Published: 10 January 2023
Version: 1.0 January 2023Print this document
Timetable and Budget
|Deadline for Stage 1 applications||14 February 2023, 1pm|
|Notification of outcome of Stage 1 application||April 2023|
|Deadline for Stage 2 application||06 June 2023, 1pm|
|Notification of outcome of Stage 2 application||October - November 2023|
|Project start||January 2024|
|Project duration||12 - 18 months|
|Budget||£350,000 per project|
The National Institute for Health and Care Research (NIHR) Policy Research Programme (PRP) invites applications for a single research project to identify and evaluate interventions that improve both mental health and violence outcomes.
The relationship between violence and mental health is complex. We know that some mental health conditions have been associated with higher risks of committing violence. However, we also know that there are many shared factors, associated with both developing mental health conditions and committing violence, such as adverse childhood experiences. Much research has been conducted into trying to understand the underlying causes of these issues.
This research project seeks to take a ‘what works’ approach to finding interventions that may both reduce violence and improve mental health outcomes.
The prevention of risks to others is central to practitioners working in mental health services, most notably those practising under the Mental Health Act. Nevertheless, research into mental health and violence has mostly focused on understanding the link and likelihood of these events and identifying the characteristics of those patients who have committed violence. There is a gap in our understanding of what works in the community to reduce the risks of violence posed by people with mental health conditions.
Other research into interventions to prevent violence has focused on identifying and addressing specific symptoms or prescribing certain types of medication. Studies also often focus on behaviour in inpatient settings.
For the purposes of this research, however, we expect a focus on the wider context of community services (including community mental health services, crisis teams, supported housing), and holistic approaches across all MH conditions (for example, not only considering medication but also talking therapies, family interventions, support groups or approaches taken from multi-systemic therapy.) We are taking this system-wide approach, covering multiple conditions, care and interventions because of the well-evidenced limitations of seeking to identify and target specific individuals to prevent extremely low-likelihood events.
The two broad areas we are interested in are:
- Interventions for those people with mental health conditions who pose a risk of perpetrating violence, but who have not been subject to the criminal justice system (i.e. people who have not been convicted of crimes or arrested). This would assess effectiveness of interventions in terms of improving mental health outcomes and reducing violence.
- Interventions for people with mental health conditions with a history of previous violence who have been subject to the criminal justice system (for example, people who have been convicted of a crime or who are subject to criminal proceedings), looking at co-benefits on MH and violence.
We expect the research to assess the effectiveness of existing interventions in improving mental health outcomes and reducing violence, rather than designing de novo interventions. This approach will enable us to ensure that we can use the outcomes of this research to inform practice more quickly by replicating and scaling up existing practices that demonstrate, or show promise of demonstrating, effectiveness in improving violence and mental health outcomes for these patients.
We propose splitting the research into three stages:
- Initial scoping review
- An initial scoping review into previous literature evaluating interventions that improve both mental health and violence outcomes.
- Identifying the measures for the two outcomes being studied – violence and mental health outcomes.
- Finding interventions already taking place, focusing on areas delivering interventions that can be compared with others. For example, NHS services working closely with the voluntary sector in a defined local area.
- Asking areas conducting these interventions to share their methodologies and any existing evidence on the effectiveness of their interventions.
- Evaluating those interventions (depending on the outcomes of the first two stages)
- Selecting the most useful interventions for evaluating further.
- Provide information on which existing interventions appear most useful and feasible for rolling out more widely, for example via adapting them for different settings or client groups.
New Guidance on Health Inequalities data collection within NIHR PRP Research:
Health Inequalities is a high priority area within the Department of Health and Social Care and the NIHR and is often present in a majority of funded projects. We are now assessing all NIHR research proposals in relation to health inequalities. We are asking applicants to identify in their application whether or not there is a health inequalities component or theme and how this research hopes to impact health inequalities. We are also asking researchers to collect relevant data, if appropriate for the research. Our goal is to collect information on health inequalities in research and data relating to the main outcome(s) of the proposed research. Please clearly identify in the research plan section whether or not your application has a health inequalities component or relevance to health inequalities and detail the core set of health inequalities breakdowns that will be reported; if none please explain why. We understand that research projects have different methodologies and focus on different populations, so please explain what data will be collected and reported for the methodology you plan to use. If a health inequalities component is not included, please explain why this does not fit within your proposed research. This should only be a few sentences.
For quantitative research we would ideally like researchers to provide one-way breakdowns of their main outcome(s) by the following equity-relevant variables: age, sex, gender, disability, region, 5 ONS Ethnic groups, and the 5 IMD quintile groups. If more detailed cross tabulations are appropriate, please include these. This table should be submitted to NIHR PRP at the end of the project. Due to data limitations, judgement calls may be necessary about which breakdowns to report and whether to merge categories to increase counts in particular cells; we ask you to make these judgement calls yourself, bearing in mind our data curation aim of enabling future evidence synthesis work in pooling results from different studies. More details and an example table can be found in Appendix A.
For qualitative research projects, this can be purely descriptive statistics giving the number of observations against the various variables.
Further details about this new request can be found in Appendix A.
A recording of the Health Inequalities in NIHR PRP Research Q&A Event which was held on 19 September 2022 is available to view on Youtube.
Areas out of scope for this programme of work
Interventions to prevent violence that focus on identifying and addressing specific symptoms thought to predict violence
Prescription of medications thought to reduce violence.
Studies exclusively focusing on behaviour in inpatient settings.
Technical requirements / Expertise required
Ability to co-design research with relevant stakeholders including those with lived experience of mental health.
Familiarity with methods to calculate and model metrics like mental health and violence outcomes, including in exceptional circumstances (for example, in the context of extremely disadvantaged groups).
Subject matter expertise on offender health, serious mental illness and violence is essential including an in-depth knowledge of the complex needs and barriers faced by these individuals.
Experience with appropriate framing of research findings concerning offender health, serious mental illness and violence to mitigate further risk of stigmatisation and marginalisation.
Applicants are asked to consider the timing and nature of deliverables in their proposals. Policymakers will need research evidence to meet key policy decisions and timescales, so resource needs to be flexible to meet these needs. A meeting to discuss policy needs with DHSC officials will be convened as a matter of priority following contracting.
For the first two work streams, findings should be provided in the form of a final report. Applicants will also be expected to arrange a meeting with DHSC and NIHR officials upon completion of the first two stages, to enable policymakers to assess progress, discuss findings and crystallise plans for the final evaluation stage.
Budget and duration
Applications will be assessed on value for money. We would not expect the cost to be more than £350K. The budget and timeline for the third stage of the research will be discussed at the interim meeting between this stage and the first two, but we expect that the initial funding offered should be sufficient to cover a limited amount of evaluation.
The duration of the project should be as short as is consistent with delivering a high-quality study.
Costings can include up to 100% full economic costing (FEC) but should exclude output VAT. Applicants are advised that value for money is one of the key criteria that peer reviewers and commissioning panel members will assess applications against.
A research advisory group including, but not limited to, representatives of DHSC, other stakeholders including NHSE and the successful applicants for the research should be established. The advisory group will provide guidance, meeting regularly over the lifetime of the research. The successful applicants should be prepared to review research objectives with the advisory group, and to share emerging findings on an ongoing basis. They will be expected to:
- Provide regular feedback on progress
- Produce timely reports to the advisory group
- Produce a final report for sign off
Research contractors will be expected to work with nominated officials in DHSC, its partners and the NIHR. Key documents including, for example, research protocols, research instruments, reports and publications must be provided to DHSC in draft form allowing sufficient time for review.
- Chang Z, Lichtenstein P, Långström N, Larsson H, Fazel S. Association Between Prescription of Major Psychotropic Medications and Violent Reoffending After Prison Release. JAMA. 2016 Nov 1;316(17):1798.
- Bhavsar V, Kosidou K, Widman L, Orsini N, Hodsoll J, Dalman C, MacCabe JH. Clozapine Treatment and Offending: A Within-Subject Study of Patients With Psychotic Disorders in Sweden. Schizophr Bull [Internet]. 2020 Feb 26;46(2):303-310. Available from: Clozapine Treatment and Offending: A Within-Subject Study of Patients With Psychotic Disorders in Sweden - PMC (nih.gov)
- Coid JW, Kallis C, Doyle M, Shaw J, Ullrich S. Identifying Causal Risk Factors for Violence among Discharged Patients. Courvoisier DS, editor. PLOS ONE. 2015 Nov 10;10(11):e0142493.
- Kendler KS, Campbell J. Interventionist causal models in psychiatry: repositioning the mind–body problem. Psychological Medicine. 2008 Oct 10;39(6):881–7.
- Whiting D, Lichtenstein P, Fazel S. Violence and mental disorders: a structured review of associations by individual diagnoses, risk factors, and risk assessment. The Lancet Psychiatry. 2020 Oct;8(2).
- Whiting D, Gulati G, Geddes JR, Fazel S. Association of Schizophrenia Spectrum Disorders and Violence Perpetration in Adults and Adolescents from 15 Countries. JAMA Psychiatry. 2021 Dec 22;79(2).
- Fazel S, Wolf A, Palm C, Lichtenstein P. Violent crime, suicide, and premature mortality in patients with schizophrenia and related disorders: a 38-year total population study in Sweden. The Lancet Psychiatry [Internet]. 2014 Jun;1(1):44–54. Available from: Violent crime, suicide, and premature mortality in patients with schizophrenia and related disorders: a 38-year total population study in Sweden
- Fazel S. Schizophrenia, Substance Abuse, and Violent Crime. JAMA [Internet]. 2009 May 20;301(19):2016. Available from: Schizophrenia, Substance Abuse, and Violent Crime - PMC
- Khalifeh H, Johnson S, Howard LM, Borschmann R, Osborn D, Dean K, et al. Violent and non-violent crime against adults with severe mental illness. British Journal of Psychiatry. 2015 Apr;206(4):275–82.
- Trevillion K, Oram S, Feder G, Howard LM. Experiences of Domestic Violence and Mental Disorders: A Systematic Review and Meta-Analysis. van Os J, editor. PLoS ONE [Internet]. 2012 Dec 26;7(12):e51740. Available from: Trevillion, K., Oram, S., Feder, G. S., & Howard, L. M. (2012). Experiences of domestic violence and mental disorders
- Hughes K, Bellis MA, Hardcastle KA, Sethi D, Butchart A, Mikton C, et al. The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. The Lancet Public Health [Internet]. 2017 Aug;2(8):e356–66. Available from:The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis
Appendix A: Further Detail on the New Guidance on Health Inequalities data collection within NIHR PRP Research
Health Inequalities is a high priority area within the Department of Health and Social Care and the NIHR and is often present in a majority of funded projects. We are now assessing all NIHR research proposals in relation to health inequalities. We are asking applicants to identify in their application whether or not there is a health inequalities component or theme and how this research hopes to impact health inequalities. We are also asking researchers to collect relevant data related to health inequalities, if appropriate for the research. Collecting specific information about health inequalities in research submitted to the programme will allow for categorisation of health inequalities research, curation of data to aid future health inequalities research and enable policymakers to better understand the implications of health inequalities within their policy areas. This is a new request from the NIHR PRP and we will be continuing to monitor queries and adapt the process as needed. If you have any feedback on this new request, please contact us at firstname.lastname@example.org.
Our goal is to facilitate more widespread and consistent reporting of health inequality breakdown data relating to the primary outcomes of NIHR funded research. We would ideally like researchers to focus on the following equity-relevant variables: age, sex, gender, disability, region*, 5 ONS Ethnic groups**, and the 5 IMD quintile groups. These variables are considered an ideal, but we understand that these are subject to change depending on the sample population and specific research question.
For qualitative research projects, this can be purely baseline characteristics of the participants, for example, the number of participants in each ethnic group.
For quantitative research projects, if there are multiple outcomes/effects with your stakeholders, select a small number of main outcomes as appropriate to report equity breakdowns. We will not be prescriptive about the number of the outcomes, as it will depend on the number of study design types and the nature of the project aims. We are asking for one way cross tabulations of each primary outcome by these equity-relevant variables, if appropriate for your research, together with the number of observations in each cell. If more detailed cross tabulations are appropriate for your proposed research, please include these as well. This request applies to both primary data collection studies and secondary analysis of routine data, and to causal inference studies as well as descriptive studies; however, if this is not possible due to data limitations then please explain. Due to sample size and other data limitations there may be difficult scientific and/or data security*** judgement calls to make about which breakdowns to report and whether to merge categories to increase counts in particular cells; we ask you to make these judgments yourself, bearing in mind our data curation aim of enabling future evidence synthesis work in pooling results from different studies. We also ask that researchers report breakdowns for the unadjusted as well as adjusted outcomes/effects, as appropriate.
We understand that research projects may employ different methodologies, and focus on different populations. Please explain how the variables and data collection methods chosen are appropriate to the methodologies used.
We ask that you please clearly identify in the research plan section of the application whether your application has a health inequalities component or not and detail the core set of health inequality breakdown data that will be collected, if applicable. Submission of the data collection will be a condition of the final report for all research with relevant methodologies regardless of whether the research has a health inequalities component that will need to be submitted to NIHR PRP when the grant has finished. This should only take a few sentences within the research plan section.
* Table below uses the nine regions in England, further regions can be used if using the UK as the study population. Please report region breakdown for large samples in nationally representative descriptive studies. There is no need to report this for small sample studies, for sub-national studies, or for quasi-experimental studies where it would require time-consuming re-estimation.
** White, Mixed/ Multiple ethnic groups, Asian/ Asian British, Black/ African/ Caribbean/ Black British, Other ethnic group. If the sample size is small then it is fine to report only some of the requested equity breakdowns and to merge some of the sub-groups as appropriate.
*** For guidance on how to handle data security concerns in reporting of sensitive data please see ONS guidance.
Example data table for submission at the end of the funded research project
(N.B. If there is more than one main outcome then you will require more tables and if you adjust your outcome then you will need two tables for the adjusted outcome and unadjusted outcome. For other methodologies, variable vs number of observations may be more appropriate to record participant data). This table is for an example only. It does not contain sub variables and does not illustrate any preference for certain variables, as these will be dependent on the proposed research.
|Variable||Outcome (an appropriate average for this subgroup, usually the mean)||Number of observations||Additional information about variation if appropriate, e.g. range, standard deviation|