Violence Prevention Information System
Scientific information on interpersonal violence – which includes child maltreatment, youth violence, intimate partner violence, elder abuse, sexual violence, multiple types of violence victimization and homicide – is scattered across a myriad of websites, statistical databases, technical reports, and often difficult-to-access academic journals. Often, these information sources only address one violence type, such as intimate partner violence. Furthermore, many existing databases have restricted coverage of interventions or prevalence of violence in one country or region of the world. The available information is also often difficult for non-academics to make sense of, and rarely is it presented in a concise, readable or visual format. Thus, at present, there is no single, authoritative, and comprehensive website which synthesises in a highly accessible form information on violence and its prevention.
In collaboration with the World Health Organization and other partners1, the Public Health Institute received funding from the UBS Optimus Foundation to fill this gap with the development of the Violence Prevention Information System (Violence Info). Violence Info aims to improve access to scientific information about all types of interpersonal violence, including findings on prevalence rates, risk factors, consequences, and prevention and response strategies, through creating a data repository and displaying the information in a user-friendly format on a website.
Development of Violence Info was conducted in three phases over a two year period. As the vision for Violence Info was to provide an overview on all aspects of several different forms of violence, traditional meta-analytical techniques for searching, extracting and synthesising data had to be adapted and refined to cope with the scale of the literature which met the criteria for inclusion in the data repository. The first phase of Violence Info development sought to systematically search and collate evidence from the meta-analyses and systematic reviews on each violence type and aspect. Single study data extracted from the systematic review, rather than the original single study, were used in this phase to populate the data repository. The exception to this strategy was homicide, which used data from the WHO Global Health Estimates. The second phase focused on filling gaps in data available in Violence Info (identified following Phase 1), through systematically searching and synthesizing evidence from single studies or large national surveys. To date this has been completed for child maltreatment, and systematic searches for the other violence types have also been developed and will be added to Violence Info as they become available (funding dependent). During the third phase, experts in the field of each violence type were contacted to suggest studies to fill identified gaps in the data. This continues to be an ongoing phase of Violence Info with experts and users being able to suggest appropriate studies.
In addition to the sheer scale of the data to be synthesised, Violence Info also aimed to present these summary estimates visually in an interactive website. The design and development of the web application took place concurrently with data extraction and methodological development. The Violence Info website is organised along two dimensions:
- Type of interpersonal violence, i.e. child maltreatment, youth violence, intimate partner violence, elder abuse and crosscutting categories such as sexual violence, violence against children, violence against women and homicide.
- The aspects of the public health approach to violence, i.e. prevalence, consequences, risk factors, prevention and response strategies.
Each violence type page presents a definition of the violence type and a summary global prevalence figure. Information on the prevalence, consequences, risk factors, and effectiveness of prevention and response strategies is extracted from published scientific studies and presented in different visualisations. All visualisations can be downloaded, shared, or embedded in other websites. There is also a help tool for each visualisation to facilitate interpretation of the data and understanding of the statistics used. Each violence type page also provides examples of interventions with some evidence for effectiveness. Key survey instruments used to gather information on the violence type are also summarised.
A studies section collates data from the different violence types and allows the user to explore the studies and their findings in depth, with options to filter the visualisations by a number of criteria. The user can also download the data repository tables in this section. These files collate information from all violence types and contain all extracted data, providing the user with more study detail than available on the website visualisations. The countries section allows the user to explore country-level data, including homicide estimates, and information on violence prevention including measures such as policies, laws, prevention programmes and victim services.
Violence Info is an ambitious project and literature reviewing and data synthesising on this scale has rarely been done before. The data repository currently contains almost 13,000 individual data points, from over 3,587 single studies. Violence Info currently holds summary estimates for more than 120 different risk factors at four ecological levels and for 45 different consequences across the different types of violence. It includes prevalence estimates across the various types (e.g. child maltreatment) and sub-types (e.g. sexual abuse) of violence for 95 countries around the world, including 59 low- and middle-income countries. Finally, summary estimates are also provided for the effectiveness of more than 38 different intervention strategies for preventing and responding to violence. The aim is to continue to update and maintain the Violence Info website and data repository, making it a one-stop shop for global violence prevention information.
1 Public Health Wales, University of Bristol and Interactive Things.