Protocol: A Quasi-Experimental Effectiveness and Cost-Effectiveness Evaluation of Emergency Department Violence Intervention Programmes in the United Kingdom

Protocol: A Quasi-Experimental Effectiveness and Cost-Effectiveness Evaluation of Emergency Department Violence Intervention Programmes in the United Kingdom | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Study protocol Protocol: A Quasi-Experimental Effectiveness and Cost-Effectiveness Evaluation of Emergency Department Violence Intervention Programmes in the United Kingdom Simon C. Moore, Sinead Brophy, Amrita Bandyopadhyay, Annemarie Newbury, and 14 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5452363/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Introduction Hospital-based violence intervention programmes (HVIPs), based in Emergency Departments (EDs) have been proposed as a public health response to violence. These programmes address the underlying reasons why patients are exposed to violence. In addressing any underlying modifiable risks and vulnerabilities HVIPs can reduce patients’ exposure to violence and therefore subsequent unplanned attendance into ED. Methods and Analysis ED patients are eligible for inclusion in the evaluation if they are normally resident in Wales, United Kingdom (UK), aged 11 years and older. A controlled longitudinal natural experiment will be undertaken. The primary outcome is derived from the Emergency Department Dataset, routinely collected for all EDs in Wales, and is subsequent unplanned ED attendance. Case patients will be matched to control patients attending EDs without an HVIP. Analysis will derive the hazard rate for subsequent unplanned ED attendances using recurrent event analysis. The total monthly count of patients identified as attending because of violence in intervention EDs will be compared to the total count of Welsh control EDs in an interrupted time series analysis to determine whether HVIPS increase violence ascertainment. To determine whether referral, versus no referral, to the HVIP represents value for money, we will undertake a cost-effectiveness analysis from the perspective of the National Health Service. Ethics and Dissemination The approval to access and analyse data housed in the Secure Anonymised Information Linkage (SAIL) databank, an ISO 27001 certified and UK Statistics Authority accredited secure data environment, was granted by the SAIL independent Information Governance Review Panel (Ref: 1421). Findings will be presented at local, national, and international conferences and disseminated by peer-review publication. ISRCTN Registration : 41868 (12 August 2022) Emergency Department Assault Violence Injury Prevention Treatment Figures Figure 1 Figure 2 Strengths and Limitations of this Study A whole population, controlled evaluation of a violence prevention team situated in an Emergency Department (ED) that accounts for patient characteristics that might influence outcomes. Provides descriptive analyses concerning patient characteristics associated with their willingness to engage in the violence prevention intervention. Describes analytic methods for evaluations utilising routine ED data. The analysis can only be conducted on patients identified in routine data and linked to the intervention data. Some patients who wish to avoid scrutiny may avoid providing details necessary for linkage. Introduction Those who experience serious injury due to violence are likely to attend an Emergency Department (ED). EDs are therefore ideal locations for hospital-based violence intervention programmes (HVIPs) 1–3 . HVIPs have recently emerged as a public health response to violent victimisation 3 4 , but despite interest in HVIPs there has been no rigorous evaluation of this public health approach to violence in the United Kingdom (UK). Moreover, little is known about the effectiveness of patient discharge planning and referral from ED into organisations able support children involved in violence 5 , patients exposed to domestic violence 6 7 , and there is a paucity of studies considering referrals for young men involved in violence, the most dominant population in respect of assault-related attendance (ARA) 8 . There are even fewer studies of the effective support available to victims of sexual violence attending ED 9 . Despite uncertainty over effectiveness in a UK context, HVIPs have begun to be implemented. For example, the Scottish Violence Reduction Unit (VRU) has placed Navigators ( www.mav.scot/navigator ) in EDs, typically youth workers who connect with patients 25 years of age and younger 10 . EDs work with patients attending for many reasons, including those presenting with violence-related injuries. EDs also engage in broader violence reduction initiatives: clinical staff typically receive training in adult and paediatric safeguarding, and many EDs have provisions to identify and refer children, and victims of domestic and sexual violence. However, methods of ascertainment and referral vary considerably and formal relationships with the police and other partners can often lack continuity with multiagency approaches not capturing the entirety of patients’ journeys. A public health approach involving the ED would be beneficial in improving overall population health outcomes. It is particularly timely, in 2019 there were 175,764 ARAs at EDs in England and Wales 11 and, despite the pandemic, 119,111 ARAs in 2020 12 . While knife crime, and therefore serious trauma, has risen by 71% from 2014 to 2018 13 . Up to 75% of ARAs are unknown to the police 14–16 , therefore EDs hold exclusive data on assault characteristics, patient vulnerabilities and modifiable risks, and are therefore well situated to play a significant role in the identification of violence, to investigate the circumstances of violence, and to challenge any underlying vulnerabilities or modifiable risks exposing patients to violence, whether that is through direct support, referral, or discharge planning. The need for HVIPs is aligned to broader UK Government initiatives, which aim to promote a whole system multi-agency (WSMA) 17 approach to violence. The 1998 Crime and Disorder Act requires the police, local government, and the National Health Service (NHS) to collaborate on joint crime reduction strategies and this includes data sharing to inform targeted responses. Violence reduction is further prioritised by the UK Government in its Serious Violence Strategy 18 and the UK government has allocated funds for the formation of VRUs in England and a Violence Prevention Unit (VPU) in Wales, across 18 Police and Crime Commissioner jurisdictions, with the explicit purpose of promoting the WSMA approach 19 . These initiatives are further aligned with a move towards active population health management, digitally enabled whole-person care and evidence-based treatment pathways outlined in the NHS future plan 20 . Integrated Care Systems in NHS England will be expected to specify violence prevention and reduction standards, which are incorporated into the 2021/22 NHS Standard Contract, and there are expectations that hubs will form Violence Prevention Teams similar to the police VRUs and VPU. Furthermore, a public sector duty on partnerships encouraging the prioritisation of reducing serious violence has received royal assent as a part of the Police, Crime, Sentencing and Courts Bill. This legislation includes a serious violence duty placing a statutory obligation on organisations to collaborate, communicate, and act. The overarching aim of the work proposed here is a robust effectiveness and cost-effectiveness evaluation of ED-based Violence Prevention Teams (VPTs). VPTs represent a formal collaboration between police and healthcare and embody the WSMA approach. To our knowledge, this is the first formal evaluation of a nurse-led, ED based HVIP in the UK and will address significant gaps in current understanding of their effectiveness and thereby facilitate future aspirations for evidenced-based referral pathways and discharge planning 20 21 . VPTs main function is to identify and support patients attending ED with assault-related injury. To facilitate this, they engage in broader pedagogical roles increasing awareness of these patients’ needs, modifiable risks and opportunities to identify and refer across the ED clinical environment. 3.1. Theoretical Framework The theoretical motivation for a WSMA approach to HVIPs is that there are many modifiable risks and vulnerabilities that, in combination, determine an individual’s exposure to violence and subsequently an ARA in an ED. Epidemiologically, these can be usefully described by shared circumstances that in turn signpost opportunities to modify risk or support patients’ vulnerability, but responsibility can fall across organisations, including local government, healthcare, and criminal justice. Risks include the consumption of alcohol and other psychoactive substances 22–25 ; criminal and/or sexual exploitation and homelessness 3 26 . Violence tends to be more prevalent in younger, socially disadvantaged groups 27–33 , with male, socio-economically deprived individuals being more likely to endure violence and experience assault-related injury. These characteristics further extend to personality features 34 , including mental health status and learning disability, and neurodevelopmental disorders 35 . This complex interplay of factors that promote exposure to violence, and hence lead to an ARA, highlight the need for a WSMA approach. For example, an environment might become synonymous with violence through a process of homophily 36 , whereby individuals with shared pursuits who are at risk of violence gather, for example street drinkers and late night drinking environments. Mitigation might include challenging reasons for frequenting such an environment, including alcohol and substance misuse counselling. Some environments might involve those who use violence to advance their interests, such as acquisitive crime, sexual assault, or sexual exploitation, in which case criminal justice or safeguarding processes to deter violence might be involved, along with support to victims. Chaotic or otherwise disadvantaged households in which domestic violence or harm to children arises might best be approached from a multi-agency process such as the Multi-Agency Risk Assessment Committee (MARAC) and formal investigation (Section 47, Children Act 1989). EDs are primary agencies receiving those who have sustained a serious injury, including those who are motivated to bypass other agencies or whose assailant is motivated to ensure their victim avoids scrutiny. Treatment for an ARA in ED aims to address symptoms (e.g., injury) that may not necessarily characterise the underlying reasons for violence (e.g., alcohol dependency), and staff do not always have the resources available to address such modifiable risks and vulnerabilities. However, without addressing them, the risk of repeat unscheduled ED attendance remains, including violence recidivism. For these reasons, services like VPTs that work within a WSMA approach to better understand reasons for ARA are required. Moreover, and for those who are most vulnerable, ED may be the only realistic opportunity for patients to enter a system of care. As such, an ARA is often a sentinel event. Intervention A process and implementation evaluation that describes both the planned and implemented VPT intervention is available elsewhere 37 , and is further described in a Template for Intervention Description and Replication (TIDieR, Appendix 20.1). 4.1. Intervention as Hypothesised VPTs, which emerged from the VPU violence prevention strategy, were funded by the UK Home Office and Youth Endowment Fund (YEF) with the funding administered by the VPU and the Office of the South Wales Police and Crime Commissioner (PCC). Other HVIPs in the UK are volunteer-based, whereas the VPTs are nurse-led. The original implementation for VPTs were focussed on identifying and supporting ED patients aged 11 to 25 years of age and to formalise the identification of modifiable risks and vulnerabilities, to support and advise patients, and to signpost to other services as appropriate. The VPTs also aimed to raise awareness of the service across ED clinical teams, with the aim of it becoming embedded within usual practice, and to train and upskill the clinical team to enable ascertainment and referral. 4.2. Intervention as Implemented Since November 2019, a collaborative VPT between the police and NHS has been operational in a South Wales Type I (consultant led with resus) ED in Cardiff (the capital and largest city in Wales) and a second VPT began in an adjacent South Wales Type I ED in Swansea (the second largest city in Wales) in January 2023. The VPTs initially sought to identify patients attending the ED due to violence. This remit was broadened with VPTs subsequently receiving referrals from across the hospital and other community healthcare teams (e.g., MIUs and GPs). The VPTs work with patients to gain an understanding of any circumstances contributing to their exposure to violence. They refer patients into care pathways (primary, secondary, and tertiary care, or third-sector organisations) to address any vulnerabilities or modifiable risks and can work alongside third sector (non-profit and charitable enterprise) to provide continual case-management. The VPTs also train other staff within the hospital to improve the identification of violence-related injury, to support clinical staff interactions with patients, and to maintain safeguarding procedures. In addition, the EDs at Cardiff and Swansea take part in Information Sharing to Tackle Violence (ISTV), in which anonymous data and intelligence regarding violent incidents are shared with Community Safety Partnerships. These anonymised data enable partner resources to be best used for violence prevention, part of the Cardiff Model for violence prevention 38 . Following implementation, both VPTs expanded the age range of patients to encompass all age groups. 4.3. Usual Care Under usual practice, clinical ED staff are obliged to undertake safeguarding activities, and provide for those attending due to violence. However, provision varies across EDs. Under usual practice, when people attend ED with an injury suspected to be caused by violence, their injuries are treated and the patient is encouraged to contact the police, or have the ED contact the police on their behalf. In cases of serious injury involving weapon use, the ED is obliged to contact the police irrespective of patient consent. In terms of general safeguarding, all patient-facing clinical staff are expected to have up-to-date safeguarding training, and thus to carry out safeguarding tasks. Patients who are experiencing domestic violence can be referred to an Independent Domestic Violence Advocate (IDVA). Patients attending due to sexual assault can be referred to an Independent Sexual Violence Advisors (ISVA). For the ten control EDs in Wales, two EDs have an IDVA, and two others have access to an IDVA not based in their ED. Furthermore, ED staff can also refer patients into a MARAC, typically cases where the criteria are not met for formal safeguarding but the clinician suspects that something is not right. To facilitate, Multi-agency Referral Forms (MARFs) are filled out for children who require safeguarding and VA1 forms (to support the referral of vulnerable adults) are completed for vulnerable adults who require safeguarding. There are IDVAs based in the intervention EDs in Cardiff and Swansea. Only Cardiff has an ISVA. Broadly, usual practice focuses on children and victims of domestic violence. The patients eligible for VPT support are therefore those who are not eligible for support from the IDVA or ISVA and are typically over ten years of age. Apart from the IDVAs in control EDs, none have additional resources specifically dedicated to the role of supporting patients attending due to violence, relying mainly on existing clinical staff to support safeguarding within their departments. This may involve naming an existing member of staff as a safeguarding ambassador or having a nurse act as safeguarding lead for the department. Across Welsh EDs, some control EDs have provisions for victims of domestic violence that includes cards with the “Live Fear Free” helpline that they can give to patients experiencing domestic violence and who do not meet the criteria for a MARAC referral. Similarly, none of the control EDs have processes in place to support patients’ referral to outside agencies. If patients disclose that they are struggling with issues, or if staff suspect patients are experiencing an issue, then referrals will made by clinical staff. However, this is not the same as VPT members working with patients to identify modifiable risks and vulnerabilities that contribute to their experience of violence. The VPTs in Cardiff and Swansea are therefore unique. Aims and Objectives The overarching aim of the Emergency Department Violence Intervention Programme (EDVIPE) evaluation is to determine whether VPTs are effective and cost-effective from the perspective of the NHS. Objective 1 To assess whether patient involvement with a VPT reduces the likelihood of unscheduled ED re-attendance. We consider case and control patients’ ED unscheduled reattendance for a minimum of 12 months following the initial ARA. Objective 2 To determine whether the presence of the VPT improves ascertainment of ARAs in ED attendances. We will consider the change in identified ED ARAs across intervention implementation in case and control EDs in Wales. Objective 3 To derive the costs of the VPT and compare those to the benefits of the intervention and understand whether the VPT represents value for money from an NHS perspective. If an effect is observed, then models will estimate the health impacts, costs and potential savings over a longer time (e.g. 10 years) period and for a national roll-out. Ethics The approval to access and analyse data housed in the SAIL Databank 39 , an ISO 27001 certified and UK Statistics Authority accredited secure data environment, was granted by the SAIL Independent Information Governance Review Panel (IGRP) (Ref: 1421). The IGRP comprises representatives from various organisations and sectors including the British Medical Association, Welsh Government, Public Health Wales, National Research Ethics Service, Digital Health and Care Wales (DHCW), Swansea Bay University Health Board, and members of the public. All routinely collected anonymised data held in SAIL are exempt from consent due to the anonymised nature of the databank (Section 251, Control of Patient Information; 2006 National Health Service Act). At no time will identifiable data be made available to the research team. ED staff will curate data pertaining to patients’ exposure to the intervention, which will be passed to DHCW, where it will be anonymised, and a project specific anonymous linkage field (ALF) added, as will a residential anonymous linkage field (RALF). These data will be passed to SAIL for linkage to anonymised VPT clinical data. Participant consent is not required because all the study outcome data involving patients are anonymised before they are incorporated into the SAIL databank. As the SAIL databank is fully anonymised, it does not fall into the remit of the National Information Governance Board who provide section 251 (formerly section 60) exemption to use identifiable data without consent. Human Ethics and Consent to Participate declarations are not therefore applicable. Patient and Public Involvement Extensive Patient and Public Involvement and engagement (PPIE) has been and will continue to be undertaken. The rationale is that many patients managed by the VPTs will be vulnerable, with some at the beginning of their journey in the support they receive. The expectations were that these patients would be unlikely to reflect meaningfully on the VPT within the study timeline and therefore alternative opportunities to explore patients’ perceptions was required. Furthermore, follow-up qualitative work with young adults in emergency care, the dominant group in ED, requires considerable resourcing and suffers from high levels of attrition 40 . We therefore sought PPIE engagement in order that those with experience of the emergency healthcare system were able to feed into the project, co-produce methods, provide their interpretation of the results and assist with interpretation and dissemination. Groups include survivors of domestic violence, carers, those who have experienced alcohol and drug dependence, homelessness, sexual exploitation, and mental health issues. One PPIE co-investigator was appointed to lead on monitoring equality and diversity, with a second, and experienced, PPIE co-investigator supporting inexperienced PPIE members. PPIE activity was captured using the short-form Guidance for Reporting Involvement of Patients and the Public (GRIPP) 41 . The PPIE groups include lay members with experience of PPIE work: Service Users for Primary and Emergency Care Research (SUPER; www.primecentre.wales/ppi.php ) who provided lay perspectives to the research team when developing, and conducting the research, with subsequent engagement undertaken to strengthen the relevance, quality and dissemination opportunities of the research. Additional PPIE members were recruited. These members had lived experience relevant to the patients who are the subject of the intervention (homelessness and domestic violence). SUPER and the PPIE co-investigators advised on how the research team should engage with those who have lived experience but had not have prior experience of PPIE involvement. 7.1. PPIE Activity SUPER provided feedback on the original proposal, and then provided advice on how the materials should be developed for the two less experienced PPIE groups. These PPIE groups with lived experience were recruited to give input on the protocol. One group consisted of people with lived experience of homelessness and related conditions, recruited from a charity supporting those who are experiencing homelessness. The second group was comprised of survivors of domestic violence and were recruited through Welsh Women’s Aid. The PPIE group members gave feedback on the VPTs and provided their first impressions of the research protocol. The groups also helped to develop a list of potential organisations to disseminate research findings, and discussed methods on how this could be achieved. They explained what may be going on in the lives of people who experience violence, and where people might turn to for support. 7.2. Results PPIE contributed to the initial development of the research proposal, and the development of the research protocol, in the following ways: Initial consultation in planning the funding application. The research proposal was reviewed in July 2021, and the input and advice contributed to a successful funding application. The research team also responded to the formative comments when developing the protocol and continue to build on them. PPIE work influenced the types of data being included in the study. For example, an issue was raised whether some patients would admit to experiencing violence in ED, and therefore opportunities for the VPTs to improve ascertainment. Data on ethnicity from the 2011 and 2021 Census, which will also have some indication of those residing in refugee centres (and nursing homes, hostels, etc.), will be included in the study following recommendations that racial violence and the experiences of asylum seekers should be considered. It was further indicated that considering school attendance and exclusions data, free school meals and special educational needs would be valuable, in addition to educational attainment. The EDVIPE stakeholder reference group (SRG) now includes representation from primary care, secondary care and the third sector, after working with these sectors was suggested. Following feedback, EDVIPE now involves PPIE groups with lived experience of different types of violence. Following consultation with PPIE lived experience groups, development and refinement of the protocol and the addition of exploratory work was undertaken to: Better understand the pathways patients follow up to their attendance in ED, notably whether General Practitioner (GP) consultations were not acted upon. To consider patient ascertainment and therefore eligibility for the intervention in ED, as some victims may not realise that they are victims of violence. Whether one or two nurses are sufficient to manage an expected high caseload of patients attending ED due to assault, and whether the lack of 24-hour VPT provision would mean some patients are missed. Some patients might choose to avoid the police and therefore be less reluctant to receive support from the police, or police aligned services. We might explore this in any VPT referral data available. There were concerns that intervention-related activity might become known to the perpetrator, therefore elevating risk of subsequent harm. We can consider the immediacy of post-intervention re-attendance by patient group. That there are unique challenges for those who are disabled, both in terms of ability to engage and the nature of the support required. Some patients, notably those with children, may be less willing to engage as they would not want to risk losing their home or children. We can consider engagement by gender and presence of dependent children in the patient’s residence. The judicial system emphasises the right for both parents to be involved with their children, if any. Involvement of parents in the court system might be associated with a blunted intervention effect. That PPIE involvement in future diffusion and dissemination activity would lend credence to the project’s communication strategy. Methods and Analysis 8.1. Design and conceptual framework. A controlled longitudinal whole population (Wales, UK) natural experiment. The intervention in Cardiff began November 2019, and January 2022 in Swansea. Due to earlier changes in EDDS coding, these data are consistent and available from January 2012. Intervention data collection therefore begins in November 2019 and ends in August 2023, allowing a 12-month follow-up of patients until August 2024. 8.1.1. Objective 1 - Effectiveness We hypothesise that engagement with the VPT will help patients overcome modifiable risks and receive support for vulnerabilities, and that therefore the intervention will reduce the recurrence of unscheduled ED attendance. Other than those who are most seriously injured, patients will register at ED reception and be triaged, at which point the most appropriate pathway through ED will be determined. At reception patients will be asked about the reason for their attendance, including whether it was due to an assault, data that becomes a part of the Patient Management System (PMS) 42 . Patients may not disclose that the reason for their attendance was assault related. They might be reluctant, the perpetrator may have accompanied them, or they may wish to avoid scrutiny. One function of the VPTs is to work across clinical teams to improve ascertainment of ARAs. The result being that patients can be stratified according to the extent that they engage with the intervention. Patients identified in the ED PMS data as having attended due to an assault, but with no further contact with the VPT. Patients identified in the ED PMS or VPT data as having attended due to an assault, but did not further engage with the VPT. Patients identified in the ED PMS and VPT data and who engaged with the VPT. The primary analysis concerns group iii. The reasons for patients not engaging are potentially related to underlying characteristics and in secondary analyses we will explore this. However, it is reasonable to assume that the three groups represent varying levels of intervention dose, and therefore secondary analyses and therefore analyses using groups i to iii will be informative. 8.1.2. Objective 2 - Ascertainment Our second hypothesis is that intervention implementation improves ARA ascertainment. Patients attending ED can do so repeatedly within periods of time. This frequency is likely associated with the modifiable risks associated with ARA and is of interest here. For Objective 2 it is therefore appropriate to determine the proportion of attendances identified as violence-related. This generates time series data. The outcome of interest is therefore the count ARAs across all EDs. ARA attendance is defined as ARA in the ED PMS or, in the case of intervention sites, in the ED, PMS or VPT data. Codes indicating Provider Site in (the ED in its hospital) is available in the Emergency Department Data Set (EDDS), and this allows comparison between intervention EDs and control EDs. While intervention EDs have been in continual service since before the time series start dates, this is not so for all Type I EDs in Wales; there have been several changes with some EDs closing and others opening or being modified to receive additional patients. This, coupled with the intervention sites located at two of the largest hospitals in Wales, reduces the scope for selecting matching control sites 43 , leading to potentially important baseline differences in the interrupted time series data 44 . The counterfactual will therefore be ARAs across all control Type I EDs. 8.1.3. Objective 3 – Cost-Effectiveness We aim to determine whether the VPT represents value for money. The primary outcome will be quality adjusted life years, which will be estimated based on effectiveness estimates comparing ED attendance, reattendance, and any injuries received for those engaging in the VPT service relative to those who do not. Costs will be captured from an NHS perspective, reflecting the cost of the intervention but also other costs to the NHS due to referral. A secondary cost-effectiveness analysis will undertake a societal perspective, which will explore additional costs across social care, the police and the third sector (see Section 9). 8.1.4. Secondary Analyses We aim to co-produce study protocols with collaborators and PPIE groups, to provide opportunities for them to shape secondary and additional epidemiological analyses. This facilitates opportunities to realise and contribute to what is a rapidly changing policy area. One example is our inclusion of school attainment, exclusions, and attendance in analyses, which have been highlighted in these early discussions. VPUs and VRUs have made little headway working with the education system to challenge the causes of violence, and this has been identified as a priority 45 . Furthermore, additional exploration is planned to characterise those excluded from the intervention but who have available ALFs. 8.2. Population and Data This is a whole-population evaluation, including all residents of Wales, UK. Data is housed in the SAIL databank 39 . 8.2.1. Data 8.2.1.1. Violence Prevention Team Data (Cardiff and Swansea) Intervention sites record patient details (name, date of birth, gender), NHS number, extent of engagement with the VPT and whether any subsequent referral was made. 8.2.2. Administrative Data Several administrative datasets are available to characterise patients and summarised in Table 1 . Table 1 – EDVIPE administrate data sets. Dataset Name and Summary ADDE Annual District Death Extract provides the week of birth and date of death, used to describe left- and right-side censoring of study participants. OPRD Outpatient Referral Dataset OPRD includes data from outpatient referrals from primary care which will help in understanding the referral pathway to secondary care. This data includes all clinical referrals from General Practitioners, General & Community dental practitioners, A&E departments, walk-ins, consultant-to-consultant referrals. PEDW Patient Episode Data for Wales Dataset PEDW covers people domiciled in Wales and treated in Welsh and English NHS Trusts and English Trusts and includes data on inpatient and day case activities. This includes spells and episode data on hospital admissions, which can used to track the frequency of patients attending services for treatment arising from an ARA. WDSD Welsh Demographic Service Dataset WDSD includes all individuals registered with a Welsh General Practitioner and via anonymisation identifies household groups. WIMD Welsh Index of Multiple Deprivation WIMD is the Welsh Government’s official measure of relative deprivation for small areas in Wales, based eight domains including income, employment, health, and access to services. Typically grouped into fifths or “quintiles”, the WIMD is included in several NHS datasets. 2011, 2021 Census A census of the UK population is taken every ten years and includes questions relating to key demographics. Patient entry into and exit from EDVIPE can mean some will be missed in the 2011 Census (e.g., born after 2011), and some will be missed in the 2021 Census (e.g., died before 2021). Hence data from both censuses is required. Table 2 summarises key patient socioeconomic and demographic characteristics, required to enable secondary analyses in relevant sub-groups. Table 2 – EDVIPE patient characteristics Characteristic Source Age (from Week of Birth, WoB) WDS Sex WDS, 2011 and 2021 Census Ethnicity Asian (Bangladeshi, Chinese, Indian, Pakistani, Other Asian) Black (Caribbean, African, Other Black) Mixed (White and Asian, White and Black African, White and Black Caribbean, Other Mixed or Multiple ethnic groups) White (English, Welsh, Scottish, Northern Irish, British, Irish, Gypsy or Irish Traveller, Roma, Other White) Other (Arab, Any other ethnic group) 2011 and 2021 Census Quintile of Residential Deprivation WIMD Urban/rural residential classification 2011 and 2021 Census 8.2.3. Secondary Outcomes To characterise the WSMA involvement of patients involved in the intervention, broader pathways will be explored across several related data sets, summarised in Table 3 . Table 3 – EDVIPE secondary outcomes Dataset Name CAFCASS Children and Family Court Advisory and Support Service Wales Family Justice Data Set CAFCASS includes information for residents of England and Wales on cases of divorce, private law, family law act, public law, adoption, family law applications. It also includes information on marriage and divorce characteristics. The information on cases also includes information on number of children involved and types of hearing. MoJ Ministry of Justice: Data First Magistrates' court defendant data Crown Court defendant data Criminal courts and prisons data Prisoner custodial journey data Family Court data. NPD National Pupil Database This dataset has four broad categories of demographics, attainment, absence and exclusion, and children in need and looked after children. Police Data 1 Police Crime Dataset Pending applications for police crime data from all four Welsh forces are progressing and will be explored. SMSD Substance Misuse Dataset SMDS, also known as Welsh National Database for Substance Misuse (WNDSM) has data for people in Wales who present for substance misuse treatment. It includes details on assessments, referrals and treatment history. WLGPD Welsh Longitudinal General Practitioner Dataset WLGP contains clinical information from General Practice in Wales, including diagnoses and referrals into tertiary care. Note: 1 Data sharing agreements are currently being developed to bring all-Wales police crime data into SAIL. 8.3. Cohort Inclusion and Exclusion Criteria All residents of Wales who 11 years of age or older are eligible for inclusion. Residents of Wales will be defined through their identification in the WDSD. The NHS assigns each patient domiciled in the UK a unique number. This NHS number links across various NHS data systems. The encrypted and anonymised ALFs are derived from these NHS numbers. Therefore, patients attending ED whose identity cannot be connected to an NHS number (e.g. overseas visitors and tourists) will not have a corresponding ALF and will, by necessity, be excluded from analyses. 8.4. Allocation Intervention patients will be identified in the VPT data and will have attended intervention EDs (in Cardiff and Swansea), subject to the above inclusion criteria. Control patients will be identified in the EDDS, and where ED attendance was not in an intervention ED. 8.5. Progression criteria This is a definitive study. As the primary focus of the study uses routinely collected data, which is available for analysis subject to information governance permissions and extraction, progression criteria are not applicable. 8.6. Sampling From the intervention sites’ data (Section 8.2.1), between October 2019 and December 2022, the Cardiff VPT contacted 2,312 patients, of whom 77% accepted VPT support (Fig. 1 ). We conservatively estimated that there will be 2,500 patients that engaged with the Cardiff VPT across the four years (2019–2024) of VPT operation, and a further 900 from the two years (2021–2024) operation in Swansea. Across the entirety of Wales, there are approximately 1M ED attendances each year. -= Insert Fig. 1 About Here =- Figure 1 - Count of patients, by month, identified by the Cardiff VPT as attending due to a violence-related injury, with the number that subsequently engaged with the VPT. Initial estimates from Cardiff VPT suggest that 3% of those engaging with the VPT reattended ED at least once within one year, compared to 23% patients who did not engage with the VPT. Data from 2015 and 2016 suggest that the frequency of unscheduled attendances for patients with at least one ARA (mean = 2.35 attendances) is greater than patients making an unscheduled attendance without evidence of an assault (mean attendances = 1.73). For a simple Cox survival model, (α = 0.05, β = 0.90) and a hazard ratio of 0.8, a total N of 845 is required. To realise the recurrent nature of analyses, simulation 46 (1,000 estimates per point estimate) was used across varying follow-up periods, which suggests a 12 month follow-up period and total N of 300 is adequate to identify a significant effect. By increasing the number of controls, statistical power will be further enhanced 47 . 8.7. Analytic Strategy 8.7.1. Objective 1 - Effectiveness Our primary outcome is unplanned ED attendance. It represents the cost to the NHS of serious healthcare events and acts as a proxy for events eliciting acute healthcare needs. EDs provide acute care for patients without prior appointment, and the aftercare of patients who have received ED treatment but where there is no alternative provision (e.g., for out of area tourists). There can, therefore, be follow-up and planned appointments in ED. These appointments in ED will be made where, for example, there is an element of diagnostic uncertainty and a review is required in the ED context, or for patients where other follow-up arrangements are likely to fail (e.g., visitors to the area without access to primary care) 48 . Follow-up and planned appointments in ED are typically a continuation of the initial unplanned attendance or referral from another healthcare provider and are not valid outcomes for EDVIPE, as they are not elicited in response to acute healthcare need. Thus, for Objective 1 we will censor the timeline. Left-side censoring at birth or when someone takes up residence in Wales. Right-side censoring when someone dies or moves out of Wales. We further interval censor the timeline, to account for repeat ED attendances associated with a health event, such as referral from a local emergency hospital to a Major Trauma Centre. 8.7.1.1. Discontinuous Risk Interval Accounting for periods when individuals are not at risk is an essential consideration in repeated time-to-event models 49–52 . The clearest example in the current context is ensuring time at risk does not extend beyond date of death or originates before birth. Similarly, time at risk will also be left side censored, if patients move into Wales, and right side censored if they moved away from Wales. With no adjustment for these discontinuous risk intervals, the time at risk will be incorrect, increasing a greater likelihood of Type II errors. How patients are routed through emergency care pathways in Wales also influences time at risk. ED attendances are mainly determined by the acuity of the patient’s condition and the urgency with which they need to be seen. These decisions can be made by the Welsh Ambulance Service Trust (WAST), a Minor Injuries Unit (MIU) or in the local ED. It is feasible that a patient initially attends a local ED to be stabilised, is assessed, and requires referral to a Major Trauma Centre (MTC), or Trauma Unit (TU). Each MTC and TU are attached to an ED, and therefore in response to severe injury, patients are registered in more than one ED if they are referred from an ED without trauma facilities, to EDs that are attached to a TU or MTC. In Wales, emergency care in Wales is provided in MIUs, EDs (Local Emergency Hospitals, LEH, and Rural Trauma Facilities, RTFs), TUs, and MTCs. There is one MTC in Cardiff UHW, which services South and West Wales, and South Powys, and acts as a TU for the local population. Morriston Hospital in Swansea is a TU, but with additional specialist services (e.g., orthoplastics) meaning some patients otherwise destined for the MTC would be referred there instead. North Wales is serviced be the MTC in the Royal Stoke University Hospital in England. Patient admission is described using spells and super-spells. A spell represents patient care by speciality, and a super-spell is a collection of spells – two spells are included in the same super-spell if admission to the second speciality is within 48 hours of discharge from the first. ED attendances included within the same super-spell (e.g., transfer from an ED to an MTC) will therefore be attributable to the same initiating event, whether planned or unplanned. Therefore, only unscheduled attendances at the initiation of a super-spell will be included, and additional ED attendances within the same super-spell will be analytically censored. The duration of a super-spell constitutes a discontinuity in a patient’s exposure to risk, which is then accommodated in our analytic approach, this is described further in Fig. 2 . -= Insert Fig. 2 About Here =- Figure 2 - A simplified example of discontinuous risk intervals as they relate to unscheduled ED attendance and super-spells. In Fig. 2 a hypothetical analytic period runs from day 0 to day 20. Patient 1 lived in Wales for the duration of the study but had no ED attendances. Patient 2 lived in Wales at the start of the analytic period, but died on day 18, they had two ED attendances, with the first resulting in a two day stay in hospital. Their time at risk for the first ED attendance is three days, and for the second six days, with a total time at risk of 16 days. Patient 3 was born in Wales on day two, had three ED attendances and was alive and living in Wales at the end of the analytic period. However, their second ED attendance (e.g., transfer to a MTC) occurred in a super-spell associated with the first ED attendance, and this is dropped from consideration. They therefore had two ED attendances, with the time at risk for the first as four days, and the time at risk for the second two days. Their total time at risk is 10 days. We assume that unscheduled ED attendances are independent and unordered, and therefore assume a common baseline hazard for all events. Thus, the hazard function, \(\:{\lambda\:}_{ik}\left(t\right)\) , for the k th ED event for each subject ( i ) is: \(\:{\lambda\:}_{ik}\left(t\right)={\lambda\:}_{0}\left(t\right){e}^{{x}_{ik}\beta\:}\) 1 However, it is feasible that operational differences across Local Health Boards (LHBs) might influence the likelihood that ED attendance varies across groups of patients. For example, initiatives that influence the likelihood patients are conveyed into emergency care, or alternative provision in the community for lower acuity patients. We introduce a random effect, \(\:{\alpha\:}_{i}\) , to describe variation in risk, or frailty, can be introduced 53 : \(\:{\lambda\:}_{ik}\left(t\right)={\lambda\:}_{0}\left(t\right){\alpha\:}_{i}{e}^{{x}_{ik}\beta\:}\) 2 8.7.1.2. Dates and Times Records on date and time of events in routine data requires consideration. With most events recorded by day, we are obliged to use day as the primary measure of time, and therefore collisions will occur. In EDDS the date and time of attendance and discharge are recorded, and several events can occur on the same day. For example, it is feasible that a patient attends ED on multiple occasions and on the same day. ED attendances on the same day are collapsed onto one another under the super-spell assumption. In the WDSD, while date of birth is approximated to the Monday of the week in which they were born (their Week of Birth, WoB), date of death is accurate to that degree of granularity. For example, a patient might attend ED at 9am and die at 2pm on the same day. To avoid inconsistencies such as this patient entering the study cohort after death, we will adopt the usual approach and add a small value (epsilon, ε) to dates to preserve the chronological order of events 54 . 8.7.1.3. Matching It is not possible to manipulate allocation to treatment, and we therefore rely on quasi-experimental methods to infer treatment effects. The purpose of matching control and intervention patients is to allow derivation of the average treatment effect under the assumption that allocation is not conditional on observed confounders. Matching on values of covariates between the treated and untreated groups avoids the bias introduced by covariates that influence the outcome variable. The goal being to find a subset of data that is closest to an exact match on observed covariates 55 . The primary analysis will be conducted using Coarsened Exact Matching, with a minimum 1:1 ratio, with secondary analyses undertaken using Propensity Score Matching. For categorial covariates, exact matching is a sensible alternative. Here a match is considered acceptable only if the values of all covariates are equal between treated and untreated potential matches. Continuous covariates can, for the purpose of using exact matching, be turned into categorial covariates by defining category intervals; the resulting method is called coarsened exact matching. For matching, covariate categories can be grouped together to form larger (and fewer) categories, improving efficiency 55–57 . This approximates to a fully blocked experiment and requires temporarily coarsening variables. Propensity score matching reduces the multi-dimensional space of covariates to a single summary covariate, the propensity score, typically derived using a logistic equation and performing regression on the treatment group for the covariates chosen to match with. The proximity of a potential match to the given subject is estimated in terms of the closeness of their propensity scores. This method simplifies subsequent analyses as matching refers only to one variate, the propensity score. However, propensity score matching can increase, rather than decrease, the imbalance of the covariates in the samples 55 . A challenge is optimising the trade-off between the quality of matches and the sample size. A larger sample reduces sampling error and can increase the study power, but including matches of lower quality may lead to greater residual imbalance 58 . The R package Matching Frontier attempts to address these issues by calculating the entire balance-sample size frontier, from which the user can easily choose one, several, or all subsamples to use for their final analysis, given their own choice of imbalance metric and quantity of interest 59 60 . 8.7.1.4. Missing Data Missing data may arise specific to the use of census data, where it is feasible that records are missing entirely, or patients did not disclose personal characteristics. However, give that these characteristics are recorded across multiple data sets the expectation is that missingness will minimally impact on the analyses. 8.7.2. Objective 2 - Ascertainment The hypothesis is that a violence-attendance is more likely to be ascertained as such with the additional VPT resource present in ED. All unplanned attendances can be coded as ARA, or not (1, 0). The time series begins in January 2012, and the two sites (Cardiff and Swansea) can be individually compared with all other Type I EDs in Wales. This indicates that a difference-in-difference model 61 is appropriate. We will analyse Cardiff first, using the implementation in Swansea as a future replication of the intervention to facilitate a more robust causal interpretation of any effect. Additional descriptive analysis can explore patient groups (age, gender, time of attendance) more likely to be ascertained by the VPTs. Unlike Objective 1, we are interested in all ED attendances as the opportunity to provision safeguarding is applicable across all ED attendances, irrespective of the frequency of attendance. 8.7.3. Objective 3 - Cost-effectiveness Analysis The aim of the economic evaluation is to understand the costs and consequences of providing a VPT versus no VPT within EDs. For this study, we will focus on the objectives of understanding whether the VPT represents value for money from an NHS perspective. 8.7.3.1. Methods A within study analysis will determine the short-term cost-effectiveness of offering VPT relative to not offering VPT within EDs. Estimates derived from this study will be used to inform a long-term decision analytic model examining the cost-effectiveness of providing, versus not providing, VPT within EDs over a longer time horizon. Alternative perspectives for the analysis will be considered, with the NHS perspective being the base case. Cost-effectiveness will be presented using standard statistics. 8.7.3.2. Exposure variable The exposure variable consists of patients engaging with the VPT. 8.7.3.3. Outcome variables Health-related outcome variables include improved mental health outcomes, measured as a reduction in primary and secondary care visits for a mental health reason, reduced substance use, improved physical health, measured as a reduction in assault-related attendances in ED, and improved health-related quality of life (HRQoL), measured using injury-related estimates from the literature. Data on all outcomes will be sourced from the SAIL databank 39 . Estimates will be measured on a quarterly basis for the follow-up period (12–48 months). HRQoL will be estimated using the literature, in line with other studies on hospital-based violence prevention programs. 8.7.3.4. Within study statistical analysis. Differences in overall mean outcomes will be analysed using a fixed effects panel regression model, where we will control for individual and time varying effects 62 . 8.7.3.5. Covariates of interest. The following covariates will be included within the fixed effects panel regression model: patient age, gender, severity of injury, number of years VPT has been operating, year, and quarter. Referral uptake will be treated as a mediating variable and not adjusted for in the models. 8.7.3.6. Long term cost-effectiveness analysis (CEA) A Markov decision model with quarterly cycles will be used to estimate the cost-effectiveness of offering VPTs in EDs. To determine cost effectiveness using a health and social care perspective (primary analysis), the Markov model will be health-state based, with key health states likely to be healthy, injured, or dead. To support a multi-sector perspective (secondary analysis), the Markov model will be event-based. 8.7.3.7. Resource use Healthcare resource use will include number of and type of visits with primary care (e.g., GP), secondary care (e.g., inpatient stays, outpatient clinic appointments), and other services (e.g., other healthcare professional visits, physiotherapy, occupational therapy, mental health therapy, drug and alcohol treatment). 8.7.3.8. Cost data Cost data will be estimated for health service use based on resource use from the SAIL databank with appropriate unit costs applied. Third sector resource use will be valued using cost data provided in the resource use questionnaires. Other Unit Cost estimates (e.g., NHS reference costs, Unit Costs of Health and Social Care) will be used to supplement where needed. Differences in overall mean total health care costs (including primary and secondary care) between groups will be analysed using GLMM. Cost data are known to be left-skewed with a substantial number of zeroes and there is no single dominant method for analysis 63 . We will test several potential models from the GLMM family and identify the correct link function and distribution to use. Potential effect modifiers include age, gender, and deprivation. Mediating variables include intervention engagement. Potential confounding variables include age, gender, deprivation, number of years VPT has been operating. The regression model will also include time variables recording the year and quarter. Costs and benefits will be discounted at 3.5% per annum as recommended by the HM Treasury 64 . 8.7.3.9. Time horizon. The time horizon for the decision analytic model will extend beyond the time frame of this study, to encompass a period of time where, if possible, all relevant costs and outcomes for this appraisal have been accrued 65 . 8.7.3.10. Cost-effectiveness Analysis Outcomes. For the primary analysis, we will report the incremental cost effectiveness ratio (ICER) as the cost per quality-adjusted life-year (QALY), the net monetary benefit (NMB), and the net health benefit (NHB), for those being offered versus not offered VPT assessment during their A&E visit. NMB and NHB will be assessed using a range of values (£15,000-£30,000 per quality-adjusted life year, or QALY) representing a health decision makers’ willingness-to-pay (WTP) to obtain the distribution of net benefits at different levels of WTP. For the secondary (multi-sector) analysis, we report the cost per unit of effectiveness in natural units (e.g., cost per reconviction avoided). We will also present the results using the extended impact inventory framework and consider alternative methods of aggregation 66 . 8.7.3.11. Sensitivity Analysis Parameter uncertainty will be assessed using a probabilistic sensitivity analysis (PSA), varying key parameters over a range of expected values, and running 1,000 Monte Carlo simulations. 8.7.3.12. Social Costs Additional exploratory analyses will estimate the broader social costs associated with the VPTs. Social costs will focus specifically on organisation costs of supporting patients who have been referred to them by the VPTs. Cost data of organisations will be assessed using top-down gross costing. Questionnaires or interviews, depending in interviewee preference, will be conducted with members of organisations to assess (i) the total annual expenditure cost for third sector and statutory organisations in the given year 67 . This cost is the total yearly cost of running the service including costs of supporting patients, staff cost, and consumables. (ii) The total number of users supported each year by the third sector and statutory organisations. (iii) The percentage of those supported that were referred by the VPTs each year. The given year will be 2022 to 2023; this is due to the Swansea team being implemented in January 2022. Top-down gross costing allows data to be aggregated and will allow us to estimate the mean cost of a ‘typical patient’ to a ‘typical organisation’. We will include the distribution for each organisation but will not attribute a specific cost to a specific organisation. Furthermore, we will ask organisations for standard cost data (e.g. cost per counselling session); this will be used as a comparison of our mean average. We will ask about waiting lists for support which will be used to inform our knowledge. Strengths and Limitations Linkage requires knowledge of the patients’ identity, anonymised and encoded as ALFs, to be included in the data return. There are instances where individuals may prefer to remain anonymous. Analyses are therefore limited to those who can be identified in routine data and linked to the WDS and therefore EDDS and related datasets. It is feasible that some may attempt to conceal their identity, and this could correspond to greater vulnerability. Furthermore, our reliance on administrative data means there is no information on the context in which their exposure to violence arose. There maybe marked difference in the modifiable risks and vulnerabilities for patients sustaining injury in and around premises licensed for the sale and onsite consumption of alcohol, and those experienced criminal or sexual exploitation. Conversely, our primary hypothesis is that the additionality of the VPT intervention teams means ED is better able to work with patients and therefore address the variation of patients that attend. Dissemination, Outputs and Anticipated Impact A diffusion and dissemination plan will be co-produced with stakeholders and PPIE groups. In so doing, we will define audiences (policy, practitioner, academic, lay, etc.) that might find the project outcomes of interest. In so doing, we will determine appropriate modalities to communicate with each, including presentations, briefing documents, and media events, with content informed by audience need. 10.1. PPI, Policy and Practitioner Focused An ongoing evaluation of VRUs and VPUs is to recommend that greater attention is paid to evidence-based interventions and that activities should consider the possible involvement of schools and therefore the Department for Education. Education data has been included and we can therefore contribute to the evidence based linking school activity to violence. In addition, our PPIE engagement highlighted the need to consider ARA predictors, and engagement with the Home Office will further identify intermediate outcomes. We therefore aim to undertake interim analyses that will inform the final analysis and respond to significant emerging policy questions as it relates to evidence-based violence prevention and reduction initiatives. While these will likely translate to academic papers, we also seek to produce more accessible outputs and exploit existing networks in that respect. Abbreviations ADDE Annual District Death Extract - All deaths registered in Wales. Including Welsh residents who died outside of Wales. ALF Anonymised Linkage Field – All individual records are assigned an ALF, allowing researchers to link different data sets together for a single person based on the ALF. ARA Assault-related Attendance – An attendance to an emergency department relating to the attendee being involved in an assault. BAWSO Black Association of Women Step Out – An All-Wales charity supporting Black and Minoritised individuals and communities. CAFCASS Children and Family Court Advisory and Support Service – A service that looks after the interests of CYP in family court proceedings. It is independent from Social Services and the courts. CEA Cost-effectiveness Analysis – Economic analysis using costs and outcomes of interventions. CG Control Group – A group that is not receiving the intervention that is being researched. The control group may receive the standard intervention or no intervention. CYP Children and Young People – Representing those from birth, usually up until the age of 18. DASH Domestic Abuse, Stalking and Honor Based Violence – A checklist used by practitioners for high-risk cases of DASH. DBS Disclosure and Barring Service – Part of the Home Office. It allows organisations to check the safety of those being recruited for work. Including criminal record checks and checks involving safety around CYP. DHCW Digital Health and Care Wales - An organisation that designs all-Wales digital health and care services. It provides digital services with the aim to improve people's health. ED Emergency Department – Area of the hospital for immediate and urgent care. EDDS Emergency Department Data Set – National data set from emergency departments. It includes why the patient attended the ED and what treatment they received. EDVIPE Emergency Department Violence Intervention Programme Effectiveness and Cost-effectiveness Evaluation – The acronym of the current study. FoI Freedom of Information – Freedom to share or consume information from government funded public agencies. This information can be requested by members of the public through a FoI request. GLMM Generalised Linear Mixed Models – An extension to a generalised linear model that includes fixed and random effects. GP General Practitioner – GPs locally treat common medical conditions. They can refer to hospitals. GRIPP Guidance for Reporting Involvement of Patients and the Public – A checklist for the inclusion of person and patient involvement in research to improve transparency and quality. HRQoL Health-Related Quality of Life – A concept that examines health's impact on quality of life. HVIP Hospital-based Violence Intervention Programme – A trauma-informed programme that ensures support for those who attend hospital with a violence-related injury. ICD10 International Statistical Classification of Diseases - 10 th revision. ICER Incremental Cost Effectiveness Ratio – Summary of the economic value of an intervention. IDVA Independent Domestic Violence Advocate - IDVAs work with those who have been affected by domestic violence through activities such as criminal justice support and representing the victim in legal proceedings. IGRP Independent Information Governance Review Panel – For the current study the IGRP was from SAIL. The IGRP includes representatives from various organisations and sectors to determine our access to SAIL databank. ISTV Information Sharing to Tackle Violence - An anonymised dataset that is collected by the NHS ED departments. ISVA Independent Sexual Violence Advisor – ISVA works with those affected by sexual violence. ITSA Interrupted Time Series Analysis – Used when looking at outcomes. It involves looking at data before and after an interruption, i.e. an intervention. LEH Local Emergency Hospitals – A hospital for immediate and urgent care. LEHs do not have a major trauma centre but are equipped to aid and transfer patients to major trauma centres. LSOA Lower Layer Service Output Areas – Geographical areas in England and Wales where data is collected, for example, population count and crime rate. MARAC Multi-Agency Risk Assessment Conference – A meeting to discuss high risk domestic abuse cases. The meeting includes professionals such as, the police, health, IDVA’s, children's services and third sector agencies. MARF Multi-Agency Referral Form – A referral form used to report a concern about a child at risk. MIU Minor Injury Unit – A walk in unit in a hospital for non-emergency care. They provide care and treat injuries such as rashes, cuts, sprains and burns. MoJ Ministry of Justice – A government department that is responsible for the criminal justice system, probation, courts, safeguarding and to examine and adapt the legal service. MoPI Management of Police Information – Management of police records and data to collect necessary and proportional data. MTC Majot Trauma Centre – Unit in a hospital that provides support to patients with major trauma. Major trauma can be defined as an illness or incident that can cause permanent disability or death. NHB Net Health Benefit – The benefit given to an intervention, that is worked out using the total expected costs of the intervention divided by the maximum cost effectiveness ratio. If the net health benefit of an intervention is positive the health of the population would be increased. NHS National Health Service – Health care that is funded by the UK Government. NMB Net Monetary Benefit – Representing the value of an intervention in monetary terms, that is worked out by looking at the difference of monetary value of QALYs and the total estimated costs of the intervention. NPD National Pupil Database – Data set of all pupils in public schools in England including absence, exclusion and demographic information. NRAC National Retention Assessment Criteria – Scheduled reviews of information retention which poses several questions that assesses risk of harm posed by nominals. If the answer is ‘yes’ to any of the criteria, the records must be retained and reviewed at a scheduled later date. OPCC Office for the Police and Crime Commissioner - Police and crime commissioners are elected police officials that are responsible for their force areas, including the police budget and how the area is policed. OPRD Outpatient Referral Dataset - Data collected from the NHS including information on where the patient was referred to, the reason for referral and the service type requested. PEDW Patient Episode Database for Wales – This dataset included all inpatient and day case activity occurring in NHS Wales institutions. It also includes data on Welsh residents that were treated in England. PMS Patient Management System - A digital computer system for health professionals to access patient records. PNC Police National Computer – A system used by the UK police and law enforcement to access information in real-time. PND Police National Database – A national system to store an accumulation of policing information from all forces. PPI Public and Patient Involvement - The inclusion of the public in a research project. PSA Probabilistic Sensitivity Analysis – a method used in economic evaluations that allows the researcher to assess the level of uncertainty or confidence in a decision. QALY Quality-Adjusted Life Year – Years that are lived in perfect health. QALYs are used to assess the value of interventions. RALF Residential Anonymous Linkage Field – An ALF that includes residential information such as location, number of residents and the change in occupants. RALFs can be used the assess the interaction between a person's residential setting and their health. (see ALF). RTF Rural Trauma Facilities – Hospitals in rural areas that are equipped to aid major trauma patients. SAIL Secure Anonymised Information Linkage – SAIL is a data bank that allows researchers to access, link and assess patient and public health information. SMDS Substance Misuse Data Set – Data set on patients referred for a substance misuse problem. Also known as The Welsh National Database for Substance Misuse (see WNDSM). SPIRIT Standard Protocol Items: Recommendation for Intervention Trial – Recommendations for content for a clinical protocol. SRG Stakeholder Reference Group – A group made up of stakeholders that provide feedback and direction on the research project. SRGs are made up of stakeholders from different sectors. SSC Study Steering Committee – A group that provides oversight of the research activities. The members usually have lived or relevant experience of what is being researched. SUPER Service Users for Primary and Emergency care Research – A group that includes Welsh residents that represent diverse backgrounds and experiences. They provide perspectives on research topics and give suggestions on how to conduct PPI. TIDier Template for Intervention Description and Replication – A checklist and guide for describing interventions that can be used for replication. TU Trauma Unit – A unit in a hospital that aids those with major trauma. Trauma units are used when the patient is not stable enough to be moved to a major trauma centre. UHW University Hospital Wales – Hospital situated in Cardiff, Wales. UK United Kingdom. VPT Violence Prevention Team – The VPTs are a type of hospital-based violence intervention programme. They are currently situated in two South Wales emergency departments. VPU Violence Prevention Unit (Wales) – A unit funded by the Home Office. The VPU consists of a multi-disciplinary team that looks at evidence and research to reduce violence in Wales. VRU Violence Reduction Unit (England and Scotland) - A unit funded by the Home Office. The VRUs consist of multi-disciplinary teams that look at evidence and research to reduce violence in England and Scotland. WAST Welsh Ambulance Service Trust – An NHS trust providing an ambulance service for Wales. WDSD Welsh Demographic Service Database - WDSD includes all individuals registered with a Welsh General Practitioner and via anonymisation it identifies household groups. WIMD Welsh Index of Multiple Deprivation - The Welsh Government’s official measure of relative deprivation for small areas in Wales. WLGP Welsh Longitudinal General Practitioner Dataset – A data set that contains clinical information from General Practitioners in Wales, including diagnoses and referrals to tertiary care. WNDSM The Welsh National Database for Substance Misuse - Data set on patients referred for a substance misuse problem. Also known as Substance Misuse dataset (See SMDS). WSMA Whole System Multi-Agency – An approach that includes people from multiple different agencies, for example, health care, police and the government. WTP Willingness To Pay – The maximum price that someone is willing to pay for a product. In health economics it can be used as how much people are willing to pay to improve health and reduce risk. YEF Youth Endowment Fund – The YEF fund research happening across England and Wales that is focused on preventing CYP becoming involved in violence. Declarations Data Availability The routine data used in these studies are housed in the SAIL databank and are available through application. Code written to manage and analyse data is available on request. Funding Funded by the National Institute for Health Research, Public Health Research Board (NIHR134055). Role of the Funder The funder had no role in the design of the study, and will not have any role during its execution, analyses, interpretation of the data, or decision to submit results. Governance A Study Steering Committee (SSC) was convened to oversee and advise on progress. The SSC included the Data Monitoring and Ethics functions and oversees any protocol amendments. Acknowledgements The authors would like to acknowledge Nurses Vicky Lee and Sarah Wilcox, leads for the intervention sites, for their support of this evaluation. Furthermore, we would like to thank the Public, Patient Involvement participants, many of whom had not prior experience working with researchers. Their enthusiasm and insights greatly impacted on this evaluation and the broader research team. Author Contributions SC Moore: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, guarantor, implemented the evaluation, analysed the data. S Brophy: drafted and revised the paper, implemented the evaluation. A Bandyopadhyay: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, analysed the data. A Newbury: designed data collection tools, drafted and revised the paper. T Lowe: drafted and revised the paper, implemented the evaluation, designed PPIE engagement, conducted and wrote-up PPIE engagement. D O'Reilly: drafted and revised the paper, implemented the evaluation. D Rawlinson: drafted and revised the paper, implemented the evaluation. L Snowdon: drafted and revised the paper, implemented the evaluation. J Shepherd: drafted and revised the paper, implemented the evaluation. V Sivarajasingam: drafted and revised the paper, implemented the evaluation. A Watkins: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, implemented the evaluation. S Walker: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, implemented the evaluation. S Borgia: drafted and revised the paper, designed PPIE engagement. A Battaglia: designed PPIE engagement, conducted PPIE engagement. H Yeomans: designed and implemented PPIE engagement, conducted PPIE engagement. S Premji: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper. R Aslam: designed PPIE engagement. M Hamilton: designed data collection tools, drafted and revised the paper. Conflicts of Interest None Author Contributions The original manuscript was submitted for publication to BMJ Open on 8 March 2024. On 13 November 2024 the manuscript had the status “awaiting reviewer assignment.” The decision was therefore made to withdraw from BMJ Open and submit to BMC Public Health. Checklists Template for Intervention Description and Replication (TIDieR) is available in Appendix 20.1. Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist is available in Appendix 20.2. References Purtle J, Dicker R, Cooper C, et al. Hospital-based violence intervention programs save lives and money. Journal of Trauma and Acute Care Surgery 2013;75(2):331-33. Cunningham R, Knox L, Fein J, et al. Before and after the trauma bay: the prevention of violent injury among youth. Annals of Emergency Medicine 2009;53(4):490-500. Kaufman E, Rising K, Wiebe DJ, et al. Recurrent violent injury: magnitude, risk factors, and opportunities for intervention from a statewide analysis. The American Journal of Emergency Medicine 2016;34(9):1823-30. Kao AM, Schlosser KA, Arnold MR, et al. Trauma recidivism and mortality following violent injuries in young adults. Journal of Surgical Research 2019;237:140-47. Newton AS, Hartling L, Soleimani A, et al. A systematic review of management strategies for children’s mental health care in the emergency department. Emergency Medicine Journal 2017;34(6):376-84. Hinsliff‐Smith K, McGarry J. Understanding management and support for domestic violence and abuse within emergency departments: A systematic literature review from 2000–2015. Journal of clinical nursing 2017;26(23-24):4013-27. Ansari S, Boyle A. Emergency department-based interventions for women suffering domestic abuse: a critical literature review. European Journal of Emergency Medicine 2017;24(1):13-18. Brice JM, Boyle AA. Are ED-based violence intervention programmes effective in reducing revictimisation and perpetration in victims of violence? A systematic review. Emergency medicine journal 2020;37(8):489-95. Koenig KL, Benjamin SB, Beÿ CK, et al. Emergency Department Management of the Sexual Assault Victim in the COVID Era: A Model SAFET-I Guideline From San Diego County. Journal of Emergency Medicine 2020 Goodall C, Jameson J, Lowe DJ. Navigator: A Tale of Two Cities. Glasgow: Violence Reduction Unit 2017. Sivarajasingam V, Guan B, Page N, et al. Violence in England and Wales in 2019. Cardiff, UK: Cardiff University 2020. Sivarajasingam V, Guan B, Page N, et al. Violence in England and Wales in 2020. Cardiff, UK: Cardiff University 2021. Introducing Public Health Measures (HO0345). London: The Home Office, 2019. Shepherd JP, Ali M, Hughes A, et al. Trends in urban violence:. Journal of the Royal Society of Medicine 1993;86(2):87. Sutherland I, Sivarajasingam V, Shepherd JP. Recording of community violence by medical and police services. Injury Prevention 2002;8(3):246-47. Gray BJ, Barton ER, Davies AR, et al. A shared data approach more accurately represents the rates and patterns of violence with injury assaults. J Epidemiol Community Health 2017;71(12):1218-24. Bath R. A whole-system multi-agency approach to serious violence prevention: A resource for local system leaders in England. London: Public Health England 2019. Serious Violence Strategy. London: HM Government 2018. Violence Reduction Unit Interim Guidance. London: Home Office 2020. The NHS Long Term Plan 2019 [Available from: www.longtermplan.nhs.uk accessed November 2020. Better Care for People With Co-occurring Mental Health and Alcohol/Drug Use Conditions. London: Public Health England 2017. Duke AA, Smith KM, Oberleitner L, et al. Alcohol, drugs, and violence: A meta-meta-analysis. Psychology of violence 2018;8(2):238. Babor TF, Berglas S, Mendelson JH, et al. Alcohol, affect, and the disinhibition of verbal behavior. Psychopharmacology 1983;80(1):53-60. Malik NS, Munoz B, de Courcey C, et al. Violence-related knife injuries in a UK city; epidemiology and impact on secondary care resources. EClinicalMedicine 2020:100296. Pallett J, Sutherland E, Glucksman E, et al. A cross-sectional study of knife injuries at a London major trauma centre. The Annals of The Royal College of Surgeons of England 2014;96(1):23-26. Boyle A, Frith C, Edgcumbe D, et al. What factors are associated with repeated domestic assault in patients attending an emergency department? A cohort study. Emergency medicine journal 2010;27(3):203-06. DeWall CN, Anderson CA, Bushman BJ. The general aggression model. Psychology of Violence 2011;1(3):245. Allen JJ, Anderson CA. General aggression model. IEME 2017:1-15. Farrington DP. Early predictors of adolescent aggression and adult violence. Violence and victims 1989;4(2):79-100. Hawkins JD, Herrenkohl T, Farrington DP, et al. A review of predictors of youth violence. 1998 Piquero AR, Jennings WG, Diamond B, et al. A systematic review of age, sex, ethnicity, and race as predictors of violent recidivism. International journal of offender therapy and comparative criminology 2015;59(1):5-26. World Report on Violence and Health. Geneva: World Health Organisation 2002. Long SJ, Fone D, Gartner A, et al. Demographic and socioeconomic inequalities in the risk of emergency hospital admission for violence. BMJ Open 2016;6(8) Ostrowsky MK. The social psychology of alcohol use and violent behavior among sports spectators. Aggression and violent behavior 2014;19(4):303-10. Fazel S, Gulati G, Linsell L, et al. Schizophrenia and violence: systematic review and meta-analysis. PLoS Med 2009;6(8):e1000120. Fu F, Nowak MA, Christakis NA, et al. The evolution of homophily. Scientific reports 2012;2(1):1-6. Van Godwin J, Moore G, Hamilton M, et al. Implementation and Process Evaluation of South Wales Hospital Based Violence Intervention Programmes. London: Youth Endowment Fund Forthcoming. Florence C, Shepherd J, Brennan I, et al. Effectiveness of anonymised information sharing and use in health service, police, and local government partnership for preventing violence related injury: experimental study and time series analysis. BMJ 2011;342:d3313. doi: 10.1136/bmj.d3313 Jones KH, Ford DV, Thompson S, et al. A profile of the Sail Databank on the UK secure research platform. International journal of population data science 2019;4(2) Irving A, Buykx P, Amos Y, et al. The acceptability of alcohol intoxication management services to users: a mixed methods study. Drug and Alcohol Review 2020;39(1):36-43. Staniszewska S, Brett J, Simera I, et al. GRIPP2 reporting checklists: tools to improve reporting of patient and public involvement in research. bmj 2017;358 Florence C, Shepherd J, Brennan I, et al. Effectiveness of anonymised information sharing and use in health service, police, and local government partnership for preventing violence related injury: experimental study and time series analysis. Bmj 2011;342 Linden A. A matching framework to improve causal inference in interrupted time‐series analysis. Journal of Evaluation in Clinical Practice 2018;24(2):408-15. Linden A, Yarnold PR. Using machine learning to evaluate treatment effects in multiple‐group interrupted time series analysis. Journal of Evaluation in Clinical Practice 2018;24(4):740-44. Evaluation of Violence Reduction Units 2020/21. London: Home Office, forthcoming. Feiveson AH. Power by simulation. The Stata Journal 2002;2(2):107-24. Hennessy S, Bilker WB, Berlin JA, et al. Factors influencing the optimal control-to-case ratio in matched case-control studies. American Journal of Epidemiology 1999;149(2):195-97. Emergency Department Follow-up Clinics. London, 2015. Law CG, Brookmeyer R. Effects of mid‐point imputation on the analysis of doubly censored data. Statistics in medicine 1992;11(12):1569-78. Rücker G, Messerer D. Remission duration: an example of interval‐censored observations. Statistics in Medicine 1988;7(11):1139-45. Turnbull BW. The empirical distribution function with arbitrarily grouped, censored and truncated data. Journal of the Royal Statistical Society: Series B (Methodological) 1976;38(3):290-95. Guo Z, Gill TM, Allore HG. Modeling repeated time-to-event health conditions with discontinuous risk intervals. Methods of information in medicine 2008;47(02):107-16. Balen TA, Putter H. A tutorial on frailty models. Statistical Methods in Medical Research 2020;29(11):3424-54. Gould W. Why can’t a subject enter and die at the same time in the Cox model? Texas, U.S.: StataCorp; 2023 [Available from: www.stata.com/support/faqs/statistics/time-and-cox-model/ accessed June 2023. King G, Nielsen R. Why Propensity Scores Should Not Be Used for Matching. Political Analysis 2019;27(4):435-54. doi: 10.1017/pan.2019.11 King G, Nielsen R, Coberley C, et al. Comparative effectiveness of matching methods for causal inference. Unpublished manuscript, Institute for Quantitative Social Science, Harvard University, Cambridge, MA 2011 Stuart EA. Matching methods for causal inference: A review and a look forward. Stat Sci 2010;25(1):1-21. doi: 10.1214/09-STS313 [published Online First: 2010/09/28] Imbens GW. Matching methods in practice: Three examples. Journal of Human Resources 2015;50(2):373-419. King G, Lucas C, Nielsen RA. Matching Frontier 2017 [Available from: https://projects.iq.harvard.edu/frontier. King G, Lucas C, Nielsen RA. The Balance-Sample Size Frontier in Matching Methods for Causal Inference. Am J Polit Sci 2017;61(2):473-89. doi: 10.1111/ajps.12272 Wing C, Simon K, Bello-Gomez RA. Designing difference in difference studies: best practices for public health policy research. Annual review of public health 2018;39(1):453-69. Jones AM. Panel data methods and applications to health economics. TC Mills and K Petterson, Palgrave Handbook of Econometrics 2007;2 Jones AM, Lomas J, Rice N. Healthcare cost regressions: going beyond the mean to estimate the full distribution. Health economics 2015;24(9):1192-212. Treasury HMs. The green book: Central government guidance on appraisal and evaluation. London: HM Treasury 2018 Drummond MF, Sculpher MJ, Claxton K, et al. Methods for the economic evaluation of health care programmes: Oxford university press 2015. Walker S, Griffin S, Asaria M, et al. Striving for a societal perspective: a framework for economic evaluations when costs and effects fall on multiple sectors and decision makers. Applied health economics and health policy 2019;17:577-90. Špacírová Z, Epstein D, García-Mochón L, et al. A general framework for classifying costing methods for economic evaluation of health care. The European Journal of Health Economics 2020;21(4):529-42. Additional Declarations No competing interests reported. Supplementary Files 1.Appendices.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5452363","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Study protocol","associatedPublications":[],"authors":[{"id":379523562,"identity":"0098b1cb-f7ae-403d-87cc-9848f19432bb","order_by":0,"name":"Simon C. Moore","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4klEQVRIie3RMQrCMBSA4VcCdXmSNS7eQHgQ6OhdpNCpg+CuQqFO4mrBQwi9QKXQLjlAR12cOhRcHDLY2FGsHR3yD8mSj+QRAJvtj5tysyJ1O0D2m8jJ9k26fRBZnM0hhAFktouKptFzJy3ze1MvScAovzJU34mnCj9JYp95KpDJiVqCATGsekgVSjbeZq5XgWRIeg0QAsPmB9E6Q3kcPVrS3sLrAQTcTJBA2RFhbul7WDuLs499EipcOWYWV9zpcuobv4xyeOr55rArU6g1Cc7927UuvpPPXBj0kTabzWbr6wW/hEtg49OeWwAAAABJRU5ErkJggg==","orcid":"","institution":"Cardiff University","correspondingAuthor":true,"prefix":"","firstName":"Simon","middleName":"C.","lastName":"Moore","suffix":""},{"id":379523563,"identity":"fe8e7ba8-b30e-4f09-9096-2b3d7d41af83","order_by":1,"name":"Sinead Brophy","email":"","orcid":"","institution":"Swansea University","correspondingAuthor":false,"prefix":"","firstName":"Sinead","middleName":"","lastName":"Brophy","suffix":""},{"id":379523565,"identity":"267ead16-8ba4-41e1-9fcb-deb14646b12f","order_by":2,"name":"Amrita Bandyopadhyay","email":"","orcid":"","institution":"Swansea University","correspondingAuthor":false,"prefix":"","firstName":"Amrita","middleName":"","lastName":"Bandyopadhyay","suffix":""},{"id":379523567,"identity":"9e647ef9-dd78-4f42-b25f-673df473688b","order_by":3,"name":"Annemarie Newbury","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Annemarie","middleName":"","lastName":"Newbury","suffix":""},{"id":379523568,"identity":"c0886074-8270-454a-be08-dcbdcc42effa","order_by":4,"name":"Megan Hamilton","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Megan","middleName":"","lastName":"Hamilton","suffix":""},{"id":379523569,"identity":"ad588f27-d5e4-4379-bbbe-1ac03f06fe95","order_by":5,"name":"Adele Battaglia","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Adele","middleName":"","lastName":"Battaglia","suffix":""},{"id":379523572,"identity":"176a273f-2cd1-4183-b583-a26946792ed7","order_by":6,"name":"Trudy Lowe","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Trudy","middleName":"","lastName":"Lowe","suffix":""},{"id":379523573,"identity":"78e36fde-e477-4401-a571-6e410422afb8","order_by":7,"name":"David O'Reilly","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"O'Reilly","suffix":""},{"id":379523574,"identity":"139e4d3e-dba0-415e-a2b7-63189888eb28","order_by":8,"name":"David Rawlinson","email":"","orcid":"","institution":"Swansea Bay University Health Board","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Rawlinson","suffix":""},{"id":379523580,"identity":"6f0eba65-56d2-4145-816e-2ae952231671","order_by":9,"name":"Lara Snowdon","email":"","orcid":"","institution":"Public Health Wales","correspondingAuthor":false,"prefix":"","firstName":"Lara","middleName":"","lastName":"Snowdon","suffix":""},{"id":379523581,"identity":"a8257795-efcf-4630-9d51-14cbb10d9037","order_by":10,"name":"Jonathan Shepherd","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Jonathan","middleName":"","lastName":"Shepherd","suffix":""},{"id":379523582,"identity":"78710da5-e8b9-4f64-aa4a-00184ce08ed9","order_by":11,"name":"Vaseekaran Sivarajasingam","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Vaseekaran","middleName":"","lastName":"Sivarajasingam","suffix":""},{"id":379523583,"identity":"cba4af89-395d-421a-aa4e-b2e0e26698b6","order_by":12,"name":"Alan Watkins","email":"","orcid":"","institution":"Swansea University","correspondingAuthor":false,"prefix":"","firstName":"Alan","middleName":"","lastName":"Watkins","suffix":""},{"id":379523584,"identity":"e0bb9885-d79d-44b8-a331-2295297fac88","order_by":13,"name":"Simon Walker","email":"","orcid":"","institution":"University of York","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Walker","suffix":""},{"id":379523585,"identity":"d308cdb3-c44d-46e6-a842-3585f27e467a","order_by":14,"name":"Shainur Premji","email":"","orcid":"","institution":"University of York","correspondingAuthor":false,"prefix":"","firstName":"Shainur","middleName":"","lastName":"Premji","suffix":""},{"id":379523586,"identity":"30ea04a7-c5cc-444e-a77e-adef5ab07600","order_by":15,"name":"Rabeea’h Aslam","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Rabeea’h","middleName":"","lastName":"Aslam","suffix":""},{"id":379523587,"identity":"5a6ae6f5-22c8-4845-8ee9-e25e12a57dde","order_by":16,"name":"Sophie Borgia","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Sophie","middleName":"","lastName":"Borgia","suffix":""},{"id":379523588,"identity":"f4fb40e4-2e24-4b94-b72d-133b361fbc92","order_by":17,"name":"Henry Yeomans","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Henry","middleName":"","lastName":"Yeomans","suffix":""}],"badges":[],"createdAt":"2024-11-14 08:53:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5452363/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5452363/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":72299811,"identity":"49177c92-9b60-4d06-9f2a-7f3c6438de54","added_by":"auto","created_at":"2024-12-25 01:12:29","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":113855,"visible":true,"origin":"","legend":"\u003cp\u003eCount of patients, by month, identified by the Cardiff VPT as attending due to a violence-related\u003c/p\u003e\n\u003cp\u003einjury, with the number that subsequently engaged with the VPT.\u003c/p\u003e","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5452363/v1/a3688c455544fcfa47b3b8d9.jpg"},{"id":72299195,"identity":"356e2a4c-fa07-4686-9ea2-66d65c999921","added_by":"auto","created_at":"2024-12-25 01:04:29","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":394605,"visible":true,"origin":"","legend":"\u003cp\u003eA simplified example of discontinuous risk intervals as they relate to unscheduled ED attendance\u003c/p\u003e\n\u003cp\u003eand super-spells.\u003c/p\u003e","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5452363/v1/6ff5c3049cb45a095245052e.jpg"},{"id":92152847,"identity":"86813066-230a-4668-93e6-7f8e81ea6b25","added_by":"auto","created_at":"2025-09-25 08:32:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2094928,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5452363/v1/5155b941-90fc-4a83-8736-a078d454f293.pdf"},{"id":72299193,"identity":"bc9694f8-b7f7-446c-b5f8-d25e47c7f7d8","added_by":"auto","created_at":"2024-12-25 01:04:29","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":69049,"visible":true,"origin":"","legend":"","description":"","filename":"1.Appendices.docx","url":"https://assets-eu.researchsquare.com/files/rs-5452363/v1/5dc3ba48f04cc1dc8a9236cf.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Protocol: A Quasi-Experimental Effectiveness and Cost-Effectiveness Evaluation of Emergency Department Violence Intervention Programmes in the United Kingdom","fulltext":[{"header":"Strengths and Limitations of this Study","content":"\u003cul\u003e\n \u003cli\u003eA whole population, controlled evaluation of a violence prevention team situated in an Emergency Department (ED) that accounts for patient characteristics that might influence outcomes.\u003c/li\u003e\n \u003cli\u003eProvides descriptive analyses concerning patient characteristics associated with their willingness to engage in the violence prevention intervention.\u003c/li\u003e\n \u003cli\u003eDescribes analytic methods for evaluations utilising routine ED data.\u003c/li\u003e\n \u003cli\u003eThe analysis can only be conducted on patients identified in routine data and linked to the intervention data. Some patients who wish to avoid scrutiny may avoid providing details necessary for linkage.\u003c/li\u003e\n\u003c/ul\u003e\n"},{"header":"Introduction","content":"\u003cp\u003eThose who experience serious injury due to violence are likely to attend an Emergency Department (ED). EDs are therefore ideal locations for hospital-based violence intervention programmes (HVIPs) \u003csup\u003e1\u0026ndash;3\u003c/sup\u003e. HVIPs have recently emerged as a public health response to violent victimisation \u003csup\u003e3 4\u003c/sup\u003e, but despite interest in HVIPs there has been no rigorous evaluation of this public health approach to violence in the United Kingdom (UK). Moreover, little is known about the effectiveness of patient discharge planning and referral from ED into organisations able support children involved in violence \u003csup\u003e5\u003c/sup\u003e, patients exposed to domestic violence \u003csup\u003e6 7\u003c/sup\u003e, and there is a paucity of studies considering referrals for young men involved in violence, the most dominant population in respect of assault-related attendance (ARA) \u003csup\u003e8\u003c/sup\u003e. There are even fewer studies of the effective support available to victims of sexual violence attending ED \u003csup\u003e9\u003c/sup\u003e. Despite uncertainty over effectiveness in a UK context, HVIPs have begun to be implemented. For example, the Scottish Violence Reduction Unit (VRU) has placed Navigators (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003ca href=\"http://www.mav.scot/navigator\" target=\"_blank\"\u003ewww.mav.scot/navigator\u003c/a\u003e\u003c/span\u003e\u003cspan address=\"http://www.mav.scot/navigator\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) in EDs, typically youth workers who connect with patients 25 years of age and younger \u003csup\u003e10\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eEDs work with patients attending for many reasons, including those presenting with violence-related injuries. EDs also engage in broader violence reduction initiatives: clinical staff typically receive training in adult and paediatric safeguarding, and many EDs have provisions to identify and refer children, and victims of domestic and sexual violence. However, methods of ascertainment and referral vary considerably and formal relationships with the police and other partners can often lack continuity with multiagency approaches not capturing the entirety of patients\u0026rsquo; journeys. A public health approach involving the ED would be beneficial in improving overall population health outcomes. It is particularly timely, in 2019 there were 175,764 ARAs at EDs in England and Wales \u003csup\u003e11\u003c/sup\u003e and, despite the pandemic, 119,111 ARAs in 2020 \u003csup\u003e12\u003c/sup\u003e. While knife crime, and therefore serious trauma, has risen by 71% from 2014 to 2018 \u003csup\u003e13\u003c/sup\u003e. Up to 75% of ARAs are unknown to the police \u003csup\u003e14\u0026ndash;16\u003c/sup\u003e, therefore EDs hold exclusive data on assault characteristics, patient vulnerabilities and modifiable risks, and are therefore well situated to play a significant role in the identification of violence, to investigate the circumstances of violence, and to challenge any underlying vulnerabilities or modifiable risks exposing patients to violence, whether that is through direct support, referral, or discharge planning.\u003c/p\u003e \u003cp\u003eThe need for HVIPs is aligned to broader UK Government initiatives, which aim to promote a whole system multi-agency (WSMA) \u003csup\u003e17\u003c/sup\u003e approach to violence. The 1998 Crime and Disorder Act requires the police, local government, and the National Health Service (NHS) to collaborate on joint crime reduction strategies and this includes data sharing to inform targeted responses. Violence reduction is further prioritised by the UK Government in its Serious Violence Strategy \u003csup\u003e18\u003c/sup\u003e and the UK government has allocated funds for the formation of VRUs in England and a Violence Prevention Unit (VPU) in Wales, across 18 Police and Crime Commissioner jurisdictions, with the explicit purpose of promoting the WSMA approach \u003csup\u003e19\u003c/sup\u003e. These initiatives are further aligned with a move towards active population health management, digitally enabled whole-person care and evidence-based treatment pathways outlined in the NHS future plan \u003csup\u003e20\u003c/sup\u003e. Integrated Care Systems in NHS England will be expected to specify violence prevention and reduction standards, which are incorporated into the 2021/22 NHS Standard Contract, and there are expectations that hubs will form Violence Prevention Teams similar to the police VRUs and VPU. Furthermore, a public sector duty on partnerships encouraging the prioritisation of reducing serious violence has received royal assent as a part of the Police, Crime, Sentencing and Courts Bill. This legislation includes a serious violence duty placing a statutory obligation on organisations to collaborate, communicate, and act.\u003c/p\u003e \u003cp\u003eThe overarching aim of the work proposed here is a robust effectiveness and cost-effectiveness evaluation of ED-based Violence Prevention Teams (VPTs). VPTs represent a formal collaboration between police and healthcare and embody the WSMA approach. To our knowledge, this is the first formal evaluation of a nurse-led, ED based HVIP in the UK and will address significant gaps in current understanding of their effectiveness and thereby facilitate future aspirations for evidenced-based referral pathways and discharge planning \u003csup\u003e20 21\u003c/sup\u003e. VPTs main function is to identify and support patients attending ED with assault-related injury. To facilitate this, they engage in broader pedagogical roles increasing awareness of these patients\u0026rsquo; needs, modifiable risks and opportunities to identify and refer across the ED clinical environment.\u003c/p\u003e \u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Theoretical Framework\u003c/h2\u003e \u003cp\u003eThe theoretical motivation for a WSMA approach to HVIPs is that there are many modifiable risks and vulnerabilities that, in combination, determine an individual\u0026rsquo;s exposure to violence and subsequently an ARA in an ED. Epidemiologically, these can be usefully described by shared circumstances that in turn signpost opportunities to modify risk or support patients\u0026rsquo; vulnerability, but responsibility can fall across organisations, including local government, healthcare, and criminal justice. Risks include the consumption of alcohol and other psychoactive substances \u003csup\u003e22\u0026ndash;25\u003c/sup\u003e; criminal and/or sexual exploitation and homelessness \u003csup\u003e3 26\u003c/sup\u003e. Violence tends to be more prevalent in younger, socially disadvantaged groups \u003csup\u003e27\u0026ndash;33\u003c/sup\u003e, with male, socio-economically deprived individuals being more likely to endure violence and experience assault-related injury. These characteristics further extend to personality features \u003csup\u003e34\u003c/sup\u003e, including mental health status and learning disability, and neurodevelopmental disorders \u003csup\u003e35\u003c/sup\u003e. This complex interplay of factors that promote exposure to violence, and hence lead to an ARA, highlight the need for a WSMA approach. For example, an environment might become synonymous with violence through a process of homophily \u003csup\u003e36\u003c/sup\u003e, whereby individuals with shared pursuits who are at risk of violence gather, for example street drinkers and late night drinking environments. Mitigation might include challenging reasons for frequenting such an environment, including alcohol and substance misuse counselling. Some environments might involve those who use violence to advance their interests, such as acquisitive crime, sexual assault, or sexual exploitation, in which case criminal justice or safeguarding processes to deter violence might be involved, along with support to victims. Chaotic or otherwise disadvantaged households in which domestic violence or harm to children arises might best be approached from a multi-agency process such as the Multi-Agency Risk Assessment Committee (MARAC) and formal investigation (Section 47, Children Act 1989). EDs are primary agencies receiving those who have sustained a serious injury, including those who are motivated to bypass other agencies or whose assailant is motivated to ensure their victim avoids scrutiny.\u003c/p\u003e \u003cp\u003eTreatment for an ARA in ED aims to address symptoms (e.g., injury) that may not necessarily characterise the underlying reasons for violence (e.g., alcohol dependency), and staff do not always have the resources available to address such modifiable risks and vulnerabilities. However, without addressing them, the risk of repeat unscheduled ED attendance remains, including violence recidivism. For these reasons, services like VPTs that work within a WSMA approach to better understand reasons for ARA are required. Moreover, and for those who are most vulnerable, ED may be the only realistic opportunity for patients to enter a system of care. As such, an ARA is often a sentinel event.\u003c/p\u003e \u003c/div\u003e"},{"header":"Intervention","content":"\u003cp\u003eA process and implementation evaluation that describes both the planned and implemented VPT intervention is available elsewhere \u003csup\u003e37\u003c/sup\u003e, and is further described in a Template for Intervention Description and Replication (TIDieR, Appendix 20.1).\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Intervention as Hypothesised\u003c/h2\u003e \u003cp\u003eVPTs, which emerged from the VPU violence prevention strategy, were funded by the UK Home Office and Youth Endowment Fund (YEF) with the funding administered by the VPU and the Office of the South Wales Police and Crime Commissioner (PCC). Other HVIPs in the UK are volunteer-based, whereas the VPTs are nurse-led. The original implementation for VPTs were focussed on identifying and supporting ED patients aged 11 to 25 years of age and to formalise the identification of modifiable risks and vulnerabilities, to support and advise patients, and to signpost to other services as appropriate. The VPTs also aimed to raise awareness of the service across ED clinical teams, with the aim of it becoming embedded within usual practice, and to train and upskill the clinical team to enable ascertainment and referral.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Intervention as Implemented\u003c/h2\u003e \u003cp\u003eSince November 2019, a collaborative VPT between the police and NHS has been operational in a South Wales Type I (consultant led with resus) ED in Cardiff (the capital and largest city in Wales) and a second VPT began in an adjacent South Wales Type I ED in Swansea (the second largest city in Wales) in January 2023. The VPTs initially sought to identify patients attending the ED due to violence. This remit was broadened with VPTs subsequently receiving referrals from across the hospital and other community healthcare teams (e.g., MIUs and GPs).\u003c/p\u003e \u003cp\u003eThe VPTs work with patients to gain an understanding of any circumstances contributing to their exposure to violence. They refer patients into care pathways (primary, secondary, and tertiary care, or third-sector organisations) to address any vulnerabilities or modifiable risks and can work alongside third sector (non-profit and charitable enterprise) to provide continual case-management. The VPTs also train other staff within the hospital to improve the identification of violence-related injury, to support clinical staff interactions with patients, and to maintain safeguarding procedures. In addition, the EDs at Cardiff and Swansea take part in Information Sharing to Tackle Violence (ISTV), in which anonymous data and intelligence regarding violent incidents are shared with Community Safety Partnerships. These anonymised data enable partner resources to be best used for violence prevention, part of the Cardiff Model for violence prevention \u003csup\u003e38\u003c/sup\u003e. Following implementation, both VPTs expanded the age range of patients to encompass all age groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Usual Care\u003c/h2\u003e \u003cp\u003eUnder usual practice, clinical ED staff are obliged to undertake safeguarding activities, and provide for those attending due to violence. However, provision varies across EDs. Under usual practice, when people attend ED with an injury suspected to be caused by violence, their injuries are treated and the patient is encouraged to contact the police, or have the ED contact the police on their behalf. In cases of serious injury involving weapon use, the ED is obliged to contact the police irrespective of patient consent. In terms of general safeguarding, all patient-facing clinical staff are expected to have up-to-date safeguarding training, and thus to carry out safeguarding tasks. Patients who are experiencing domestic violence can be referred to an Independent Domestic Violence Advocate (IDVA). Patients attending due to sexual assault can be referred to an Independent Sexual Violence Advisors (ISVA). For the ten control EDs in Wales, two EDs have an IDVA, and two others have access to an IDVA not based in their ED. Furthermore, ED staff can also refer patients into a MARAC, typically cases where the criteria are not met for formal safeguarding but the clinician suspects that something is not right. To facilitate, Multi-agency Referral Forms (MARFs) are filled out for children who require safeguarding and VA1 forms (to support the referral of vulnerable adults) are completed for vulnerable adults who require safeguarding.\u003c/p\u003e \u003cp\u003eThere are IDVAs based in the intervention EDs in Cardiff and Swansea. Only Cardiff has an ISVA. Broadly, usual practice focuses on children and victims of domestic violence. The patients eligible for VPT support are therefore those who are not eligible for support from the IDVA or ISVA and are typically over ten years of age. Apart from the IDVAs in control EDs, none have additional resources specifically dedicated to the role of supporting patients attending due to violence, relying mainly on existing clinical staff to support safeguarding within their departments. This may involve naming an existing member of staff as a safeguarding ambassador or having a nurse act as safeguarding lead for the department. Across Welsh EDs, some control EDs have provisions for victims of domestic violence that includes cards with the \u0026ldquo;Live Fear Free\u0026rdquo; helpline that they can give to patients experiencing domestic violence and who do not meet the criteria for a MARAC referral. Similarly, none of the control EDs have processes in place to support patients\u0026rsquo; referral to outside agencies. If patients disclose that they are struggling with issues, or if staff suspect patients are experiencing an issue, then referrals will made by clinical staff. However, this is not the same as VPT members working with patients to identify modifiable risks and vulnerabilities that contribute to their experience of violence. The VPTs in Cardiff and Swansea are therefore unique.\u003c/p\u003e \u003c/div\u003e"},{"header":"Aims and Objectives","content":"\u003cp\u003eThe overarching aim of the Emergency Department Violence Intervention Programme (EDVIPE) evaluation is to determine whether VPTs are effective and cost-effective from the perspective of the NHS.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObjective 1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo assess whether patient involvement with a VPT reduces the likelihood of unscheduled ED re-attendance. We consider case and control patients\u0026rsquo; ED unscheduled reattendance for a minimum of 12 months following the initial ARA.\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObjective 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo determine whether the presence of the VPT improves ascertainment of ARAs in ED attendances. We will consider the change in identified ED ARAs across intervention implementation in case and control EDs in Wales.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObjective 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo derive the costs of the VPT and compare those to the benefits of the intervention and understand whether the VPT represents value for money from an NHS perspective. If an effect is observed, then models will estimate the health impacts, costs and potential savings over a longer time (e.g. 10 years) period and for a national roll-out.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"Ethics","content":"\u003cp\u003eThe approval to access and analyse data housed in the SAIL Databank \u003csup\u003e39\u003c/sup\u003e, an ISO 27001 certified and UK Statistics Authority accredited secure data environment, was granted by the SAIL Independent Information Governance Review Panel (IGRP) (Ref: 1421). The IGRP comprises representatives from various organisations and sectors including the British Medical Association, Welsh Government, Public Health Wales, National Research Ethics Service, Digital Health and Care Wales (DHCW), Swansea Bay University Health Board, and members of the public. All routinely collected anonymised data held in SAIL are exempt from consent due to the anonymised nature of the databank (Section 251, Control of Patient Information; 2006 National Health Service Act). At no time will identifiable data be made available to the research team. ED staff will curate data pertaining to patients\u0026rsquo; exposure to the intervention, which will be passed to DHCW, where it will be anonymised, and a project specific anonymous linkage field (ALF) added, as will a residential anonymous linkage field (RALF). These data will be passed to SAIL for linkage to anonymised VPT clinical data.\u003c/p\u003e \u003cp\u003eParticipant consent is not required because all the study outcome data involving patients are anonymised before they are incorporated into the SAIL databank. As the SAIL databank is fully anonymised, it does not fall into the remit of the National Information Governance Board who provide section 251 (formerly section 60) exemption to use identifiable data without consent. Human Ethics and Consent to Participate declarations are not therefore applicable.\u003c/p\u003e"},{"header":"Patient and Public Involvement","content":"\u003cp\u003eExtensive Patient and Public Involvement and engagement (PPIE) has been and will continue to be undertaken. The rationale is that many patients managed by the VPTs will be vulnerable, with some at the beginning of their journey in the support they receive. The expectations were that these patients would be unlikely to reflect meaningfully on the VPT within the study timeline and therefore alternative opportunities to explore patients\u0026rsquo; perceptions was required. Furthermore, follow-up qualitative work with young adults in emergency care, the dominant group in ED, requires considerable resourcing and suffers from high levels of attrition \u003csup\u003e40\u003c/sup\u003e. We therefore sought PPIE engagement in order that those with experience of the emergency healthcare system were able to feed into the project, co-produce methods, provide their interpretation of the results and assist with interpretation and dissemination. Groups include survivors of domestic violence, carers, those who have experienced alcohol and drug dependence, homelessness, sexual exploitation, and mental health issues. One PPIE co-investigator was appointed to lead on monitoring equality and diversity, with a second, and experienced, PPIE co-investigator supporting inexperienced PPIE members. PPIE activity was captured using the short-form Guidance for Reporting Involvement of Patients and the Public (GRIPP) \u003csup\u003e41\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThe PPIE groups include lay members with experience of PPIE work: Service Users for Primary and Emergency Care Research (SUPER; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.primecentre.wales/ppi.php\u003c/span\u003e\u003c/span\u003e) who provided lay perspectives to the research team when developing, and conducting the research, with subsequent engagement undertaken to strengthen the relevance, quality and dissemination opportunities of the research. Additional PPIE members were recruited. These members had lived experience relevant to the patients who are the subject of the intervention (homelessness and domestic violence). SUPER and the PPIE co-investigators advised on how the research team should engage with those who have lived experience but had not have prior experience of PPIE involvement.\u003c/p\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e7.1. PPIE Activity\u003c/h2\u003e\n \u003cp\u003eSUPER provided feedback on the original proposal, and then provided advice on how the materials should be developed for the two less experienced PPIE groups. These PPIE groups with lived experience were recruited to give input on the protocol. One group consisted of people with lived experience of homelessness and related conditions, recruited from a charity supporting those who are experiencing homelessness. The second group was comprised of survivors of domestic violence and were recruited through Welsh Women\u0026rsquo;s Aid. The PPIE group members gave feedback on the VPTs and provided their first impressions of the research protocol. The groups also helped to develop a list of potential organisations to disseminate research findings, and discussed methods on how this could be achieved. They explained what may be going on in the lives of people who experience violence, and where people might turn to for support.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003e7.2. Results\u003c/h2\u003e\n \u003cp\u003ePPIE contributed to the initial development of the research proposal, and the development of the research protocol, in the following ways:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eInitial consultation in planning the funding application.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe research proposal was reviewed in July 2021, and the input and advice contributed to a successful funding application. The research team also responded to the formative comments when developing the protocol and continue to build on them.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003ePPIE work influenced the types of data being included in the study. For example, an issue was raised whether some patients would admit to experiencing violence in ED, and therefore opportunities for the VPTs to improve ascertainment.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eData on ethnicity from the 2011 and 2021 Census, which will also have some indication of those residing in refugee centres (and nursing homes, hostels, etc.), will be included in the study following recommendations that racial violence and the experiences of asylum seekers should be considered.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eIt was further indicated that considering school attendance and exclusions data, free school meals and special educational needs would be valuable, in addition to educational attainment.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe EDVIPE stakeholder reference group (SRG) now includes representation from primary care, secondary care and the third sector, after working with these sectors was suggested.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eFollowing feedback, EDVIPE now involves PPIE groups with lived experience of different types of violence.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eFollowing consultation with PPIE lived experience groups, development and refinement of the protocol and the addition of exploratory work was undertaken to:\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eBetter understand the pathways patients follow up to their attendance in ED, notably whether General Practitioner (GP) consultations were not acted upon.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eTo consider patient ascertainment and therefore eligibility for the intervention in ED, as some victims may not realise that they are victims of violence.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eWhether one or two nurses are sufficient to manage an expected high caseload of patients attending ED due to assault, and whether the lack of 24-hour VPT provision would mean some patients are missed.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eSome patients might choose to avoid the police and therefore be less reluctant to receive support from the police, or police aligned services. We might explore this in any VPT referral data available.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThere were concerns that intervention-related activity might become known to the perpetrator, therefore elevating risk of subsequent harm. We can consider the immediacy of post-intervention re-attendance by patient group.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThat there are unique challenges for those who are disabled, both in terms of ability to engage and the nature of the support required.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eSome patients, notably those with children, may be less willing to engage as they would not want to risk losing their home or children. We can consider engagement by gender and presence of dependent children in the patient\u0026rsquo;s residence.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe judicial system emphasises the right for both parents to be involved with their children, if any. Involvement of parents in the court system might be associated with a blunted intervention effect.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThat PPIE involvement in future diffusion and dissemination activity would lend credence to the project\u0026rsquo;s communication strategy.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Methods and Analysis","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e8.1. Design and conceptual framework.\u003c/h2\u003e \u003cp\u003eA controlled longitudinal whole population (Wales, UK) natural experiment. The intervention in Cardiff began November 2019, and January 2022 in Swansea. Due to earlier changes in EDDS coding, these data are consistent and available from January 2012. Intervention data collection therefore begins in November 2019 and ends in August 2023, allowing a 12-month follow-up of patients until August 2024.\u003c/p\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e8.1.1. Objective 1 - Effectiveness\u003c/h2\u003e \u003cp\u003eWe hypothesise that engagement with the VPT will help patients overcome modifiable risks and receive support for vulnerabilities, and that therefore the intervention will reduce the recurrence of unscheduled ED attendance.\u003c/p\u003e \u003cp\u003eOther than those who are most seriously injured, patients will register at ED reception and be triaged, at which point the most appropriate pathway through ED will be determined. At reception patients will be asked about the reason for their attendance, including whether it was due to an assault, data that becomes a part of the Patient Management System (PMS) \u003csup\u003e42\u003c/sup\u003e. Patients may not disclose that the reason for their attendance was assault related. They might be reluctant, the perpetrator may have accompanied them, or they may wish to avoid scrutiny. One function of the VPTs is to work across clinical teams to improve ascertainment of ARAs. The result being that patients can be stratified according to the extent that they engage with the intervention.\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003ePatients identified in the ED PMS data as having attended due to an assault, but with no further contact with the VPT.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003ePatients identified in the ED PMS or VPT data as having attended due to an assault, but did not further engage with the VPT.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003ePatients identified in the ED PMS and VPT data and who engaged with the VPT.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eThe primary analysis concerns group iii. The reasons for patients not engaging are potentially related to underlying characteristics and in secondary analyses we will explore this. However, it is reasonable to assume that the three groups represent varying levels of intervention dose, and therefore secondary analyses and therefore analyses using groups i to iii will be informative.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e \u003ch2\u003e8.1.2. Objective 2 - Ascertainment\u003c/h2\u003e \u003cp\u003eOur second hypothesis is that intervention implementation improves ARA ascertainment.\u003c/p\u003e \u003cp\u003ePatients attending ED can do so repeatedly within periods of time. This frequency is likely associated with the modifiable risks associated with ARA and is of interest here. For Objective 2 it is therefore appropriate to determine the proportion of attendances identified as violence-related. This generates time series data. The outcome of interest is therefore the count ARAs across all EDs. ARA attendance is defined as ARA in the ED PMS or, in the case of intervention sites, in the ED, PMS or VPT data. Codes indicating Provider Site in (the ED in its hospital) is available in the Emergency Department Data Set (EDDS), and this allows comparison between intervention EDs and control EDs. While intervention EDs have been in continual service since before the time series start dates, this is not so for all Type I EDs in Wales; there have been several changes with some EDs closing and others opening or being modified to receive additional patients. This, coupled with the intervention sites located at two of the largest hospitals in Wales, reduces the scope for selecting matching control sites \u003csup\u003e43\u003c/sup\u003e, leading to potentially important baseline differences in the interrupted time series data \u003csup\u003e44\u003c/sup\u003e. The counterfactual will therefore be ARAs across all control Type I EDs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e \u003ch2\u003e8.1.3. Objective 3 \u0026ndash; Cost-Effectiveness\u003c/h2\u003e \u003cp\u003eWe aim to determine whether the VPT represents value for money. The primary outcome will be quality adjusted life years, which will be estimated based on effectiveness estimates comparing ED attendance, reattendance, and any injuries received for those engaging in the VPT service relative to those who do not. Costs will be captured from an NHS perspective, reflecting the cost of the intervention but also other costs to the NHS due to referral. A secondary cost-effectiveness analysis will undertake a societal perspective, which will explore additional costs across social care, the police and the third sector (see Section 9).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section3\"\u003e \u003ch2\u003e8.1.4. Secondary Analyses\u003c/h2\u003e \u003cp\u003eWe aim to co-produce study protocols with collaborators and PPIE groups, to provide opportunities for them to shape secondary and additional epidemiological analyses. This facilitates opportunities to realise and contribute to what is a rapidly changing policy area. One example is our inclusion of school attainment, exclusions, and attendance in analyses, which have been highlighted in these early discussions. VPUs and VRUs have made little headway working with the education system to challenge the causes of violence, and this has been identified as a priority \u003csup\u003e45\u003c/sup\u003e. Furthermore, additional exploration is planned to characterise those excluded from the intervention but who have available ALFs.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e8.2. Population and Data\u003c/h2\u003e \u003cp\u003eThis is a whole-population evaluation, including all residents of Wales, UK. Data is housed in the SAIL databank \u003csup\u003e39\u003c/sup\u003e.\u003c/p\u003e \u003cdiv id=\"Sec19\" class=\"Section3\"\u003e \u003ch2\u003e8.2.1. Data\u003c/h2\u003e \u003cdiv id=\"Sec20\" class=\"Section4\"\u003e \u003ch2\u003e8.2.1.1. Violence Prevention Team Data (Cardiff and Swansea)\u003c/h2\u003e \u003cp\u003eIntervention sites record patient details (name, date of birth, gender), NHS number, extent of engagement with the VPT and whether any subsequent referral was made.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e \u003ch2\u003e8.2.2. Administrative Data\u003c/h2\u003e \u003cp\u003eSeveral administrative datasets are available to characterise patients and summarised in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u0026ndash; EDVIPE administrate data sets.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDataset\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eName and Summary\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eADDE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAnnual District Death Extract provides the week of birth and date of death, used to describe left- and right-side censoring of study participants.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOPRD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOutpatient Referral Dataset\u003c/p\u003e \u003cp\u003eOPRD includes data from outpatient referrals from primary care which will help in understanding the referral pathway to secondary care. This data includes all clinical referrals from General Practitioners, General \u0026amp; Community dental practitioners, A\u0026amp;E departments, walk-ins, consultant-to-consultant referrals.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEDW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePatient Episode Data for Wales Dataset\u003c/p\u003e \u003cp\u003ePEDW covers people domiciled in Wales and treated in Welsh and English NHS Trusts and English Trusts and includes data on inpatient and day case activities. This includes spells and episode data on hospital admissions, which can used to track the frequency of patients attending services for treatment arising from an ARA.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWDSD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWelsh Demographic Service Dataset\u003c/p\u003e \u003cp\u003eWDSD includes all individuals registered with a Welsh General Practitioner and via anonymisation identifies household groups.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWIMD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWelsh Index of Multiple Deprivation\u003c/p\u003e \u003cp\u003eWIMD is the Welsh Government\u0026rsquo;s official measure of relative deprivation for small areas in Wales, based eight domains including income, employment, health, and access to services. Typically grouped into fifths or \u0026ldquo;quintiles\u0026rdquo;, the WIMD is included in several NHS datasets.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2011, 2021 Census\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eA census of the UK population is taken every ten years and includes questions relating to key demographics. Patient entry into and exit from EDVIPE can mean some will be missed in the 2011 Census (e.g., born after 2011), and some will be missed in the 2021 Census (e.g., died before 2021). Hence data from both censuses is required.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e summarises key patient socioeconomic and demographic characteristics, required to enable secondary analyses in relevant sub-groups.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u0026ndash; EDVIPE patient characteristics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (from Week of Birth, WoB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWDS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWDS, 2011 and 2021 Census\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEthnicity\u003c/p\u003e \u003cp\u003eAsian (Bangladeshi, Chinese, Indian, Pakistani, Other Asian)\u003c/p\u003e \u003cp\u003eBlack (Caribbean, African, Other Black)\u003c/p\u003e \u003cp\u003eMixed (White and Asian, White and Black African, White and Black Caribbean, Other Mixed or Multiple ethnic groups)\u003c/p\u003e \u003cp\u003eWhite (English, Welsh, Scottish, Northern Irish, British, Irish, Gypsy or Irish Traveller, Roma, Other White)\u003c/p\u003e \u003cp\u003eOther (Arab, Any other ethnic group)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2011 and 2021 Census\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQuintile of Residential Deprivation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWIMD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUrban/rural residential classification\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2011 and 2021 Census\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section3\"\u003e \u003ch2\u003e8.2.3. Secondary Outcomes\u003c/h2\u003e \u003cp\u003eTo characterise the WSMA involvement of patients involved in the intervention, broader pathways will be explored across several related data sets, summarised in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u0026ndash; EDVIPE secondary outcomes\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDataset\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eName\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCAFCASS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eChildren and Family Court Advisory and Support Service Wales Family Justice Data Set\u003c/p\u003e \u003cp\u003eCAFCASS includes information for residents of England and Wales on cases of divorce, private law, family law act, public law, adoption, family law applications. It also includes information on marriage and divorce characteristics. The information on cases also includes information on number of children involved and types of hearing.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMoJ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMinistry of Justice: Data First\u003c/p\u003e \u003cp\u003eMagistrates' court defendant data\u003c/p\u003e \u003cp\u003eCrown Court defendant data\u003c/p\u003e \u003cp\u003eCriminal courts and prisons data\u003c/p\u003e \u003cp\u003ePrisoner custodial journey data\u003c/p\u003e \u003cp\u003eFamily Court data.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNational Pupil Database\u003c/p\u003e \u003cp\u003eThis dataset has four broad categories of demographics, attainment, absence and exclusion, and children in need and looked after children.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePolice Data\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePolice Crime Dataset\u003c/p\u003e \u003cp\u003ePending applications for police crime data from all four Welsh forces are progressing and will be explored.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSMSD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSubstance Misuse Dataset\u003c/p\u003e \u003cp\u003eSMDS, also known as Welsh National Database for Substance Misuse (WNDSM) has data for people in Wales who present for substance misuse treatment. It includes details on assessments, referrals and treatment history.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWLGPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWelsh Longitudinal General Practitioner Dataset\u003c/p\u003e \u003cp\u003eWLGP contains clinical information from General Practice in Wales, including diagnoses and referrals into tertiary care.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003eNote: \u003csup\u003e1\u003c/sup\u003e Data sharing agreements are currently being developed to bring all-Wales police crime data into SAIL.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e8.3. Cohort Inclusion and Exclusion Criteria\u003c/h2\u003e \u003cp\u003eAll residents of Wales who 11 years of age or older are eligible for inclusion. Residents of Wales will be defined through their identification in the WDSD.\u003c/p\u003e \u003cp\u003eThe NHS assigns each patient domiciled in the UK a unique number. This NHS number links across various NHS data systems. The encrypted and anonymised ALFs are derived from these NHS numbers. Therefore, patients attending ED whose identity cannot be connected to an NHS number (e.g. overseas visitors and tourists) will not have a corresponding ALF and will, by necessity, be excluded from analyses.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e8.4. Allocation\u003c/h2\u003e \u003cp\u003eIntervention patients will be identified in the VPT data and will have attended intervention EDs (in Cardiff and Swansea), subject to the above inclusion criteria. Control patients will be identified in the EDDS, and where ED attendance was not in an intervention ED.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e8.5. Progression criteria\u003c/h2\u003e \u003cp\u003eThis is a definitive study. As the primary focus of the study uses routinely collected data, which is available for analysis subject to information governance permissions and extraction, progression criteria are not applicable.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e8.6. Sampling\u003c/h2\u003e \u003cp\u003eFrom the intervention sites\u0026rsquo; data (Section 8.2.1), between October 2019 and December 2022, the Cardiff VPT contacted 2,312 patients, of whom 77% accepted VPT support (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We conservatively estimated that there will be 2,500 patients that engaged with the Cardiff VPT across the four years (2019\u0026ndash;2024) of VPT operation, and a further 900 from the two years (2021\u0026ndash;2024) operation in Swansea. Across the entirety of Wales, there are approximately 1M ED attendances each year.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e-= Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e About Here =-\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e - Count of patients, by month, identified by the Cardiff VPT as attending due to a violence-related injury, with the number that subsequently engaged with the VPT.\u003c/p\u003e \u003cp\u003eInitial estimates from Cardiff VPT suggest that 3% of those engaging with the VPT reattended ED at least once within one year, compared to 23% patients who did not engage with the VPT. Data from 2015 and 2016 suggest that the frequency of unscheduled attendances for patients with at least one ARA (mean\u0026thinsp;=\u0026thinsp;2.35 attendances) is greater than patients making an unscheduled attendance without evidence of an assault (mean attendances\u0026thinsp;=\u0026thinsp;1.73). For a simple Cox survival model, (α\u0026thinsp;=\u0026thinsp;0.05, β\u0026thinsp;=\u0026thinsp;0.90) and a hazard ratio of 0.8, a total N of 845 is required.\u003c/p\u003e \u003cp\u003eTo realise the recurrent nature of analyses, simulation \u003csup\u003e46\u003c/sup\u003e (1,000 estimates per point estimate) was used across varying follow-up periods, which suggests a 12 month follow-up period and total N of 300 is adequate to identify a significant effect. By increasing the number of controls, statistical power will be further enhanced \u003csup\u003e47\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section2\"\u003e \u003ch2\u003e8.7. Analytic Strategy\u003c/h2\u003e \u003cdiv id=\"Sec28\" class=\"Section3\"\u003e \u003ch2\u003e8.7.1. Objective 1 - Effectiveness\u003c/h2\u003e \u003cp\u003eOur primary outcome is unplanned ED attendance. It represents the cost to the NHS of serious healthcare events and acts as a proxy for events eliciting acute healthcare needs. EDs provide acute care for patients without prior appointment, and the aftercare of patients who have received ED treatment but where there is no alternative provision (e.g., for out of area tourists). There can, therefore, be follow-up and planned appointments in ED. These appointments in ED will be made where, for example, there is an element of diagnostic uncertainty and a review is required in the ED context, or for patients where other follow-up arrangements are likely to fail (e.g., visitors to the area without access to primary care) \u003csup\u003e48\u003c/sup\u003e. Follow-up and planned appointments in ED are typically a continuation of the initial unplanned attendance or referral from another healthcare provider and are not valid outcomes for EDVIPE, as they are not elicited in response to acute healthcare need. Thus, for Objective 1 we will censor the timeline. Left-side censoring at birth or when someone takes up residence in Wales. Right-side censoring when someone dies or moves out of Wales. We further interval censor the timeline, to account for repeat ED attendances associated with a health event, such as referral from a local emergency hospital to a Major Trauma Centre.\u003c/p\u003e \u003cdiv id=\"Sec29\" class=\"Section4\"\u003e \u003ch2\u003e8.7.1.1. Discontinuous Risk Interval\u003c/h2\u003e \u003cp\u003eAccounting for periods when individuals are not at risk is an essential consideration in repeated time-to-event models \u003csup\u003e49\u0026ndash;52\u003c/sup\u003e. The clearest example in the current context is ensuring time at risk does not extend beyond date of death or originates before birth. Similarly, time at risk will also be left side censored, if patients move into Wales, and right side censored if they moved away from Wales. With no adjustment for these discontinuous risk intervals, the time at risk will be incorrect, increasing a greater likelihood of Type II errors.\u003c/p\u003e \u003cp\u003eHow patients are routed through emergency care pathways in Wales also influences time at risk. ED attendances are mainly determined by the acuity of the patient\u0026rsquo;s condition and the urgency with which they need to be seen. These decisions can be made by the Welsh Ambulance Service Trust (WAST), a Minor Injuries Unit (MIU) or in the local ED. It is feasible that a patient initially attends a local ED to be stabilised, is assessed, and requires referral to a Major Trauma Centre (MTC), or Trauma Unit (TU). Each MTC and TU are attached to an ED, and therefore in response to severe injury, patients are registered in more than one ED if they are referred from an ED without trauma facilities, to EDs that are attached to a TU or MTC.\u003c/p\u003e \u003cp\u003eIn Wales, emergency care in Wales is provided in MIUs, EDs (Local Emergency Hospitals, LEH, and Rural Trauma Facilities, RTFs), TUs, and MTCs. There is one MTC in Cardiff UHW, which services South and West Wales, and South Powys, and acts as a TU for the local population. Morriston Hospital in Swansea is a TU, but with additional specialist services (e.g., orthoplastics) meaning some patients otherwise destined for the MTC would be referred there instead. North Wales is serviced be the MTC in the Royal Stoke University Hospital in England.\u003c/p\u003e \u003cp\u003ePatient admission is described using spells and super-spells. A spell represents patient care by speciality, and a super-spell is a collection of spells \u0026ndash; two spells are included in the same super-spell if admission to the second speciality is within 48 hours of discharge from the first. ED attendances included within the same super-spell (e.g., transfer from an ED to an MTC) will therefore be attributable to the same initiating event, whether planned or unplanned. Therefore, only unscheduled attendances at the initiation of a super-spell will be included, and additional ED attendances within the same super-spell will be analytically censored. The duration of a super-spell constitutes a discontinuity in a patient\u0026rsquo;s exposure to risk, which is then accommodated in our analytic approach, this is described further in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e-= Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e About Here =-\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e - A simplified example of discontinuous risk intervals as they relate to unscheduled ED attendance and super-spells.\u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea hypothetical analytic period runs from day 0 to day 20. Patient 1 lived in Wales for the duration of the study but had no ED attendances. Patient 2 lived in Wales at the start of the analytic period, but died on day 18, they had two ED attendances, with the first resulting in a two day stay in hospital. Their time at risk for the first ED attendance is three days, and for the second six days, with a total time at risk of 16 days. Patient 3 was born in Wales on day two, had three ED attendances and was alive and living in Wales at the end of the analytic period. However, their second ED attendance (e.g., transfer to a MTC) occurred in a super-spell associated with the first ED attendance, and this is dropped from consideration. They therefore had two ED attendances, with the time at risk for the first as four days, and the time at risk for the second two days. Their total time at risk is 10 days.\u003c/p\u003e \u003cp\u003eWe assume that unscheduled ED attendances are independent and unordered, and therefore assume a common baseline hazard for all events. Thus, the hazard function, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\lambda\\:}_{ik}\\left(t\\right)\\)\u003c/span\u003e\u003c/span\u003e, for the \u003cem\u003ek\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e ED event for each subject (\u003cem\u003ei\u003c/em\u003e) is:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{\\lambda\\:}_{ik}\\left(t\\right)={\\lambda\\:}_{0}\\left(t\\right){e}^{{x}_{ik}\\beta\\:}\\)\u003c/span\u003e \u003c/span\u003e1\u003c/p\u003e \u003cp\u003eHowever, it is feasible that operational differences across Local Health Boards (LHBs) might influence the likelihood that ED attendance varies across groups of patients. For example, initiatives that influence the likelihood patients are conveyed into emergency care, or alternative provision in the community for lower acuity patients. We introduce a random effect, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{i}\\)\u003c/span\u003e\u003c/span\u003e, to describe variation in risk, or frailty, can be introduced \u003csup\u003e53\u003c/sup\u003e:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{\\lambda\\:}_{ik}\\left(t\\right)={\\lambda\\:}_{0}\\left(t\\right){\\alpha\\:}_{i}{e}^{{x}_{ik}\\beta\\:}\\)\u003c/span\u003e \u003c/span\u003e2\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec30\" class=\"Section4\"\u003e \u003ch2\u003e8.7.1.2. Dates and Times\u003c/h2\u003e \u003cp\u003eRecords on date and time of events in routine data requires consideration. With most events recorded by day, we are obliged to use day as the primary measure of time, and therefore collisions will occur.\u003c/p\u003e \u003cp\u003eIn EDDS the date and time of attendance and discharge are recorded, and several events can occur on the same day. For example, it is feasible that a patient attends ED on multiple occasions and on the same day. ED attendances on the same day are collapsed onto one another under the super-spell assumption. In the WDSD, while date of birth is approximated to the Monday of the week in which they were born (their Week of Birth, WoB), date of death is accurate to that degree of granularity.\u003c/p\u003e \u003cp\u003eFor example, a patient might attend ED at 9am and die at 2pm on the same day. To avoid inconsistencies such as this patient entering the study cohort after death, we will adopt the usual approach and add a small value (epsilon, ε) to dates to preserve the chronological order of events \u003csup\u003e54\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec31\" class=\"Section4\"\u003e \u003ch2\u003e8.7.1.3. Matching\u003c/h2\u003e \u003cp\u003eIt is not possible to manipulate allocation to treatment, and we therefore rely on quasi-experimental methods to infer treatment effects. The purpose of matching control and intervention patients is to allow derivation of the average treatment effect under the assumption that allocation is not conditional on observed confounders. Matching on values of covariates between the treated and untreated groups avoids the bias introduced by covariates that influence the outcome variable. The goal being to find a subset of data that is closest to an exact match on observed covariates \u003csup\u003e55\u003c/sup\u003e. The primary analysis will be conducted using Coarsened Exact Matching, with a minimum 1:1 ratio, with secondary analyses undertaken using Propensity Score Matching.\u003c/p\u003e \u003cp\u003eFor categorial covariates, exact matching is a sensible alternative. Here a match is considered acceptable only if the values of all covariates are equal between treated and untreated potential matches. Continuous covariates can, for the purpose of using exact matching, be turned into categorial covariates by defining category intervals; the resulting method is called coarsened exact matching. For matching, covariate categories can be grouped together to form larger (and fewer) categories, improving efficiency \u003csup\u003e55\u0026ndash;57\u003c/sup\u003e. This approximates to a fully blocked experiment and requires temporarily coarsening variables.\u003c/p\u003e \u003cp\u003ePropensity score matching reduces the multi-dimensional space of covariates to a single summary covariate, the propensity score, typically derived using a logistic equation and performing regression on the treatment group for the covariates chosen to match with. The proximity of a potential match to the given subject is estimated in terms of the closeness of their propensity scores. This method simplifies subsequent analyses as matching refers only to one variate, the propensity score. However, propensity score matching can increase, rather than decrease, the imbalance of the covariates in the samples \u003csup\u003e55\u003c/sup\u003e. A challenge is optimising the trade-off between the quality of matches and the sample size. A larger sample reduces sampling error and can increase the study power, but including matches of lower quality may lead to greater residual imbalance \u003csup\u003e58\u003c/sup\u003e. The R package Matching Frontier attempts to address these issues by calculating the entire balance-sample size frontier, from which the user can easily choose one, several, or all subsamples to use for their final analysis, given their own choice of imbalance metric and quantity of interest \u003csup\u003e59 60\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section4\"\u003e \u003ch2\u003e8.7.1.4. Missing Data\u003c/h2\u003e \u003cp\u003eMissing data may arise specific to the use of census data, where it is feasible that records are missing entirely, or patients did not disclose personal characteristics. However, give that these characteristics are recorded across multiple data sets the expectation is that missingness will minimally impact on the analyses.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec33\" class=\"Section3\"\u003e \u003ch2\u003e8.7.2. Objective 2 - Ascertainment\u003c/h2\u003e \u003cp\u003eThe hypothesis is that a violence-attendance is more likely to be ascertained as such with the additional VPT resource present in ED. All unplanned attendances can be coded as ARA, or not (1, 0). The time series begins in January 2012, and the two sites (Cardiff and Swansea) can be individually compared with all other Type I EDs in Wales. This indicates that a difference-in-difference model \u003csup\u003e61\u003c/sup\u003e is appropriate. We will analyse Cardiff first, using the implementation in Swansea as a future replication of the intervention to facilitate a more robust causal interpretation of any effect. Additional descriptive analysis can explore patient groups (age, gender, time of attendance) more likely to be ascertained by the VPTs. Unlike Objective 1, we are interested in all ED attendances as the opportunity to provision safeguarding is applicable across all ED attendances, irrespective of the frequency of attendance.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section3\"\u003e \u003ch2\u003e8.7.3. Objective 3 - Cost-effectiveness Analysis\u003c/h2\u003e \u003cp\u003eThe aim of the economic evaluation is to understand the costs and consequences of providing a VPT versus no VPT within EDs. For this study, we will focus on the objectives of understanding whether the VPT represents value for money from an NHS perspective.\u003c/p\u003e \u003cdiv id=\"Sec35\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.1. Methods\u003c/h2\u003e \u003cp\u003eA within study analysis will determine the short-term cost-effectiveness of offering VPT relative to not offering VPT within EDs. Estimates derived from this study will be used to inform a long-term decision analytic model examining the cost-effectiveness of providing, versus not providing, VPT within EDs over a longer time horizon. Alternative perspectives for the analysis will be considered, with the NHS perspective being the base case. Cost-effectiveness will be presented using standard statistics.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec36\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.2. Exposure variable\u003c/h2\u003e \u003cp\u003eThe exposure variable consists of patients engaging with the VPT.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec37\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.3. Outcome variables\u003c/h2\u003e \u003cp\u003eHealth-related outcome variables include improved mental health outcomes, measured as a reduction in primary and secondary care visits for a mental health reason, reduced substance use, improved physical health, measured as a reduction in assault-related attendances in ED, and improved health-related quality of life (HRQoL), measured using injury-related estimates from the literature. Data on all outcomes will be sourced from the SAIL databank \u003csup\u003e39\u003c/sup\u003e. Estimates will be measured on a quarterly basis for the follow-up period (12\u0026ndash;48 months). HRQoL will be estimated using the literature, in line with other studies on hospital-based violence prevention programs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec38\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.4. Within study statistical analysis.\u003c/h2\u003e \u003cp\u003eDifferences in overall mean outcomes will be analysed using a fixed effects panel regression model, where we will control for individual and time varying effects \u003csup\u003e62\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec39\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.5. Covariates of interest.\u003c/h2\u003e \u003cp\u003eThe following covariates will be included within the fixed effects panel regression model: patient age, gender, severity of injury, number of years VPT has been operating, year, and quarter. Referral uptake will be treated as a mediating variable and not adjusted for in the models.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec40\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.6. Long term cost-effectiveness analysis (CEA)\u003c/h2\u003e \u003cp\u003eA Markov decision model with quarterly cycles will be used to estimate the cost-effectiveness of offering VPTs in EDs. To determine cost effectiveness using a health and social care perspective (primary analysis), the Markov model will be health-state based, with key health states likely to be healthy, injured, or dead. To support a multi-sector perspective (secondary analysis), the Markov model will be event-based.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec41\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.7. Resource use\u003c/h2\u003e \u003cp\u003eHealthcare resource use will include number of and type of visits with\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eprimary care (e.g., GP),\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003esecondary care (e.g., inpatient stays, outpatient clinic appointments), and\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eother services (e.g., other healthcare professional visits, physiotherapy, occupational therapy, mental health therapy, drug and alcohol treatment).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec42\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.8. Cost data\u003c/h2\u003e \u003cp\u003eCost data will be estimated for health service use based on resource use from the SAIL databank with appropriate unit costs applied. Third sector resource use will be valued using cost data provided in the resource use questionnaires. Other Unit Cost estimates (e.g., NHS reference costs, Unit Costs of Health and Social Care) will be used to supplement where needed.\u003c/p\u003e \u003cp\u003eDifferences in overall mean total health care costs (including primary and secondary care) between groups will be analysed using GLMM. Cost data are known to be left-skewed with a substantial number of zeroes and there is no single dominant method for analysis \u003csup\u003e63\u003c/sup\u003e. We will test several potential models from the GLMM family and identify the correct link function and distribution to use. Potential effect modifiers include age, gender, and deprivation. Mediating variables include intervention engagement. Potential confounding variables include age, gender, deprivation, number of years VPT has been operating. The regression model will also include time variables recording the year and quarter.\u003c/p\u003e \u003cp\u003eCosts and benefits will be discounted at 3.5% per annum as recommended by the HM Treasury \u003csup\u003e64\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec43\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.9. Time horizon.\u003c/h2\u003e \u003cp\u003eThe time horizon for the decision analytic model will extend beyond the time frame of this study, to encompass a period of time where, if possible, all relevant costs and outcomes for this appraisal have been accrued \u003csup\u003e65\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec44\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.10. Cost-effectiveness Analysis Outcomes.\u003c/h2\u003e \u003cp\u003eFor the primary analysis, we will report the incremental cost effectiveness ratio (ICER) as the cost per quality-adjusted life-year (QALY), the net monetary benefit (NMB), and the net health benefit (NHB), for those being offered versus not offered VPT assessment during their A\u0026amp;E visit. NMB and NHB will be assessed using a range of values (\u0026pound;15,000-\u0026pound;30,000 per quality-adjusted life year, or QALY) representing a health decision makers\u0026rsquo; willingness-to-pay (WTP) to obtain the distribution of net benefits at different levels of WTP.\u003c/p\u003e \u003cp\u003eFor the secondary (multi-sector) analysis, we report the cost per unit of effectiveness in natural units (e.g., cost per reconviction avoided). We will also present the results using the extended impact inventory framework and consider alternative methods of aggregation \u003csup\u003e66\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec45\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.11. Sensitivity Analysis\u003c/h2\u003e \u003cp\u003eParameter uncertainty will be assessed using a probabilistic sensitivity analysis (PSA), varying key parameters over a range of expected values, and running 1,000 Monte Carlo simulations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec46\" class=\"Section4\"\u003e \u003ch2\u003e8.7.3.12. Social Costs\u003c/h2\u003e \u003cp\u003eAdditional exploratory analyses will estimate the broader social costs associated with the VPTs. Social costs will focus specifically on organisation costs of supporting patients who have been referred to them by the VPTs. Cost data of organisations will be assessed using top-down gross costing. Questionnaires or interviews, depending in interviewee preference, will be conducted with members of organisations to assess (i) the total annual expenditure cost for third sector and statutory organisations in the given year \u003csup\u003e67\u003c/sup\u003e. This cost is the total yearly cost of running the service including costs of supporting patients, staff cost, and consumables. (ii) The total number of users supported each year by the third sector and statutory organisations. (iii) The percentage of those supported that were referred by the VPTs each year.\u003c/p\u003e \u003cp\u003eThe given year will be 2022 to 2023; this is due to the Swansea team being implemented in January 2022. Top-down gross costing allows data to be aggregated and will allow us to estimate the mean cost of a \u0026lsquo;typical patient\u0026rsquo; to a \u0026lsquo;typical organisation\u0026rsquo;. We will include the distribution for each organisation but will not attribute a specific cost to a specific organisation. Furthermore, we will ask organisations for standard cost data (e.g. cost per counselling session); this will be used as a comparison of our mean average. We will ask about waiting lists for support which will be used to inform our knowledge.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Strengths and Limitations","content":"\u003cp\u003eLinkage requires knowledge of the patients\u0026rsquo; identity, anonymised and encoded as ALFs, to be included in the data return. There are instances where individuals may prefer to remain anonymous. Analyses are therefore limited to those who can be identified in routine data and linked to the WDS and therefore EDDS and related datasets. It is feasible that some may attempt to conceal their identity, and this could correspond to greater vulnerability. Furthermore, our reliance on administrative data means there is no information on the context in which their exposure to violence arose. There maybe marked difference in the modifiable risks and vulnerabilities for patients sustaining injury in and around premises licensed for the sale and onsite consumption of alcohol, and those experienced criminal or sexual exploitation. Conversely, our primary hypothesis is that the additionality of the VPT intervention teams means ED is better able to work with patients and therefore address the variation of patients that attend.\u003c/p\u003e"},{"header":"Dissemination, Outputs and Anticipated Impact","content":"\u003cp\u003eA diffusion and dissemination plan will be co-produced with stakeholders and PPIE groups. In so doing, we will define audiences (policy, practitioner, academic, lay, etc.) that might find the project outcomes of interest. In so doing, we will determine appropriate modalities to communicate with each, including presentations, briefing documents, and media events, with content informed by audience need.\u003c/p\u003e \u003cdiv id=\"Sec49\" class=\"Section2\"\u003e \u003ch2\u003e10.1. PPI, Policy and Practitioner Focused\u003c/h2\u003e \u003cp\u003eAn ongoing evaluation of VRUs and VPUs is to recommend that greater attention is paid to evidence-based interventions and that activities should consider the possible involvement of schools and therefore the Department for Education. Education data has been included and we can therefore contribute to the evidence based linking school activity to violence. In addition, our PPIE engagement highlighted the need to consider ARA predictors, and engagement with the Home Office will further identify intermediate outcomes. We therefore aim to undertake interim analyses that will inform the final analysis and respond to significant emerging policy questions as it relates to evidence-based violence prevention and reduction initiatives. While these will likely translate to academic papers, we also seek to produce more accessible outputs and exploit existing networks in that respect.\u003c/p\u003e \u003c/div\u003e"},{"header":" Abbreviations","content":"\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eADDE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eAnnual District Death Extract - All deaths registered in Wales. Including Welsh residents who died outside of Wales.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eALF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eAnonymised Linkage Field \u0026ndash; All individual records are assigned an ALF, allowing researchers to link different data sets together for a single person based on the ALF.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eARA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eAssault-related Attendance \u0026ndash; An attendance to an emergency department relating to the attendee being involved in an assault.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eBAWSO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eBlack Association of Women Step Out \u0026ndash; An All-Wales charity supporting Black and Minoritised individuals and communities.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eCAFCASS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eChildren and Family Court Advisory and Support Service \u0026ndash; A service that looks after the interests of CYP in family court proceedings. It is independent from Social Services and the courts.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eCEA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eCost-effectiveness Analysis \u0026ndash; Economic analysis using costs and outcomes of interventions.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eCG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eControl Group \u0026ndash; A group that is not receiving the intervention that is being researched. The control group may receive the standard intervention or no intervention.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eCYP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eChildren and Young People \u0026ndash; Representing those from birth, usually up until the age of 18.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eDASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eDomestic Abuse, Stalking and Honor Based Violence \u0026ndash; A checklist used by practitioners for high-risk cases of DASH.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eDBS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eDisclosure and Barring Service \u0026ndash; Part of the Home Office. It allows organisations to check the safety of those being recruited for work. Including criminal record checks and checks involving safety around CYP.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eDHCW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eDigital Health and Care Wales - An organisation that designs all-Wales digital health and care services. It provides digital services with the aim to improve people\u0026apos;s health.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eED\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eEmergency Department \u0026ndash; Area of the hospital for immediate and urgent care.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eEDDS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eEmergency Department Data Set \u0026ndash; National data set from emergency departments. It includes why the patient attended the ED and what treatment they received.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eEDVIPE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eEmergency Department Violence Intervention Programme Effectiveness and Cost-effectiveness Evaluation \u0026ndash; The acronym of the current study.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eFoI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eFreedom of Information \u0026ndash; Freedom to share or consume information from government funded public agencies. This information can be requested by members of the public through a FoI request.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eGLMM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eGeneralised Linear Mixed Models \u0026ndash; An extension to a generalised linear model that includes fixed and random effects.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eGP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eGeneral Practitioner \u0026ndash; GPs locally treat common medical conditions. They can refer to hospitals. \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eGRIPP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eGuidance for Reporting Involvement of Patients and the Public \u0026ndash; A checklist for the inclusion of person and patient involvement in research to improve transparency and quality.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eHRQoL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eHealth-Related Quality of Life \u0026ndash; A concept that examines health\u0026apos;s impact on quality of life.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eHVIP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eHospital-based Violence Intervention Programme \u0026ndash; A trauma-informed programme that ensures support for those who attend hospital with a violence-related injury.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eICD10\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eInternational Statistical Classification of Diseases - 10\u003csup\u003eth\u003c/sup\u003e revision.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eICER\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eIncremental Cost Effectiveness Ratio \u0026ndash; Summary of the economic value of an intervention.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eIDVA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eIndependent Domestic Violence Advocate - IDVAs work with those who have been affected by domestic violence through activities such as criminal justice support and representing the victim in legal proceedings.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eIGRP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eIndependent Information Governance Review Panel \u0026ndash; For the current study the IGRP was from SAIL. The IGRP includes representatives from various organisations and sectors to determine our access to SAIL databank.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eISTV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eInformation Sharing to Tackle Violence - An anonymised dataset that is collected by the NHS ED departments.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eISVA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eIndependent Sexual Violence Advisor \u0026ndash; ISVA works with those affected by sexual violence.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eITSA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eInterrupted Time Series Analysis \u0026ndash; Used when looking at outcomes. It involves looking at data before and after an interruption, i.e. an intervention. \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eLEH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eLocal Emergency Hospitals \u0026ndash; A hospital for immediate and urgent care. LEHs do not have a major trauma centre but are equipped to aid and transfer patients to major trauma centres.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eLSOA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eLower Layer Service Output Areas \u0026ndash; Geographical areas in England and Wales where data is collected, for example, population count and crime rate.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eMARAC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eMulti-Agency Risk Assessment Conference \u0026ndash; A meeting to discuss high risk domestic abuse cases. The meeting includes professionals such as, the police, health, IDVA\u0026rsquo;s, children\u0026apos;s services and third sector agencies.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eMARF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eMulti-Agency Referral Form \u0026ndash; A referral form used to report a concern about a child at risk.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eMIU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eMinor Injury Unit \u0026ndash; A walk in unit in a hospital for non-emergency care. They provide care and treat injuries such as rashes, cuts, sprains and burns. \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eMoJ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eMinistry of Justice \u0026ndash; A government department that is responsible for the criminal justice system, probation, courts, safeguarding and to examine and adapt the legal service.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eMoPI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eManagement of Police Information \u0026ndash; Management of police records and data to collect necessary and proportional data.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eMTC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eMajot Trauma Centre \u0026ndash; Unit in a hospital that provides support to patients with major trauma. Major trauma can be defined as an illness or incident that can cause permanent disability or death.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eNHB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eNet Health Benefit \u0026ndash; The benefit given to an intervention, that is worked out using the total expected costs of the intervention divided by the maximum cost effectiveness ratio. If the net health benefit of an intervention is positive the health of the population would be increased.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eNHS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eNational Health Service \u0026ndash; Health care that is funded by the UK Government.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eNMB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eNet Monetary Benefit \u0026ndash; Representing the value of an intervention in monetary terms, that is worked out by looking at the difference of monetary value of QALYs and the total estimated costs of the intervention.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eNPD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eNational Pupil Database \u0026ndash; Data set of all pupils in public schools in England including absence, exclusion and demographic information.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eNRAC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eNational Retention Assessment Criteria \u0026ndash; Scheduled reviews of information retention which poses several questions that assesses risk of harm posed by nominals. If the answer is \u0026lsquo;yes\u0026rsquo; to any of the criteria, the records must be retained and reviewed at a scheduled later date.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eOPCC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eOffice for the Police and Crime Commissioner - Police and crime commissioners are elected police officials that are responsible for their force areas, including the police budget and how the area is policed.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eOPRD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eOutpatient Referral Dataset - Data collected from the NHS including information on where the patient was referred to, the reason for referral and the service type requested.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003ePEDW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003ePatient Episode Database for Wales \u0026ndash; This dataset included all inpatient and day case activity occurring in NHS Wales institutions. It also includes data on Welsh residents that were treated in England.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003ePMS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003ePatient Management System - A digital computer system for health professionals to access patient records.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003ePNC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003ePolice National Computer \u0026ndash; A system used by the UK police and law enforcement to access information in real-time.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003ePND\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003ePolice National Database \u0026ndash; A national system to store an accumulation of policing information from all forces.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003ePPI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003ePublic and Patient Involvement - The inclusion of the public in a research project.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003ePSA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eProbabilistic Sensitivity Analysis \u0026ndash; a method used in economic evaluations that allows the researcher to assess the level of uncertainty or confidence in a decision.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eQALY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eQuality-Adjusted Life Year \u0026ndash; Years that are lived in perfect health. QALYs are used to assess the value of interventions.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eRALF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eResidential Anonymous Linkage Field \u0026ndash; An ALF that includes residential information such as location, number of residents and the change in occupants. RALFs can be used the assess the interaction between a person\u0026apos;s residential setting and their health. (see ALF).\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eRTF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eRural Trauma Facilities \u0026ndash; Hospitals in rural areas that are equipped to aid major trauma patients.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eSAIL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eSecure Anonymised Information Linkage \u0026ndash; SAIL is a data bank that allows researchers to access, link and assess patient and public health information.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eSMDS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eSubstance Misuse Data Set \u0026ndash; Data set on patients referred for a substance misuse problem. Also known as The Welsh National Database for Substance Misuse (see WNDSM).\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eSPIRIT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eStandard Protocol Items: Recommendation for Intervention Trial \u0026ndash; Recommendations for content for a clinical protocol.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eSRG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eStakeholder Reference Group \u0026ndash; A group made up of stakeholders that provide feedback and direction on the research project. SRGs are made up of stakeholders from different sectors.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eSSC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eStudy Steering Committee \u0026ndash; A group that provides oversight of the research activities. The members usually have lived or relevant experience of what is being researched.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eSUPER\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eService Users for Primary and Emergency care Research \u0026ndash; A group that includes Welsh residents that represent diverse backgrounds and experiences. They provide perspectives on research topics and give suggestions on how to conduct PPI.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eTIDier\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eTemplate for Intervention Description and Replication \u0026ndash; A checklist and guide for describing interventions that can be used for replication.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eTU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eTrauma Unit \u0026ndash; A unit in a hospital that aids those with major trauma. Trauma units are used when the patient is not stable enough to be moved to a major trauma centre.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eUHW\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eUniversity Hospital Wales \u0026ndash; Hospital situated in Cardiff, Wales.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eUK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eUnited Kingdom.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eVPT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eViolence Prevention Team \u0026ndash; The VPTs are a type of hospital-based violence intervention programme. They are currently situated in two South Wales emergency departments.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eVPU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eViolence Prevention Unit (Wales) \u0026ndash; A unit funded by the Home Office. \u0026nbsp;The VPU consists of a multi-disciplinary team that looks at evidence and research to reduce violence in Wales.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eVRU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eViolence Reduction Unit (England and Scotland) - A unit funded by the Home Office. The VRUs consist of multi-disciplinary teams that look at evidence and research to reduce violence in England and Scotland.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWAST\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eWelsh Ambulance Service Trust \u0026ndash; An NHS trust providing an ambulance service for Wales.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWDSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eWelsh Demographic Service Database - WDSD includes all individuals registered with a Welsh General Practitioner and via anonymisation it identifies household groups.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWIMD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eWelsh Index of Multiple Deprivation - The Welsh Government\u0026rsquo;s official measure of relative deprivation for small areas in Wales.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWLGP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eWelsh Longitudinal General Practitioner Dataset \u0026ndash; A data set that contains clinical information from General Practitioners in Wales, including diagnoses and referrals to tertiary care.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWNDSM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eThe Welsh National Database for Substance Misuse - Data set on patients referred for a substance misuse problem. Also known as Substance Misuse dataset (See SMDS).\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWSMA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eWhole System Multi-Agency \u0026ndash; An approach that includes people from multiple different agencies, for example, health care, police and the government.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eWTP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eWillingness To Pay \u0026ndash; The maximum price that someone is willing to pay for a product. In health economics it can be used as how much people are willing to pay to improve health and reduce risk.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 12.605%;\"\u003e\n \u003cp\u003eYEF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 87.395%;\"\u003e\n \u003cp\u003eYouth Endowment Fund \u0026ndash; The YEF fund research happening across England and Wales that is focused on preventing CYP becoming involved in violence.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003eData Availability\u003c/p\u003e\n\n\u003cp\u003eThe routine data used in these studies are housed in the SAIL databank and are available through application. Code written to manage and analyse data is available on request.\u003c/p\u003e\n\n\u003cp\u003eFunding\u003c/p\u003e\n\n\u003cp\u003eFunded by the National Institute for Health Research, Public Health Research Board (NIHR134055).\u003c/p\u003e\n\n\u003cp\u003eRole of the Funder\u003c/p\u003e\n\n\u003cp\u003eThe funder had no role in the design of the study, and will not have any role during its execution, analyses, interpretation of the data, or decision to submit results.\u003c/p\u003e\n\n\u003cp\u003eGovernance\u003c/p\u003e\n\n\u003cp\u003eA Study Steering Committee (SSC) was convened to oversee and advise on progress. The SSC included the Data Monitoring and Ethics functions and oversees any protocol amendments.\u003c/p\u003e\n\n\u003cp\u003eAcknowledgements\u003c/p\u003e\n\n\u003cp\u003eThe authors would like to acknowledge Nurses Vicky Lee and Sarah Wilcox, leads for the intervention sites, for their support of this evaluation. Furthermore, we would like to thank the Public, Patient Involvement participants, many of whom had not prior experience working with researchers. Their enthusiasm and insights greatly impacted on this evaluation and the broader research team.\u003c/p\u003e\n\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\n\u003cp\u003eSC Moore: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, guarantor, implemented the evaluation, analysed the data. S Brophy: drafted and revised the paper, implemented the evaluation. A Bandyopadhyay: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, analysed the data. A Newbury: designed data collection tools, drafted and revised the paper. T Lowe: drafted and revised the paper, implemented the evaluation, designed PPIE engagement, conducted and wrote-up PPIE engagement. D O\u0026apos;Reilly: drafted and revised the paper, implemented the evaluation. D Rawlinson: drafted and revised the paper, implemented the evaluation. L Snowdon: drafted and revised the paper, implemented the evaluation. J Shepherd: drafted and revised the paper, implemented the evaluation. V Sivarajasingam: drafted and revised the paper, implemented the evaluation. A Watkins: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, implemented the evaluation. S Walker: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper, implemented the evaluation. S Borgia: drafted and revised the paper, designed PPIE engagement. A Battaglia: designed PPIE engagement, conducted PPIE engagement. H Yeomans: designed and implemented PPIE engagement, conducted PPIE engagement. S Premji: designed data collection tools, wrote the statistical analysis plan, drafted and revised the paper. R Aslam: designed PPIE engagement. M Hamilton: designed data collection tools, drafted and revised the paper.\u003c/p\u003e\n\n\u003cp\u003eConflicts of Interest\u003c/p\u003e\n\n\u003cp\u003eNone\u003c/p\u003e\n\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\n\u003cp\u003eThe original manuscript was submitted for publication to BMJ Open on 8 March 2024. On 13 November 2024 the manuscript had the status \u0026ldquo;awaiting reviewer assignment.\u0026rdquo; The decision was therefore made to withdraw from BMJ Open and submit to BMC Public Health.\u003c/p\u003e\n\n\u003cp\u003eChecklists\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eTemplate for Intervention Description and Replication\u003c/strong\u003e (TIDieR) is available in Appendix 20.1.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eStandard Protocol Items: Recommendations for Interventional Trials\u003c/strong\u003e (SPIRIT) checklist is available in Appendix 20.2.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePurtle J, Dicker R, Cooper C, et al. Hospital-based violence intervention programs save lives and money. \u003cem\u003eJournal of Trauma and Acute Care Surgery\u003c/em\u003e 2013;75(2):331-33.\u003c/li\u003e\n\u003cli\u003eCunningham R, Knox L, Fein J, et al. Before and after the trauma bay: the prevention of violent injury among youth. \u003cem\u003eAnnals of Emergency Medicine\u003c/em\u003e 2009;53(4):490-500.\u003c/li\u003e\n\u003cli\u003eKaufman E, Rising K, Wiebe DJ, et al. Recurrent violent injury: magnitude, risk factors, and opportunities for intervention from a statewide analysis. \u003cem\u003eThe American Journal of Emergency Medicine\u003c/em\u003e 2016;34(9):1823-30.\u003c/li\u003e\n\u003cli\u003eKao AM, Schlosser KA, Arnold MR, et al. Trauma recidivism and mortality following violent injuries in young adults. \u003cem\u003eJournal of Surgical Research\u003c/em\u003e 2019;237:140-47.\u003c/li\u003e\n\u003cli\u003eNewton AS, Hartling L, Soleimani A, et al. A systematic review of management strategies for children\u0026rsquo;s mental health care in the emergency department. \u003cem\u003eEmergency Medicine Journal\u003c/em\u003e 2017;34(6):376-84.\u003c/li\u003e\n\u003cli\u003eHinsliff‐Smith K, McGarry J. Understanding management and support for domestic violence and abuse within emergency departments: A systematic literature review from 2000\u0026ndash;2015. \u003cem\u003eJournal of clinical nursing\u003c/em\u003e 2017;26(23-24):4013-27.\u003c/li\u003e\n\u003cli\u003eAnsari S, Boyle A. Emergency department-based interventions for women suffering domestic abuse: a critical literature review. \u003cem\u003eEuropean Journal of Emergency Medicine\u003c/em\u003e 2017;24(1):13-18.\u003c/li\u003e\n\u003cli\u003eBrice JM, Boyle AA. Are ED-based violence intervention programmes effective in reducing revictimisation and perpetration in victims of violence? A systematic review. \u003cem\u003eEmergency medicine journal\u003c/em\u003e 2020;37(8):489-95.\u003c/li\u003e\n\u003cli\u003eKoenig KL, Benjamin SB, Be\u0026yuml; CK, et al. Emergency Department Management of the Sexual Assault Victim in the COVID Era: A Model SAFET-I Guideline From San Diego County. \u003cem\u003eJournal of Emergency Medicine\u003c/em\u003e 2020\u003c/li\u003e\n\u003cli\u003eGoodall C, Jameson J, Lowe DJ. Navigator: A Tale of Two Cities. Glasgow: Violence Reduction Unit 2017.\u003c/li\u003e\n\u003cli\u003eSivarajasingam V, Guan B, Page N, et al. Violence in England and Wales in 2019. Cardiff, UK: Cardiff University 2020.\u003c/li\u003e\n\u003cli\u003eSivarajasingam V, Guan B, Page N, et al. Violence in England and Wales in 2020. Cardiff, UK: Cardiff University 2021.\u003c/li\u003e\n\u003cli\u003eIntroducing Public Health Measures (HO0345). London: The Home Office, 2019.\u003c/li\u003e\n\u003cli\u003eShepherd JP, Ali M, Hughes A, et al. Trends in urban violence:. \u003cem\u003eJournal of the Royal Society of Medicine\u003c/em\u003e 1993;86(2):87.\u003c/li\u003e\n\u003cli\u003eSutherland I, Sivarajasingam V, Shepherd JP. Recording of community violence by medical and police services. \u003cem\u003eInjury Prevention\u003c/em\u003e 2002;8(3):246-47.\u003c/li\u003e\n\u003cli\u003eGray BJ, Barton ER, Davies AR, et al. A shared data approach more accurately represents the rates and patterns of violence with injury assaults. \u003cem\u003eJ Epidemiol Community Health\u003c/em\u003e 2017;71(12):1218-24.\u003c/li\u003e\n\u003cli\u003eBath R. A whole-system multi-agency approach to serious violence prevention: A resource for local system leaders in England. London: Public Health England 2019.\u003c/li\u003e\n\u003cli\u003eSerious Violence Strategy. London: HM Government 2018.\u003c/li\u003e\n\u003cli\u003eViolence Reduction Unit Interim Guidance. London: Home Office 2020.\u003c/li\u003e\n\u003cli\u003eThe NHS Long Term Plan 2019 [Available from: www.longtermplan.nhs.uk accessed November 2020.\u003c/li\u003e\n\u003cli\u003eBetter Care for People With Co-occurring Mental Health and Alcohol/Drug Use Conditions. London: Public Health England 2017.\u003c/li\u003e\n\u003cli\u003eDuke AA, Smith KM, Oberleitner L, et al. Alcohol, drugs, and violence: A meta-meta-analysis. \u003cem\u003ePsychology of violence\u003c/em\u003e 2018;8(2):238.\u003c/li\u003e\n\u003cli\u003eBabor TF, Berglas S, Mendelson JH, et al. Alcohol, affect, and the disinhibition of verbal behavior. \u003cem\u003ePsychopharmacology\u003c/em\u003e 1983;80(1):53-60.\u003c/li\u003e\n\u003cli\u003eMalik NS, Munoz B, de Courcey C, et al. Violence-related knife injuries in a UK city; epidemiology and impact on secondary care resources. \u003cem\u003eEClinicalMedicine\u003c/em\u003e 2020:100296.\u003c/li\u003e\n\u003cli\u003ePallett J, Sutherland E, Glucksman E, et al. A cross-sectional study of knife injuries at a London major trauma centre. \u003cem\u003eThe Annals of The Royal College of Surgeons of England\u003c/em\u003e 2014;96(1):23-26.\u003c/li\u003e\n\u003cli\u003eBoyle A, Frith C, Edgcumbe D, et al. What factors are associated with repeated domestic assault in patients attending an emergency department? A cohort study. \u003cem\u003eEmergency medicine journal\u003c/em\u003e 2010;27(3):203-06.\u003c/li\u003e\n\u003cli\u003eDeWall CN, Anderson CA, Bushman BJ. The general aggression model. \u003cem\u003ePsychology of Violence\u003c/em\u003e 2011;1(3):245.\u003c/li\u003e\n\u003cli\u003eAllen JJ, Anderson CA. General aggression model. \u003cem\u003eIEME\u003c/em\u003e 2017:1-15.\u003c/li\u003e\n\u003cli\u003eFarrington DP. Early predictors of adolescent aggression and adult violence. \u003cem\u003eViolence and victims\u003c/em\u003e 1989;4(2):79-100.\u003c/li\u003e\n\u003cli\u003eHawkins JD, Herrenkohl T, Farrington DP, et al. A review of predictors of youth violence. 1998\u003c/li\u003e\n\u003cli\u003ePiquero AR, Jennings WG, Diamond B, et al. A systematic review of age, sex, ethnicity, and race as predictors of violent recidivism. \u003cem\u003eInternational journal of offender therapy and comparative criminology\u003c/em\u003e 2015;59(1):5-26.\u003c/li\u003e\n\u003cli\u003eWorld Report on Violence and Health. Geneva: World Health Organisation 2002.\u003c/li\u003e\n\u003cli\u003eLong SJ, Fone D, Gartner A, et al. Demographic and socioeconomic inequalities in the risk of emergency hospital admission for violence. \u003cem\u003eBMJ Open\u003c/em\u003e 2016;6(8)\u003c/li\u003e\n\u003cli\u003eOstrowsky MK. The social psychology of alcohol use and violent behavior among sports spectators. \u003cem\u003eAggression and violent behavior\u003c/em\u003e 2014;19(4):303-10.\u003c/li\u003e\n\u003cli\u003eFazel S, Gulati G, Linsell L, et al. Schizophrenia and violence: systematic review and meta-analysis. \u003cem\u003ePLoS Med\u003c/em\u003e 2009;6(8):e1000120.\u003c/li\u003e\n\u003cli\u003eFu F, Nowak MA, Christakis NA, et al. The evolution of homophily. \u003cem\u003eScientific reports\u003c/em\u003e 2012;2(1):1-6.\u003c/li\u003e\n\u003cli\u003eVan Godwin J, Moore G, Hamilton M, et al. Implementation and Process Evaluation of South Wales Hospital Based Violence Intervention Programmes. London: Youth Endowment Fund Forthcoming.\u003c/li\u003e\n\u003cli\u003eFlorence C, Shepherd J, Brennan I, et al. Effectiveness of anonymised information sharing and use in health service, police, and local government partnership for preventing violence related injury: experimental study and time series analysis. \u003cem\u003eBMJ\u003c/em\u003e 2011;342:d3313. doi: 10.1136/bmj.d3313\u003c/li\u003e\n\u003cli\u003eJones KH, Ford DV, Thompson S, et al. A profile of the Sail Databank on the UK secure research platform. \u003cem\u003eInternational journal of population data science\u003c/em\u003e 2019;4(2)\u003c/li\u003e\n\u003cli\u003eIrving A, Buykx P, Amos Y, et al. The acceptability of alcohol intoxication management services to users: a mixed methods study. \u003cem\u003eDrug and Alcohol Review\u003c/em\u003e 2020;39(1):36-43.\u003c/li\u003e\n\u003cli\u003eStaniszewska S, Brett J, Simera I, et al. GRIPP2 reporting checklists: tools to improve reporting of patient and public involvement in research. \u003cem\u003ebmj\u003c/em\u003e 2017;358\u003c/li\u003e\n\u003cli\u003eFlorence C, Shepherd J, Brennan I, et al. Effectiveness of anonymised information sharing and use in health service, police, and local government partnership for preventing violence related injury: experimental study and time series analysis. \u003cem\u003eBmj\u003c/em\u003e 2011;342\u003c/li\u003e\n\u003cli\u003eLinden A. A matching framework to improve causal inference in interrupted time‐series analysis. \u003cem\u003eJournal of Evaluation in Clinical Practice\u003c/em\u003e 2018;24(2):408-15.\u003c/li\u003e\n\u003cli\u003eLinden A, Yarnold PR. Using machine learning to evaluate treatment effects in multiple‐group interrupted time series analysis. \u003cem\u003eJournal of Evaluation in Clinical Practice\u003c/em\u003e 2018;24(4):740-44.\u003c/li\u003e\n\u003cli\u003eEvaluation of Violence Reduction Units 2020/21. London: Home Office, forthcoming.\u003c/li\u003e\n\u003cli\u003eFeiveson AH. Power by simulation. \u003cem\u003eThe Stata Journal\u003c/em\u003e 2002;2(2):107-24.\u003c/li\u003e\n\u003cli\u003eHennessy S, Bilker WB, Berlin JA, et al. Factors influencing the optimal control-to-case ratio in matched case-control studies. \u003cem\u003eAmerican Journal of Epidemiology\u003c/em\u003e 1999;149(2):195-97.\u003c/li\u003e\n\u003cli\u003eEmergency Department Follow-up Clinics. London, 2015.\u003c/li\u003e\n\u003cli\u003eLaw CG, Brookmeyer R. Effects of mid‐point imputation on the analysis of doubly censored data. \u003cem\u003eStatistics in medicine\u003c/em\u003e 1992;11(12):1569-78.\u003c/li\u003e\n\u003cli\u003eR\u0026uuml;cker G, Messerer D. Remission duration: an example of interval‐censored observations. \u003cem\u003eStatistics in Medicine\u003c/em\u003e 1988;7(11):1139-45.\u003c/li\u003e\n\u003cli\u003eTurnbull BW. The empirical distribution function with arbitrarily grouped, censored and truncated data. \u003cem\u003eJournal of the Royal Statistical Society: Series B (Methodological)\u003c/em\u003e 1976;38(3):290-95.\u003c/li\u003e\n\u003cli\u003eGuo Z, Gill TM, Allore HG. Modeling repeated time-to-event health conditions with discontinuous risk intervals. \u003cem\u003eMethods of information in medicine\u003c/em\u003e 2008;47(02):107-16.\u003c/li\u003e\n\u003cli\u003eBalen TA, Putter H. A tutorial on frailty models. \u003cem\u003eStatistical Methods in Medical Research\u003c/em\u003e 2020;29(11):3424-54.\u003c/li\u003e\n\u003cli\u003eGould W. Why can\u0026rsquo;t a subject enter and die at the same time in the Cox model? Texas, U.S.: StataCorp; 2023 [Available from: www.stata.com/support/faqs/statistics/time-and-cox-model/ accessed June 2023.\u003c/li\u003e\n\u003cli\u003eKing G, Nielsen R. Why Propensity Scores Should Not Be Used for Matching. \u003cem\u003ePolitical Analysis\u003c/em\u003e 2019;27(4):435-54. doi: 10.1017/pan.2019.11\u003c/li\u003e\n\u003cli\u003eKing G, Nielsen R, Coberley C, et al. Comparative effectiveness of matching methods for causal inference. \u003cem\u003eUnpublished manuscript, Institute for Quantitative Social Science, Harvard University, Cambridge, MA\u003c/em\u003e 2011\u003c/li\u003e\n\u003cli\u003eStuart EA. Matching methods for causal inference: A review and a look forward. \u003cem\u003eStat Sci\u003c/em\u003e 2010;25(1):1-21. doi: 10.1214/09-STS313 [published Online First: 2010/09/28]\u003c/li\u003e\n\u003cli\u003eImbens GW. Matching methods in practice: Three examples. \u003cem\u003eJournal of Human Resources\u003c/em\u003e 2015;50(2):373-419.\u003c/li\u003e\n\u003cli\u003eKing G, Lucas C, Nielsen RA. Matching Frontier 2017 [Available from: https://projects.iq.harvard.edu/frontier.\u003c/li\u003e\n\u003cli\u003eKing G, Lucas C, Nielsen RA. The Balance-Sample Size Frontier in Matching Methods for Causal Inference. \u003cem\u003eAm J Polit Sci\u003c/em\u003e 2017;61(2):473-89. doi: 10.1111/ajps.12272\u003c/li\u003e\n\u003cli\u003eWing C, Simon K, Bello-Gomez RA. Designing difference in difference studies: best practices for public health policy research. \u003cem\u003eAnnual review of public health\u003c/em\u003e 2018;39(1):453-69.\u003c/li\u003e\n\u003cli\u003eJones AM. Panel data methods and applications to health economics. \u003cem\u003eTC Mills and K Petterson, Palgrave Handbook of Econometrics\u003c/em\u003e 2007;2\u003c/li\u003e\n\u003cli\u003eJones AM, Lomas J, Rice N. Healthcare cost regressions: going beyond the mean to estimate the full distribution. \u003cem\u003eHealth economics\u003c/em\u003e 2015;24(9):1192-212.\u003c/li\u003e\n\u003cli\u003eTreasury HMs. The green book: Central government guidance on appraisal and evaluation. \u003cem\u003eLondon: HM Treasury\u003c/em\u003e 2018\u003c/li\u003e\n\u003cli\u003eDrummond MF, Sculpher MJ, Claxton K, et al. Methods for the economic evaluation of health care programmes: Oxford university press 2015.\u003c/li\u003e\n\u003cli\u003eWalker S, Griffin S, Asaria M, et al. Striving for a societal perspective: a framework for economic evaluations when costs and effects fall on multiple sectors and decision makers. \u003cem\u003eApplied health economics and health policy\u003c/em\u003e 2019;17:577-90.\u003c/li\u003e\n\u003cli\u003e\u0026Scaron;pac\u0026iacute;rov\u0026aacute; Z, Epstein D, Garc\u0026iacute;a-Moch\u0026oacute;n L, et al. A general framework for classifying costing methods for economic evaluation of health care. \u003cem\u003eThe European Journal of Health Economics\u003c/em\u003e 2020;21(4):529-42.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Emergency Department, Assault, Violence, Injury, Prevention, Treatment","lastPublishedDoi":"10.21203/rs.3.rs-5452363/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5452363/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIntroduction\u003c/p\u003e\n\u003cp\u003eHospital-based violence intervention programmes (HVIPs), based in Emergency Departments (EDs) have been proposed as a public health response to violence. These programmes address the underlying reasons why patients are exposed to violence. In addressing any underlying modifiable risks and vulnerabilities HVIPs can reduce patients’ exposure to violence and therefore subsequent unplanned attendance into ED.\u003c/p\u003e\n\u003cp\u003eMethods and Analysis\u003c/p\u003e\n\u003cp\u003eED patients are eligible for inclusion in the evaluation if they are normally resident in Wales, United Kingdom (UK), aged 11 years and older. A controlled longitudinal natural experiment will be undertaken. The primary outcome is derived from the Emergency Department Dataset, routinely collected for all EDs in Wales, and is subsequent unplanned ED attendance. Case patients will be matched to control patients attending EDs without an HVIP. Analysis will derive the hazard rate for subsequent unplanned ED attendances using recurrent event analysis. The total monthly count of patients identified as attending because of violence in intervention EDs will be compared to the total count of Welsh control EDs in an interrupted time series analysis to determine whether HVIPS increase violence ascertainment. To determine whether referral, versus no referral, to the HVIP represents value for money, we will undertake a cost-effectiveness analysis from the perspective of the National Health Service.\u003c/p\u003e\n\u003cp\u003eEthics and Dissemination\u003c/p\u003e\n\u003cp\u003eThe approval to access and analyse data housed in the Secure Anonymised Information Linkage (SAIL) databank, an ISO 27001 certified and UK Statistics Authority accredited secure data environment, was granted by the SAIL independent Information Governance Review Panel (Ref: 1421). Findings will be presented at local, national, and international conferences and disseminated by peer-review publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eISRCTN Registration\u003c/strong\u003e: 41868 (12 August 2022)\u003c/p\u003e","manuscriptTitle":"Protocol: A Quasi-Experimental Effectiveness and Cost-Effectiveness Evaluation of Emergency Department Violence Intervention Programmes in the United Kingdom","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-25 01:04:24","doi":"10.21203/rs.3.rs-5452363/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"14287dc0-7252-40d5-991c-96b37b813d66","owner":[],"postedDate":"December 25th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-09-25T08:24:05+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-25 01:04:24","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5452363","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5452363","identity":"rs-5452363","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}