Mental illness, mental health service use, and lifetime domestic violence perpetration: an observational study.

King's College London (KCL) wishes to retain the data to examine the association between mental health and lifetime perpetration of domestic violence. KCL researchers aim to identify whether there is a significant relationship between mental health and perpetration of domestic violence and, if such a relationship exists, whether or not this can be explained by prior trauma; use of alcohol and/or drugs, and/or socioeconomic position. Given that much violence in the general population is gendered, KCL researchers wish to also assess whether there is a link between mental health and perpetration/victimisation in men and women. It is for this reason that KCL are requesting to continue to access the 2014 Adult Psychiatric Morbidity Survey (APMS) data set.

Continued access to the pseudonymised 2014 APMS data set will support an ongoing body of work carried out by KCL’s Section for Women's Mental Health. The Section was set up to investigate the mental health needs of women, in response to the needs for the NHS, PHE, and the broader scientific community, with the aim of improving the mental health of women while further understanding and addressing violence and abuse. Prior to this request several meetings with mental health service users have been held to shape the purpose of the research. In these meetings it was identified that there is insufficient support for service users who were perpetrators of domestic violence.

It is intended that the findings from this analysis will inform a programme of work on the response of mental health services to the perpetration of domestic violence enabling an assessment of how best to identify mental health service users presenting the greatest risk of domestic violence and determining the treatments that might be most effective in reducing the risk of domestic violence.

The lawful basis for processing in this instance is GDPR Article 6(1)(e) ‘public task’ whereby processing is necessary to perform a task in the public interest. The processing of special category data can be justified under GDPR Article 9(2)(j), where processing is necessary for archiving, research and statistics (with a basis in law). This research is in the public interest because the processing of this data may yield benefits for the users of mental health services, carers/family members of people with mental illness, mental health professionals, and wider society. In addition to this processing may allow the identification of specific groups at relatively greater risk of domestic violence perpetration and/or victimisation.

There is currently limited epidemiological evidence on the perpetration of domestic violence, and improved knowledge in this area will enhance wider societal response to perpetration of domestic violence, as part of a future national perpetrator strategy. Globally, one in three women report domestic violence in their lifetime and improving evidence on perpetrators could result in added improvements in this area, leading to violence reduction.

All individuals who participated in the APMS were consenting adults living in England and Wales and, as such, the processing of their data for the purposes described here should not come as unexpected. As APMS is a nationally representative survey it provides an opportunity to generate the first nationally representative data on perpetration of domestic violence in relation to mental illness, as such KCL wish to receive national APMS data.

KCL is the sole data controller for the purpose of processing outlined above. KCL will be the only organisation processing the data for this purpose.

All organisations party to this agreement must comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by “Personnel” (as defined within the Data Sharing Framework Contract i.e.: employees, agents and contractors of the Data Recipient who may have access to that data).

The 2014 APMS data set is held on behalf of NHS England by the UK Data Service (UKDS) (www.ukdataservice.ac.uk) and the UKDS is responsible for dissemination under direction by NHS England. King's College London will receive the pseudonymised APMS data set. There is no facility to select individual variables. KCL will be able to download the data set from UKDS for the period specified within the Data Sharing Agreement and must securely destroy all local copies of the data set when the Agreement expires and notify NHS England in line with standard procedures. This 2014 version of the data set available has been redacted on Disclosure Control Procedure advice to minimise the likelihood of individuals being able to identify anyone taking part in the survey.

UKDS have already under the previous iteration of this agreement, transferred the pseudonymised 2014 APMS data set to KCL. No other organisations will be involved in the flow of data.

It is intended that the APMS data received from NHS England will be described in statistical software, and analysed for association between exposure and outcome, as well as other variables. The statistical analyses performed should allow researchers to determine 1) whether there is a significant relationship between domestic violence perpetration and mental health, 2) whether this relationship can be explained by other factors (e.g. prior trauma, use of alcohol and/or drugs, socioeconomic position), and 3) if this relationship varies across different genders.

Data will be inspected to look for errors and cleaned as necessary. Correlations will be examined between the survey items (e.g. , survey weighting, age and sociodemographic variables, violence-related items, and mental health items). Researchers will assess the overlap between mental health and domestic violence-related variables, for errors and pre-analysis processing. Basic descriptions of bivariate distributions with victimisation will be tested, before and after statistical modelling (to account for other explanations of a link between domestic violence and produced, to identify basic data patterns. Analyses will then assess association between victimisation and a range of different mental health indicators).

Results will be displayed in tables with small numbers suppressed and reports -related characteristics, including depression, anxiety, and psychotic experiences. Adjustments will include socioeconomic and demographic characteristics.

The analysis will examine the overlap between domestic violence and factors. This may involve examining associations of lifetime perpetration of domestic violence with age, gender, and educational attainment, and mental health indicators including common mental disorder. Inclusion of covariates such as age, gender, educational attainment, and alcohol use in the assessment of association between mental health (including depression and anxiety).

This analysis plan respects the essence of the right to data protection, and clear and specific measures will be implemented to safeguard the rights and interests of the data subject.

Under this agreement data received from NHS England will not be linked and there will be no attempt to re-identify individuals who completed the APMS survey.

KCL will store the data on servers behind a firewall which are inaccessible from the outside world, except via a Virtual Private Network (VPN), which provides a password-protected, encrypted, private communication channel for staff working outside the office. All analyses will therefore be done on KCL equipment behind secure firewalls.

Data previously provided under this agreement will only be processed by substantive employees and students of King’s College London, all of whom have been appropriately trained in data protection and confidentiality.

The KCL Privacy notice can be found online here: https://www.kcl.ac.uk/terms/privacy .

The data processing described above is expected to result in a project report and scientific papers submitted for peer review. Presentations are expected to be produced for the KCL Section of Women’s Mental Health, and the UKRI Violence, Abuse, and Mental Health Network.

Findings are expected to be disseminated to researchers via epidemiological networks including the Society for Epidemiological Research and the International Federation of Psychiatric Epidemiology. Further to this, findings will be presented via workshops to networks of researchers with a focus on violence and abuse, including the UKRI Violence, Abuse and Mental Health Network.

Existing infrastructure will be used to communicate findings to individuals who took part in the collection of the APMS dataset (APMS individuals). Findings are expected to be shared with the Economic Social Research Council (ESRC) Centre for Society and Mental Health to reach a wide range of academic and policy-making communities. Scientific findings are expected to be primarily communicated in psychiatric journals (BJPsych, Lancet Psychiatry) and violence focused journals (Trauma, Violence and Abuse, and the Journal of Interpersonal Violence).

Active public engagement with scientific messages is hoped to be encouraged over social media using tweet threads, blog posts, and press releases. It is expected where possible, given support arrangements at the time of publication, scientific reports will be published in open access journals to ensure the widest possible reach, for the purpose of the results reaching carer networks, victims/survivors and people with lived experience of mental illness.

To allow replication of scientific findings in other data, code is expected to be published alongside scientific reporting. Target dates for the final scientific report and related manuscripts and all other outputs is December 2024.

Communication of scientific findings will be done sensitively, and mindful of not contributing to mental health stigma. Children's data will not be processed.

The data will not be made available to any third parties other than those specified except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

Violence is a detriment to population health and places a large burden on the UK health and social care services. This research may allow health care professionals to better identify, assess and respond to domestic violence in the context of mental illness, and to better understand the possible role of mental health treatments.

It is hoped that the findings yielded from this study will inform national efforts towards a perpetrator strategy (linking different sectors but including mental health services) and will help to identify directions for more detailed scientific study of existing responses. As many as 40% of general population have experienced violence and abuse and, based on existing evidence of association of perpetration of domestic violence with mental disorder, mental health services may play an important role in a national response to perpetrators- particularly if psychiatric or behavioural interventions play a role in shaping occurrence of violence in this group.

Evidence produced because of data processing may be in a position to shape national guidance on commissioning, and clinical management guidance on domestic violence (for example, in the context of alcohol/drug misuse).

Indicators of benefit may be citation in local policy guidance on the management of domestic violence, citation in national reports on domestic violence perpetration, scientific citation. The incorporation of this evidence into guidelines could be expected to improve knowledge on the characteristics of perpetrators of domestic violence (DV), and the quality-of-service delivery, through better identification of risks.

It is expected that these benefits may be achieved in the five years following the publication of the scientific report detailed in section 5c.

We have published the first report in the British Journal of Psychiatry Open and presented these results at academic conferences.

Update under version 1.9

We published a scientific paper on the prevalence of perpetration of intimate partner violence in users of secondary mental health services in England, in the British Journal of Psychiatry Open in May 2023. We disseminated the findings to an international psychosis conference in October 2023.