Archives

Measuring outcomes of safe accommodation: A rapid evidence review

Safe accommodation refers to safe, time limited accommodation for anyone escaping domestic abuse. It includes a range of settings such as refuge accommodation, specialist and dispersed accommodation, sanctuary schemes, and move-on or second-stage housing. In the short-term, support in safe accommodation provides immediate safety for victim-survivors fleeing domestic abuse, and, in the longer-term the aim is for victim survivors to be free from abuse and rebuild their lives.

The most recent national estimates from the Crime Survey for England and Wales showed that 3.8 million people aged 16 years and over experienced domestic abuse in year ending March 2025: 2.2 million females and 1.5 million males. The demand for safe accommodation for those experiencing domestic abuse persistently outweighs supply. Also, there is a shortage of safe accommodation and barriers to access for marginalised groups such as those with no recourse to public funds, or complex mental health and/or substance use need and/or disability access.

While existing evidence provides insight into demand and provision in the context of supported safe accommodation, it does not capture the outcomes experienced by survivors who access safe accommodation. Previous studies have highlighted the diversity and inconsistency of outcomes used to evaluate the impact of domestic abuse services on adults and children, and the need for a shared outcomes framework to develop a coherent national picture of what works.

In response, the Ministry of Housing, Communities and Local Government (MHCLG) commissioned a study to assess the feasibility of establishing a shared outcomes framework for support delivered in safe accommodation. The first stage of the research was a rapid evidence review to synthesise outcomes measured to assess the impact of support for victim-survivors and their children provided in the context of safe accommodation, conducted by Dr Annie Bunce (City St George’s, University of London), Dr Katie Smith and Dr Estela Capelas Barbosa (University of Bristol).

The review constitutes a targeted update and extension of a recently published scoping review of outcomes used to measure the effectiveness of domestic abuse support services conducted by VISION researchers (Carlisle et al., 2024), which identified 426 outcome measures across 80 studies. The current rapid evidence review was restricted to forms of supported safe accommodation, the original searches were updated to include studies published from June 2022 onwards, and barriers to and facilitators of implementing a shared outcomes framework were also reviewed. Seven academic databases were searched alongside three grey literature databases and 164 organisational websites. To be included, papers were required to be UK specific, published in English and focused on interventions delivered within safe accommodation. For the outcomes element of the review, papers were only included if they reported empirically measured outcomes (i.e., outcomes based on data collected and analysed by the authors, rather than purely theoretical, proposed, or assumed outcomes). 

Across the 17 reports meeting the inclusion criteria, outcomes clustered around three core domains: safety, wellbeing, and empowerment, which map broadly onto the Domestic Violence and Abuse Core Outcomes Set. Safety outcomes were the most frequently reported, followed by mental health and psychological wellbeing, while empowerment outcomes such as independence and confidence appeared less consistently. Service-level indicators such as length of stay and move-on destinations were also widely reported but fall outside existing frameworks.

This finding that outcomes measured in the context of safe accommodation broadly cluster around three core domains of safety, wellbeing and empowerment indicates some convergence in how impact is conceptualised. However, outcomes are measured inconsistently across services, with considerable variation in what is captured and how. Furthermore, significant barriers remain to implementing a shared outcomes framework. Developing a consistent approach to outcome measurement could strengthen accountability and support learning across areas.

Recommendations

  1. Frameworks should allow for the capture of relational and subjective aspects of recovery not easily measured through conventional quantitative indicators.
  2. Any national framework must be supported by appropriate data systems, training and analytical capacity to avoid placing additional burden on frontline staff and further disadvantaging smaller (particularly by-and-for) organisations and victim-survivors from marginalised communities.
  3. Future research should focus on further development and validation of measurement tools suitable in the context of safe accommodation, with particular attention to the relevance/meaningfulness of outcomes for victim-survivors themselves. Coordination between research teams is needed to avoid duplication and ensure data collection is harmonised between stakeholders.

To download the rapid evidence review: Measuring Outcomes of Safe Accommodation: A Rapid Evidence Review 

To download the main research report: Domestic_Abuse_Safe_Accommodation_Main_Research_Report.pdf

AI illustration from Adobe Photo Stock subscription

5th VISION annual conference: Presentations and illustrations now available

The fifth VISION annual conference was held on 7 July at The Light Auditorium in Friends House. The theme was Violence prevention in dialogue: Research and Policy.

After five years of research looking at violence prevention and the resulting health inequalities and focusing on improved measurement, policy, and practice, and delivering through a strong cross-sector partnership, we showcased our research in partnership with stakeholders Refuge, High Trees, Migrants’ Rights Network, Gloucestershire police constabulary, and Revolving Doors. In addition, we invited those working in violence prevention to critically assess our contributions and identify where further research and prioritised focus are required.

Jo Todd of Respect and Professor Gene Feder, VISION Director, reflected on 25 years of working in violence prevention as frontline practitioners and researchers. Drawing on their extensive experience, they highlighted both the progress made and the challenges that remain. While collaboration between government, academia, and the third sector has improved, longstanding silos—particularly across the health and criminal justice sectors—persist. Overcoming these barriers requires sustained commitment to relationship-building and cross-sector working.

They noted that significant progress has been made in violence prevention policy, including the publication of the government’s Violence Against Women and Girls (VAWG) strategy last year. However, there remains considerable interest in how the incoming Prime Minister will take forward the Labour government’s commitment to halve violence against women and girls within a decade, and what this will mean for future policy, practice, and investment.

Professor Liz Kelly from London Metropolitan University was our 2nd keynote speaker and shared her thoughts on the language and naming of violence in the 21st century. Liz highlighted that  language changes not just what we say but what we know and recognise. The words we use in the field as researchers, analysts, policymakers, officials, and practitioners matter and should be clear and grounded in victim / survivor agency and freedom.

VISION research was highlighted in three symposiums:

  • Insight: Key findings on definitions and measurement within violence prevention
    • Dr Vishal Bhavsar (Kings College London) and Professor Miranda Horvath (University of Suffolk) presented on intimate partner violence perpetration.
    • Dr Madeleine Janickyj (University College London) and Emma Pickering (Refuge) presented on technology facilitated abuse.
  • Inclusion: Violence prevention research with marginalised populations
    • Dr Alexandria Innes (City St George’s UoL), Aya Khedairi (Migrants’ Rights Network) and Professor Gargi Bhattacharyya (University College London) presented on insecure migration status. Andri and Aya discussed their project, Understanding Violence: the risks for migrants with rising far-right fascism and the resulting output, Workbook: Migrant Community Insights on Building Safety. Gargi highlighted how the far right is misusing genuine work and research and that media discussion of VAWG frequently descends into racism and incitement.
    • Dr Annie Bunce (City St George’s UoL) and Lauren Bennett and Anthony Donohoe from Revolving Doors presented their work on developing principles for co-analysis with Lived Experience.
  • Involve: A look at two ongoing, impactful pieces of VISION research with partners
    • Dr Elizabeth Cook (City St George’s UoL) and Harry Jenkins (High Trees) presented their collaborative research developing Built on Trust, a report with youth from Lambeth in London on the role of youth spaces and trusted adults in reducing young people’s exposure to violence.
    • Dr Ruth Weir (City St George’s UoL) and Sophie Jarrett (Gloucestershire Police) presented on their system-wide partnership with central and local government, police, education trusts, and the third sector to understand and prevent teenage relationship abuse (TRA). This work has resulted in the establishment of a national group on TRA and a local group wtihin Gloucestershire.

VISION co-Deputy Directors, Professor Sally McManus (City St George’s UoL) and Dr Estela Capelas Barbosa (University of Bristol) closed the conference by reflecting on the achievements of the VISION consortium and outlining priorities for the future of violence prevention research, policy, and practice. Sally highlighted the wide range of datasets, tools, resources, and publications developed by VISION over the past five years. Looking ahead, Estela emphasised the need for further research on violence affecting LGBTQI+ communities and minoritised women and children. She also stressed the importance of breaking down the persistent silos between the health and criminal justice sectors, noting that the impacts of violence extend across both systems. Effective prevention and support therefore require joined-up services, coordinated policies, and cross-sector collaboration.

Visual artists, Abigail and Chloe Baldwin, from Buttercrumble, beautifully captured the day in a series of illustrations which can be downloaded below along with the keynote and symposium presentations.

To download the presentations:

  1. Insight Symposium: Insight Symposium_Perpetrators of IPV; Insight Symposium_Tech Abuse
  2. Inclusion Symposium: Inclusion Symposium_Insecure migration status; Inclusion Symposium_Co-design of analysis principles
  3. Involve Symposium: Involve Symposium_Partnering with young people; Involve Symposium_Teenage Relationship Abuse
  4. Professor Liz Kelly, 2nd Keynote: Keynote 2_The language and naming of violence
  5. Professor Sally McManus, VISION Achievements: VISION achievements

To download the illustrations:

  1. Conference Summary
  2. 1st Keynote – Jo Todd and Gene Feder and 1st Keynote – 2nd illustration
  3. 2nd Keynote – Liz Kelly
  4. Insight Symposium – Data and Definitions
  5. Inclusion Symposium – Marginalised Populations
  6. Involve Symposium – Partnerships
  7. VISION Achievements
  8. Violence Prevention in Dialogue

We keep each other safe – a guide for migrants in the UK

 

A negative shift in how migrants, including refugees and people seeking asylum, are viewed in the UK has been documented in research, including VISION work, and media. Rhetoric about migration has become more aggressive and emboldened racist demonstrations and attacks on asylum accommodation.

VISION partnered with Migrants’ Rights Network (MRN) to co-develop a research project with migrants that maps experiences of harm and identifies community-led safety strategies. The research centred on two cities, London and Belfast, working with communities who have experience of the asylum system / no recourse to public funds. Through a series of focus groups and workshops, migrants shared their experiences of far-right violence, and how they are already keeping themselves and each other safe. ​

The project brings together MRN’s narrative work, which actively challenges disinformation about migration, while trying to better understand and document the impact of hostile language on people currently in the immigration system.

The key findings and themes​ of the research are:

  • Impact on mental health and change in behaviours ​
  • Role of the state in making migrants unsafe, and low trust in institutions​
  • Everyone is experiencing it, but no one is talking about it​
  • We keep each other safe – what are we already doing, and what else can we do together

These insights coalesced into a free workbook available for downloading on the MRN website. The practical guide is for migrants in the UK co-developed with migrants and migrant-led organisations. It features shared knowledge, scenarios, and messages of solidarity.

To download the free resource: We keep each other safe workbook: Migrant community insights on building safety

For further information on the VISION – MRN collaboration, please see the blog, Understanding Violence: The risks for migrants with rising far-right fascism, written by Aya Khedairi, MRN Community Engagement Manager.

AI photo from Adobe Stock subscription

The link between Adverse Childhood Experiences and school engagement

By their eighth birthday, an estimated nine in ten New Zealand children will have experienced a form of serious adversity. They might have been neglected, grown up with family violence, lived through a separation, experienced bullying or coped with a parent’s mental illness or substance use problem. These experiences are relevant to education as they can have important and lasting consequences for academic achievement. When a child faces more adversity, it weighs heavily on how they fare at school. School engagement has a significant impact on students’ academic success underscoring the need for a thorough examination of the risk factors contributing to student disengagement within educational environments. The research has been less clear about why that link exists.

A recent study led by Maryam Ghasemi at the University of Auckland, with VISION Senior Research Fellow Ladan Hashemi (City St George’s UoL) and others, examines the longitudinal association between a relatively broad Adverse Childhood Experiences (ACEs) index within Growing Up in New Zealand (GUiNZ) measured up to age 8, and low school engagement at age 12.

The team also explores the potential mediating role of children’s self-esteem and scholastic self-concept in this relationship. Using data from an ethnically diverse longitudinal birth cohort of 3858 children in New Zealand, they investigated how early exposure to adversity influences school engagement at age 12.

The research demonstrates that exposure to a greater number of adversities by age 8 is associated with lower levels of school engagement at age 12, even after adjusting for gender and area deprivation level. This effect occurs by undermining children’s self-esteem and scholastic self-concept, which highlights how early adversity can shape educational trajectories through internalized negative self-perceptions.

The findings highlight the importance of preventing ACEs, early identification of at-risk children, and prioritizing trauma-informed, equitable, and culturally sensitive interventions to improve school engagement. Children who stay engaged at school are more likely to do well later in life. Those who disengage face a greater risk of lower academic achievement, poorer mental health and leaving school early. To keep children connected to their school, supportive teachers, positive friendships, opportunities to succeed and extracurricular activities can all make a meaningful difference. So can giving students more say in their learning and recognising their achievements. Ultimately, schools cannot undo the adversity a child has faced at home. But they can still do much to help children believe in themselves, thrive in the classroom and reach their potential.

To download the peer-reviewed article: Understanding the Link Between Adverse Childhood Experiences (ACEs) and School Engagement: Investigating the Role of Child Self-concept as a Mediator

To download the article in The Conversation: Adversity can follow NZ kids to the classroom. Can schools make a difference?

To cite: Ghasemi, M., Meissel, K., McIntosh, T. et al. Understanding the Link Between Adverse Childhood Experiences (ACEs) and School Engagement: Investigating the Role of Child Self-concept as a Mediator. ADV RES SCI7, 43 (2026). https://doi.org/10.1007/s42844-026-00229-z

Photograph from Adobe Photo Stock subscription

Using Natural Language Processing in Domestic Homicide Reviews

Since 2011, there has been a statutory requirement in England and Wales to conduct a Domestic Homicide Review (DHR) into any domestic abuse-related death: a multi-agency review into the death of a person aged 16 or over that appears to have resulted from violence, abuse or neglect from an intimate partner, family member or household member.

However, analyses of large numbers of DHRs are rare. One of the core challenges is the time and effort required to analyse narrative text within reports. Doing so manually is both time-consuming and resource-intensive and is a primary reason why researchers typically focus on only a portion of the available data. Natural Language Processing (NLP)—a sub-branch of artificial intelligence that enables computers to interpret and process natural language—provides a viable and scalable alternative by offsetting much of the heavy data processing to a computer.

In this study protocol, developed by VISION Research Fellow Dr Darren Cook and VISION Co-Investigator Dr Elizabeth Cook (both at City St George’s University of London) with Sumanta Roy and Rani Selvarajah of Imkaan, and VISION Co-Investigator Professor Ravi Thiara (University of Warwick), they outline a study to assess the feasibility of applying NLP to DHRs.

The VISION and Imkaan team outline a collaborative approach which balances the speed and scale of automation with the embedded knowledge and expertise of practitioners. This approach helps to ensure that outputs of NLP are sensitive and transparent about the biases common within datasets on violence and abuse.

Based on initial consultations, the team have identified a series of priority research questions for investigation. In addition, they outline details of an ongoing collaboration with one partner, Imkaan. The protocol describes the data access, and retrieval and analysis stages before summarising how feasibility will be evaluated. The protocol concludes by arguing that working with practitioners who hold deep contextual knowledge about the social realities of violence and abuse, including language, risks, and experiences, means that tools can be developed that are accountable to communities and appropriately applied to real-world problems.

To download the protocol: A collaborative approach to applying Natural Language Processing (NLP) to Domestic Homicide Reviews (DHRs): A study protocol

To cite: Cook D, Cook EA, Roy S, Thiara R, Selvarajah R (2026) A collaborative approach to applying Natural Language Processing (NLP) to Domestic Homicide Reviews (DHRs): A study protocol. PLoS One 21(5): e0348948. https://doi.org/10.1371/journal.pone.0348948

For further information: Please contact Lizzie at elizabeth.cook@citystgeorges.ac.uk

Illustration from Unsplash website

Commissioning Pathways for Domestic Homicide / Abuse-Related Death Reviews: Are All Deaths Counted?

Domestic homicide reviews (being renamed domestic abuse-related death reviews) have been undertaken in England and Wales since 2011. However, relatively little is known about the commissioning process for these reviews, including where notifications come from, if the types of cases being referred are changing, and the outcomes. Knowledge is also limited about who is involved in these decisions and who is informed when a decision is made.

For this project, Dr James Rowlands (University of Durham), VISION Co-Investigator Dr Elizabeth Cook (City St George’s University of London) and research consultant Dr Althea Cribb, used data requested from the partnership bodies responsible for commissioning domestic homicide / abuse-related death reviews about notifications and decision-making between January 2017 and December 2024. Their findings highlight the changing profile of cases, variability in decision-making, and gaps in communication and oversight.

To download the paper: Commissioning Pathways for Domestic Homicide / Abuse-Related Death Reviews: Are All Deaths Counted?

To cite: Rowlands, J., Cook, E., & Cribb, A. (2026, May 11). Commissioning Pathways for Domestic Homicide / Abuse-Related Death Reviews: Are All Deaths Counted?. https://doi.org/10.15128/r1kp78gg500

Illustration from Adobe Photo Stock subscription

Violence, Injustice, and Inequality: The State of International Migration

On June 14, 2023, a fishing trawler with over 700 asylum seekers and refugees predominantly from Syria, Palestine, Pakistan, Egypt and Afghanistan on board, sank in the Messenia region of the Mediterranean. The travelers on board had no food or water, the engine was overheating, and a number of people on board were alleged to have died. The boat eventually capsized and sank after floundering for two days in Greek search and rescue waters while under observation by the Hellenic Coast Guard. During this time no rescue operation was launched. It was only after the boat capsized that a search and rescue operation occurred. Just over 100 men were rescued, and 84 bodies were recovered from the water, including all the women and children on board.

On June 18, 2023, a private submersible vessel launched on a sightseeing tour of the Titantic shipwreck. The small vessel was occupied by four passengers and the captain. The passengers on board had paid approximately £250,000 each for the voyage. Shortly after the vessel began its dive, contact was lost and soon afterward a global search and rescue operation was launched. Later, debris of the submersible was confirmed on the ocean floor, and all five occupants were declared dead.

The temporal juxtaposition of these two maritime disasters, and the disparity in both the number of lives lost and the effort spent to save the lives of those on the vessels throw international inequality and injustice into sharp relief. While there is a huge and obvious wealth disparity between the two groups cited in the examples above, economics alone cannot explain the contrast between the framing of and responses to these two disasters.

In her latest publication, Violence, Injustice, and Inequality: The State of International Migration, VISION Co-Investigator Dr Alexandria Innes focuses on the gross inequality that is evident in international migration governance and management. Framed in the concept of inequality as it is variously yet incompletely theorized in international relations scholarship and drawing on Ranciere’s equality as practice, she situates violence as an indicator of inequality that reveals injustice.

Using a case study of domestic violence in the context of the UK’s “hostile environment,” Andri demonstrates how states, exemplified by the UK, adopt domestic violence as a mechanism of immigration deterrence. She argues that, despite the acceptance of domestic violence as a social wrong, and the evidence that domestic violence is pervasive in society, migrant women in insecure status are denied access to necessary forms of protection, which leads to prolonged exposure to domestic violence and reveals continuous violence against migrants in insecure status.

Attending to violence, and in particular state violence, in the global politics of migration reveals the injustice of embedded inequality in the international system. While injustice is immediately legible in violent events, injustice is also embedded in the unequal social order, continuously ordering and bordering protection from and submission to violence.

To download the paper: Violence, Injustice, and Inequality: The State of International Migration

To cite: Alexandria Innes, Violence, Injustice, and Inequality: The State of International Migration, Global Studies Quarterly, Volume 6, Issue 2, April 2026, ksag058, https://doi.org/10.1093/isagsq/ksag058

Photograph from Adobe Photo Stock subscription

Intimate partner relationship abuse between men: An interdisciplinary seminar

 

Wednesday 17 June 2026, 13:00  – 16:45, in person only

This event is in the past.

This free, interdisciplinary seminar brings together leading voices from research, lived experience, performance, and specialist practice to examine intimate partner violence in relationships between men and consider implications for policy, research, professional practice, and wider system response.

Register here: Ticket Tailor

Despite increasing recognition that abuse within intimate male relationships is a significant public health, social justice, and service delivery issue, it remains under recognised across many systems and disciplines. Survivors frequently encounter barriers to recognition, disclosure, and support, while policy, professional training, and public discourse have historically paid limited attention to these experiences. 

This event will explore the dynamics, impacts, and structural challenges surrounding abuse in this context, and consider how research, policy, practice, and public understanding can evolve to better respond. 

The programme combines research presentations, creative performance, lived experience testimony, and expert commentary to offer a multi-dimensional exploration of the issue. The seminar will include catering and: 

  • A brief research presentation on prevalence, dynamics, barriers to recognition, and current evidence gaps from Dr Steven Maxwell 
  • A monologue from He Kept Me Safe, a research based verbatim play developed from survivor narratives exploring abuse within intimate male relationships by Dr Edgar Rodriguez-Doran 
  • A keynote contribution from broadcaster, comedian and survivor advocate James Barr, reflecting on his lived experience through comedy and the role of storytelling 
  • A talk from Tanaka Mhishi, writer and researcher in masculinity, trauma and sexual violence, exploring masculinity, trauma, violence, and engaging men in conversations about healing 
  • Brief facilitated interdisciplinary discussion on implications for research, policy, education, commissioning, service design, and practice 

This event aims to stimulate critical interdisciplinary dialogue and contribute to wider thinking on inclusive responses to domestic abuse and sexual violence. 

 

Intended Audience / Invitees 

The seminar is intended for a broad interdisciplinary audience including: 

  • Researchers and academics across social science, health, psychology, gender studies, criminology, law, public health, and related disciplines 
  • Policy makers and strategic leaders working in domestic abuse, sexual violence, health, justice, equalities, LGBTQ+ inclusion, and public policy 
  • Health and social care professionals 
  • Mental health practitioners and psychological therapists 
  • Domestic abuse and sexual violence services 
  • Policing, criminal justice, and legal professional 
  • LGBTQ+ and community sector organisations 
  • Sexual health and public health professionals 
  • Education and professional training providers 
  • Students and trainees with relevant academic or professional interests 
  • Wider stakeholders with interest in trauma, masculinity, violence prevention, and inclusive service/system design

 

Programme 

1.00 pm: Registration and Lunch 

1.30 pm: Welcome and Opening Remarks 

1.40 pm: Research Presentation: Understanding Intimate Partner Violence in Male Relationships: Prevalence, Dynamics and System Challenges 

2.00 pm: Monologue: He Kept Me Safe 

2.40 pm: Facilitated Discussion and Audience Reflection 

3.00 pm: Break 

3.15 pm: James Barr: Lived Experience, Comedy, and Public Discourse 

3.45 pm: Audience Q&A 

4.00 pm: Tanaka Mhishi: Masculinity, Trauma and Engaging Men in Conversations About Violence and Healing 

4.30 pm: Panel Discussion / Closing Audience Q&A 

4.45 pm: Close 

Join us at this free seminar, 17 June, Northampton Suite, City St George’s UoL, Clerkenwell, London, 1300 – 1645. To book your place please register here: Ticket Tailor 

Photograph provided via Adobe Stock subscription

Measuring perpetration of intimate partner violence: A systematic review

Understanding how common intimate partner violence (IPV) is in different groups and settings can inform better policies and interventions to prevent IPV and reduce its impact on public health. Much of the existing research on IPV has measured IPV victimisation in populations. But IPV reported by victims, e.g. in nationally representative crime or health surveys, reflects the behaviour of perpetrators. Asking people in surveys about their perpetration of IPV, and how frequently harmful behaviour(such as physical, sexual, or psychological harm) occur, is also important.

A new systematic review by VISION collaborator Vishal Bhavsar and VISION co-investigator Sian Oram gathers evidence on how IPV perpetration is measured in health surveys of the general population and presents a set of measurement principles. Based on 39 published articles, they find 27 health surveys containing information on IPV perpetration. These measures vary in detail, ranging from one or two binary items to more comprehensive scales. In many cases, the more detailed measures are adaptations of victimisation items reframed to capture perpetration, while a smaller number of studies use instruments specifically developed for IPV perpetration. Information relevant to understanding the public health implications of IPV perpetration, such as the sex and number of victims per perpetrator, the nature of relationships (e.g. marital, cohabiting), patterns of frequency or escalation, and harmful impact on victims, is not consistently collected. Details on the development and validation of measurement items are also often lacking.

Distinguishing IPV perpetrators from victims in measurement is important for the development of interventions, particularly where policies or programmes are designed to engage individuals who perpetrate IPV. Although health surveys do include some measures of perpetration, the extent and detail of these measures are limited. Some of the principles developed and applied in the review might also apply to the measurement of IPV perpetration and other forms of domestic abuse in electronic health records or policing data. A future cross-sector consensus on how to measure IPV perpetration might be helpful in informing the societal response to preventing IPV.

To download the paper: The Measurement of Perpetration of Intimate Partner Violence in General Population Health Surveys: Systematic Review

To cite: Bhavsar, V., Oram, S. The Measurement of Perpetration of Intimate Partner Violence in General Population Health Surveys: Systematic Review. Trauma, Violence, & Abuse. OnlineFirst, March 31, 2026.  https://doi.org/10.1186/s40163-026-00272-2

Illustration from Adobe Photo Stock subscription

Using AI to investigate publicly available documents on violence prevention

 

Artificial intelligence (AI) systems are increasingly applied in public health, yet their use for analysing fragmented, multi-sectoral policy landscapes remains underdeveloped. Many applications have focused on service delivery, such as AI-powered chatbots, data surveillance and monitoring, and tracking social media interactions for emerging risks, with less attention paid to how AI might support policy analysis. This is especially true for the violence prevention sector, where AI is gaining traction as a solution for triaging help-seeking calls, detecting threatening messages, predicting conflict and improving police data, but not for understanding the policy landscape.

Policy responses to violence are undergoing scrutiny in the UK, coinciding with the recent publication of an updated cross-government strategy addressing violence against women and girls. This renewed focus places increased demands on researchers and policymakers to rapidly synthesise large and fragmented bodies of policy evidence spanning multiple sectors and both local and national government. Traditional approaches to policy review formed around a wholly manual approach may struggle to meet these demands within policy-relevant timeframes.

This research, an exploratory, proof-of-concept case study, aimed to describe the development and preliminary exploration of an AI-enabled tool designed to synthesise evidence from violence-related policy documents in the UK. The team was led by VISION Research Fellow Dr Darren Cook and inlcuded several members from the wider VISION consortium, Dr Elizabeth Cook, Kimberly Cullen, Professor Sally McManus, Professor Gene Feder and Professor Mark Bellis. 

For their article, Artificial intelligence in critical synthesis of public health responses to violence: A novel application to UK violence prevention policy, the team compiled a corpus of publicly available UK policy and strategy documents on violence (N = 343) through expert review, manual searches of government and third sector organisation websites, and automated web scraping.

Then, they used the corpus to train an existing AI framework and deployed it through a question-answer interface. Stakeholders working in violence prevention (academics, practitioners in specialist services and government officials) were invited to pose natural-language questions about violence policy and consider the system’s utility and the usefulness of its outputs. Their feedback indicated that the AI generated reports were well-grounded in the underlying source documents. Syntheses aligned closely with the documents in the tool, and the inclusion of document references and page-level citations supported credibility assessments. Corpus coverage statistics were considered particularly helpful when judging the robustness of responses. 

This research contributes by documenting the early application of an AI-enabled tool designed to support exploratory policy analysis. The team illustrates an emerging analytic capability and its potential role within policy-oriented research workflows. By demonstrating how a document-grounded, closed-domain AI system can be used to interrogate policy framings and identify potential siloes, this work addresses a gap in current public health applications of AI, specifically in the context of violence prevention.

To access the VISION AI tool to ask your own questions about violence prevention: VISION: Violence, Health & Society  

To download the paper: Artificial intelligence in critical synthesis of public health responses to violence: A novel application to UK violence prevention policy

To cite: Cook, D., Cook, E., Cullen, K., Zachos, K., McManus, S., Feder, G., Bellis, M., Maiden, N. Artificial intelligence in critical synthesis of public health responses to violence: A novel application to UK violence prevention policy. Science Direct (2026). https://doi.org/10.1186/s40163-026-00272-2

Illustration from Adobe Photo Stock subscription