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 analysisBMJ 2011; 342 doi: https://doi.org/10.1136/bmj.d3313 (Published 16 June 2011) Cite this as: BMJ 2011;342:d3313
All rapid responses
Florence C, Shepherd J, Brennan I and Simon T in their paper have
demonstrated that systematic collection, analyses and use of anonymised
emergency department(ED) recorded information on violence led to
substantial and significant reductions in violence-related hospital
Although violence has long been a public health issue, there
continues to be ambivalence when it comes to making it a health priority.
After all, crime reduction is the responsibility of Government departments
other than health and other public services - the police, courts, and
probation and prison services - are there to tackle it. However,
reflecting the fact that, globally, 1,200 people are murdered every day
and violence, not counting war, is the fifth commonest cause of death of
15-29 year olds, the WHO has made it a priority, emphasising the need for
multisectoral approaches to prevention.
It has been known for some time that police records tend to
underestimate violence levels and that emergency department injury records
are more representative of serious community violence. We have become
aware of similar data matching studies in Denmark and Norway and bring
them together here to suggest, for the first time, that there is
considerable consistency in Europe in the extent to which serious violence
is under ascertained by police services (Table).
Each series of data was obtained by ascertaining a consecutive series
of cases attending EDs for assault injuries, and determining how many of
them also appeared in police records. Column A represents cases known only
to EDs and column B represents cases ascertained from both police and ED
sources. The under recording ratio is defined as B/(A+B). 95% confidence
limits for this ratio are given for each series (7). In all series less
than half the ED cases were recorded by the police, although the B/(A+B)
ratio differs between series. There is no significant gender difference
where gender-specific recording rates are available, for the Funen Island
and Swansea series.
It therefore appears that, on average, police only record one third
or less of violence which results in ED treatment in three north European
countries. The principal reasons for under-recording have been identified
in the UK as fear of reprisals, not knowing the identity of the assailant,
antagonism towards the police and the injured person not wanting their own
conduct scrutinised - these factors may exert a remarkably similar
influence across national boundaries and sharing data between health
agencies and agencies which have traditionally had the responsibility for
tackling violence for example, the police, seems important.
Whatever the reasons, there are important policy implications. These
are that police data are a poor measure of serious violence; that health
services are a principal source of information about violence which is not
available anywhere else; and that violence in a city for example, can only
be understood and targeted if police and health data are combined. Doing
this, and the involvement of trauma service doctors in community safety
partnerships, is proving effective, particularly by directing police
activity (8). Reflecting this, the College of Emergency Medicine has
published guidance (9). It is noteworthy that countries like the US, where
weapon availability and use are far greater than in Europe, have not
developed policy in this area. This may reflect the dominance in public
policy thinking of firearm injury, punishment and other offender-
It is clear from the data presented here that public health and trauma
services have much to contribute to violence prevention across Europe and
1 Florence C, Shepherd J, Brennan I, Simon T. 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. British Medical Journal 2011;
2 Faergemann C. Interpersonal violence in the Odense municipality,
Denmark 1991-2002. PhD thesis. Odense. University of Southern Denmark
3 Brink O, Sorensen V. Is the dark figure of violence decreasing?
Nordisk Tidsskriff for Kriminalvidenskab 2001;88:230-9.
4 Steen K, Hunskaar S. Violence: a prospective study of police and health
care registrations in an urban community in Norway. Medicine, Science,
5 Sutherland I, Sivarajasingam V, Shepherd JP. Recording of community
violence by medical and police services. Injury Prevention 2002; 8:246-7.
6 Shepherd JP, Shapland M, Scully C. Recording by the police of violent
offences: an accident and emergency perspective. Medicine, Science, Law
7 Wilson EB. Probable inference, the law of succession, and statistical
inference. Journal of the American Statistical Association 1927;22:209-12.
8 Warburton Al, Shepherd JP. Tackling alcohol related violence in city
centres: effect of emergency medicine and police intervention. Emergency
Medicine Journal 2006;23:12-17.
9 Boyle A, Shepherd JP, Sheehan D (2009). Guidance on data sharing for
violence prevention. London. College of Emergency Medicine.
V Sivarajasingam, JP Shepherd, RG Newcombe
Violence and Society Research Group, School of Dentistry (JPS, VS) and
School of Medicine (RGN)
Cardiff University, Cardiff, UK
E mail: firstname.lastname@example.org
Comparison of A&E and police violent injury figures from several studies.
Competing interests: No competing interests
Florence et al present powerful evidence that sharing emergency
department data with community safety partnerships can bring about
meaningful reductions in community violence.(1) Information sharing is a
simple intervention that is almost entirely without harm or cost. The
magnitude of the benefit is far greater than any change to the Licensing
Act has achieved.
The linked editorial acknowledges that implementation of this model of
information sharing across the UK has been poor.(2) Why is this?
Sharing data for injury prevention is a culture shift for acute hospitals.
Many hospital clinicians and managers see injury prevention as a public
health and primary care role.
There are (groundless) concerns about data protection and breeches of
confidentiality. The Data Protection Act makes specific provision for data
sharing for injury prevention.
Hospital managers are incentivised by the Payment By Results system not to
reduce activity of low acuity patients at emergency departments. There are
currently no incentives to facilitate information sharing in the NHS
The Department of Health and Commissioners should develop a system of
incentives that promotes information sharing between emergency departments
and community partnerships.
1. BMJ 2011 342:d3313; doi:10.1136/bmj.d3313
2. BMJ 2011; 342:d2882
Competing interests: AB sits on the Department of Health's Information Sharing Steering Group. KS works for the Department of Health to implement Information Sharing.