructure is built on a Trusted Research Environment service provided by AIMES. The data will be stored and all analysis will reside in the secure AIMES data centre in Liverpool. Researchers access the environment via a secure VPN connection with 2-factor authentication. As such no data will sit on the end users machine.

Any additional aggregate data that is created using the original data that was shared in these tabulations must employ the same disclosure control methods where the output is to be used externally. Any public facing views of the shared data must employ disclosure control as in the public facing published outputs in order to ensure appropriate levels of confidentiality are maintained.

There are several outputs for the data processing. Ministerial reports, presentations and slides with monitoring of trends and basic descriptive statistics for internal-to-government and police audiences, and dashboards for internal audiences and ministers and police/Police and Crime Commissioners. The ministerial reports occur at least on a monthly basis so the first of these is intended to be produced in January 2022 and then at monthly intervals thereafter. Data used in evaluations will be published in peer-reviewed evaluations in the public domain, the first of these evaluation reports will be published in Autumn 2022 and then annually thereafter. All of the data presented, whether internal or public, will follow the publishable standard of suppression as according to the HES Analysis Guide.

Most data will be aggregated to national level or in some cases police force area level. Much of this will be annual totals, although some may be in other time periods. All of the data will be suppressed and contain the relevant caveats on interpretation, disclosure, and sharing.

Research analysis of government funded programmes with evaluations will be published using the suppressed data, to demonstrate findings and as part of evaluating value for money using public funds. It will also add to the evidence base of what works in serious violence, and the relevance of the data to assessing this. The results will be made available in the public domain through a peer-reviewed publication, and will also be promoted to policy makers internally and with other government departments, as well as other researchers. It may be used in academic and public conferences, and also with police forces and relevant parties to the police.

Monitoring data will be produced monthly internally, with more detailed analysis and reports less frequently (likely annually). Evaluations will be tied to publishing timetables set by the publishing team at the Home Office, and will be published in line with any funding requirements of evaluation set by HM Treasury, so the timelines are not pre-specified.

Serious Violence is a high harm form of crime, and has wider impacts on society, including health, poverty, education, and mental health. Serious Violence has both direct and indirect effects on health and social care. Serious Violence leads to injury, which puts a pressure on healthcare, especially emergency services and hospitals. Indirect effects on healthcare include poor mental health of those affected by Serious Violence. Accessing the data will help the Home Office reduce Serious Violence. This is anticipated to allow the NHS Hospitals most affected by Serious Violence to reallocate the resources they would gain back from a reduction in Serious Violence incidents to areas as they see fit, providing benefits to health resources.

The use of the requested hospital admissions data for monitoring and evaluating policies is expected to carry key benefits to health and social care. It aims to:

1. Allow the development of policies to reduce serious violence, and the evaluation to ensure the effectiveness of such policies. It will allow better targeting of funding to reduce serious violence in areas which have high volumes of serious violence, which will reduce serious violence injuries requiring treatment within the health system and reduce impact on social care resources.

2. Inform Government response, and feed into ministerial decision making around policies and programmes as well as funding decisions, ensuring these are reaching areas and people most in need of support.

3. Be useful in disseminating knowledge gained to the public and to non-governmental research bodies to improve the evidence base and understanding of what works in relation to reducing serious violence. This will have a positive impact for health care in reduced serious violence related injuries, saving lives, and reducing costs in health and social care.

It is therefore in the public interest to provide policies to reduce Serious Violence, and to ensure that public funds spent are targeted to the areas that need it most, and where it can have the most impact.

Target dates are at the end of policies with evaluation cycles, likely annually but will be determined by publication timetables.

For the evaluation on Violence Reduction Units (VRUs) being conducted by University of Hull, annual reports will be produced, with analysis undertaken in one financial year planned for publication in autumn the following financial year. Any future use for evaluations conducted internally by the serious violence team are likely to follow similar timescales. Frequent monitoring will be reported on internally on a regular basis, with monthly monitoring to ministers, policy and analysts in the Home Office, and quarterly reports to minsters, other government departments, and police (all suppressed).

All of these outputs have the potential to benefit health and social care, with the monitoring to ministers leading to informed decision making, and engagement with the police force areas and local authorities to understand the drivers and the policies that may have an impact. The evaluations offer a more robust approach to assessing effectiveness of any funding or policies, to improve them or increase rollout based on results. The VRU programme is calculated to have monetised benefits of over £330 million during the next three years alone, but by having the unsuppressed data the Home Office will be able to assess whether the impact manifests. Communication to police forces and other government departments can lead to a more informed approach to tackling serious violence. These can all impact on health and social care, through reducing serious violence, and building knowledge on how to intervene effectively to prevent involvement in serious violence.

Timings for achievement of benefits also vary. For funding decisions, these are likely to use the data immediately, but will be implemented at the times of spending reviews (annually or every three years, depending on Treasury), or for one-off funding programmes.

For monitoring, impacts will be immediate in use, for understanding and communication. Impacts of evaluation of programmes, and understanding what works in serious violence, will be after a year initially, and annually thereafter. Actual reductions in serious violence are harder to predict, given the range of other factors that can also influence trends, and depend on the policy aims and targets.

Serious Violence is a high harm form of crime, and has wider impacts on society, including health, poverty, education, and mental health. Serious Violence has both direct and indirect effects on health and social care. Serious Violence leads to injury, which puts a pressure on healthcare, especially emergency services and hospitals. Indirect effects on healthcare include poor mental health of those affected by Serious Violence. Accessing the data helps the Home Office reduce Serious Violence. This has been demonstrated by the targeted funding of the Violence Reduction Units in areas with high SV volumes as measured by the data, and also by the VRU evaluation, which showed a statistically significant reduction in violence as measured using the relevant data.

It has:

1. Allow the development of policies to reduce serious violence, and the evaluation to ensure the effectiveness of such policies. It allowed better targeting of funding to reduce serious violence in areas which have high volumes of serious violence, reducing serious violence injuries requiring treatment within the health system and reduce impact on social care resources (as shown in the published VRU evaluation).

2. Informed Government response, and fed into ministerial decision making around policies and programmes as well as funding decisions, ensuring these reached areas and people most in need of support. This included high SV areas geographically, which led to the hotspot policing programme - this also showed a statistically significant reduction in violence in areas of high SV (the police forces in need were identified using the data source).

3. Been useful in disseminating knowledge gained to the public and to non-governmental research bodies to improve the evidence base and understanding of what works in relation to reducing serious violence. This has been done by publishing the above evaluations, as well as regular contact with public and non-governmental bodies to highlight these findings and improve responses. This will have a positive impact for health care in reduced serious violence related injuries, saving lives, and reducing costs in health and social care.

It is therefore in the public interest to provide policies to reduce Serious Violence, and to ensure that public funds spent are targeted to the areas that need it most, and where it can have the most impact. This has been demonstrated by how the data have been successfully applied and used in evaluations to date.