Violence, Health and Society: VISION

Violence causes harms to health. The harms to mental health can be more long-lasting than the immediate harms to physical health and have consequences that reverberate through a person's life impacting on their functioning in society. Reducing such 'upstream' determinants of poor mental health would significantly improve the health of the population. This would reduce health inequalities since being a victim of violence is more prevalent among those who are already disadvantaged.
The Consortium would investigate the effectiveness of interventions to reduce violence and, thus, reduce health inequalities. Within the field of violence, we have special interest in domestic and sexual violence because these are significant causes of inequalities in mental health, which have been relatively neglected in the scientific and statistical evidence base. We address how to mainstream these issues across multiple sectors rather than seeing them as only of specialised concern.
Multiple institutions are relevant to preventing violence. They include not only health services, but also criminal law enforcement (most violence is a crime), civil law (e.g. domestic protection orders), specialised services (Third Sector organisations that help victim/survivors of violence), and governmental bodies concerned with law, policy and data quality. The connections between violence and ill health are complicated since they are mediated by many of these institutions. Identifying these connections would aid the development of more effective interventions while a complex systems analysis captures the adaptive behaviour between these systems.
The data needed to assess the effectiveness and cost-effectiveness of interventions is currently weak. This is partly because each specialised academic discipline and profession has a different way of measuring violence, which makes cooperation across these differences difficult. Not only do we need harmonised core metrics for the evaluation of interventions and cross-sector cost-benefit comparisons, we also need to adapt and extend our metrics to capture newly identified forms of abuse such as that facilitated by technology. The Consortium aims to improve the measurement framework and data availability to aid the evaluation of interventions. This is premised on cooperation between academics and practitioners. The project seeks to identify profiles of persons and incidents exposed to violence and link data from multiple services and surveys. We would assist services to make their own data more useable and more available. This involves care and attention to issues of data protection and the development of bespoke agreements on data sharing that respect communities that generate data.
We would unlock the potential in multiple data sources rather than collect new data. These datasets include major national surveys such as the Adult Psychiatric Morbidity Survey, and the Crime Survey for England and Wales, and also administrative data sets from professions and practitioners, including the police, solicitors, health and specialised services. These datasets will be linked in a new integrated dataset and provide an evidence base upon which a cost-benefit framework and risk assessment tools can be developed.
With the linked data and new tools, we would assess key interventions. These are interventions at the level of institutions and systems. Our focus is the prevention of violence in the population rather than the treatment of trauma in individuals. The Consortium seeks to mainstream evidence of the significance of violence for health in policy making. We would engage with decision-makers concerned with the commissioning of services and policy makers concerned with priorities for public expenditure, as well as wider publics.
The aim is to reduce the harm to health, especially mental health, by identifying the most effective and cost-effective interventions to reduce violence in the population.

The vision underlying the Consortium on 'Violence, Health and Society' is that improving the knowledge base on violence and using this knowledge to inform changes to policy and practice will improve population health and reduce health inequalities. It will do so by acting on 'upstream' harms to health caused by violence. Research on such interventions is developing but is held back by weak theory and weak data. Theory is weak because it is often focused on individuals rather than on the system level. We will develop systems analysis using complexity theory that allows consideration of feedback loops that generate wicked problems and perverse outcomes. We will embed questions about the significance of multiple intersecting inequalities including gender and ethnicity into the theory underpinning the systems framework. Data is weak because, collected by multiple agencies for their own purposes, it is fragmented and incommensurable. We will work with data providers to develop survey and administrative data in health, justice and specialised services and translate it into our shared measurement framework. We will curate existing datasets rather than collect new data. We will use natural language processing to turn free-text narratives into quantitative data. We will integrate data using probabilistic individual profiles, which offers a powerful new route to data linkage that avoids the dangers of identifying real people. We will interrogate our newly improved data with questions about the nature of the causal pathways connecting violence, health and society to identify promising sites of intervention. We will develop cost-benefit analysis and evaluate interventions, using findings to build the theory of change. With our partners in health, justice, specialised services and government, we will seek to embed the new measurement in practice, to enable evidence-based feedback on developments to reduce violence and thereby improve health and reduce health inequalities.