What made you first become interested in the question of violence?
There were three things that made me interested in the question of violence. First of all, I’m a maxillofacial surgeon, which means that I, and my colleagues in hospitals across the country, are treating many people who have been injured in violence.
Second, when training in Leeds in the 1980s, I began to see the patterns of violence from this injury perspective. I was startled when one of my colleagues said, “We see more people injured in violence when there’s a miners’ strike.” And I noticed that a few licensed premises were the site of a large proportion of violent injuries.
Because of this, when I was in Bristol looking for a PhD topic, violence from a health standpoint was very attractive. The third reason for my continued interest in violence are the discoveries in this PhD research. For instance, I was surprised to discover – and this finding has since been replicated many times – that half of violence that results in hospital treatment is never reported to the police. This has real implications for the importance of health data in tracking violence – there are so many holes in police data.
You have been involved in Cardiff’s pioneering work around violence reduction for more than 20 years. Tell us about that.
Cardiff’s violence prevention work has been led by two bodies which I founded in 1996 and then chaired for 20 years. The first, the Cardiff Violence Prevention Board, is composed of practitioners and executives who are in a position to implement interventions to prevent violence. This is highlighted in the Crime and Disorder Act as an example of good multi-agency partnership. It also provides a testbed for new ideas emerging from the second body, Cardiff University’s Violence Research Group: a cross-disciplinary alliance of researchers from across the University.
The Board developed a systematic approach to using anonymised data collected in hospital emergency departments on violence locations, weapons and assailants. These ‘Cardiff Model’ data, when combined with police data, provide a far more complete picture of violence than police data alone, enabling a more targeted response, for example deployment of police and CCTV cameras to violence hotspots. Published evaluations show that this approach brings about violence reductions of over 40% and substantial cost savings relative to cities where this approach is not implemented.
We also use these data to identify victims of domestic violence. As a result, hospitals can more frequently trigger a multi-agency domestic violence risk assessment and support strategy which help victims escape violent relationships.
The research group also uses data about violence to find new preventive measures. For example, we found that a lot of violence is spontaneous and triggered by the most trivial incident, like intoxicated people bumping into each other. This led the Board, through the local authority, to pedestrianise streets in the city’s entertainment district to decrease the risk of confrontation.
Another example is our work on glass. Our research showed that glasses and bottles had been used as weapons in ten per cent of the violent incidents which led to hospital attendance. Ten per cent may not sound like a lot, but across the UK that’s tens of thousands of violent acts. After lab testing of pint glasses we did a randomised trial of toughened glass in bars across West Midlands and South Wales and found that it was associated with significantly fewer glass breakages and injuries.
This kind of quantitative evaluation is a core part of our work. We had some evidence that suggests CCTV might be effective, but it was only when we did a controlled study comparing cities that had CCTV with cities that didn’t (back when some cities still didn’t have CCTV) that we concluded that it improved police awareness of violence and cut violence-related A&E attendance.
A lack of controlled studies like this is a real drawback in policing. Without real evidence from well-designed experiments, how can we know what works?
People talk about your work as a ‘public health approach’ to violence. But what does that mean?
The broad answer to that is that it is now well-recognised that the NHS has unique contributions to make in terms of violence prevention. We mustn’t leave violence prevention to the police and the criminal justice system alone. A public health approach means multi-agency working where agencies which have unique interventions to offer contribute to violence reduction, including the NHS with its distinctive and evidence-based approach. It’s not about finding a biological agent which causes violence. It’s about seeing, understanding and working to prevent the health harms that violence produces. That’s not just the physical injuries, it’s also the mental health harms: the trauma for victims, families and communities.
Early intervention is part of a public health approach. When health visitors work with young single mums in deprived areas for example, an approach known as nurse-family partnership, randomised trials show that violence, bullying, truancy, anti-social behaviour and drug taking can be reduced over many years as the children develop. Overall, a public health approach means three layers of violence prevention:
- Primary prevention: stopping violence from happening in the first place. This includes early intervention, the deterrent effect of the criminal justice system, and design solutions to reduce violence.
- Secondary prevention: stopping violence happening again once it’s happened the first time. This includes preventing re-victimisation through protecting people injured in domestic violence, and tackling risk factors like alcohol intoxication.
- Tertiary prevention: reducing the impact of violence once it’s happened. Examples of this are trauma centres in the United States which demonstrably save lives, and psychological treatment which nips post-traumatic stress disorder in the bud.
The ‘Cardiff Model’ incorporates these three strategies and is an example of an integrated public health approach.
What would you add to current debates on violence in the media?
A public health approach which measures violence reliably should reassure, and hopefully is reassuring, the public and professionals alike. Injury data across England and Wales tell us that violence levels in 2017 were unchanged from 2016, and that, as the authoritative Crime Survey also shows, levels of violence have fallen steadily and substantially since 2000. But concern now is rightly focused on knife and gun violence which, though numbers are much smaller than less serious violence, has recently risen in some cities, especially in London, according to hospital and police data.
Overall though, police data are a poor measure of violence. For example, a change in reporting rules means that tens of thousands of offences which would once have been categorised as public order are now classified as violent offences. And the definition of actual bodily harm has been changed to cover any violence which leaves a visible mark. This increases the number of offences recorded, but doesn’t, of course, make us a more violent society. Journalists who use these data to generate sensational media stories whip up fear unnecessarily and harmfully and should be ashamed of themselves.
Under Boris Johnson, and now Sadiq Khan, Cardiff Model data are collected and shared by all London’s 29 A&E departments, analysed and used – for example by the integrated gang unit in Hackney. Not surprisingly, gang members don’t report injuries sustained in violence or associated with drug trafficking to the police and will often use A&Es far away from the scene of assaults. The Cardiff Model, by focusing on injury and numbers of assailants, identifies such violence so that it can be targeted. More than one assailant is a good indicator of gang activity; this information is now being used by the Met and other police forces.
What’s the one lesson that other cities should learn from Cardiff’s approach to violence reduction?
The Cardiff Model is widely recognised now, including in the United States and Australia. Many cities have adopted this approach. But sharing data collected in emergency departments isn’t going to prevent anything if it’s not used. Astonishingly, this is the part of the Model which is most often neglected. Every city should have a violence prevention board, like Cardiff’s, where, based firmly on good data, the local authority, police and health come together to understand and prevent violence. Agencies working together in that way is the key to reducing violence.
Jonathan Shepherd is Professor of Oral and Maxillofacial Surgery at Cardiff University. His Violence Research Group won a 2009 Queen's Prize in Higher Education. He is Emeritus Director of the University's Crime and Security Research Institute. His research on clinical decisions, community violence and the evidence ecosystem has made many contributions to clinical and public policy and to legislation. His innovations include the Universities' Police Science Institute in Wales, the ‘Cardiff Model’ for violence prevention which was adopted in the 2008 UK alcohol strategy, by the Coalition Government in 2010 and by the present UK Government in its 2018 serious violence strategy, and a comprehensive care pathway for people harmed by violence. He is a member of the Cabinet Office What Works Council, and the Home Office Science Advisory Council.
This article is included in the Centre for Justice Innovation's Monument Fellowship book, Curing violence: How we can become a less violent society.