AIMS: The aim of this study was twofold. Firstly we identified victims of violence in national registers and discussed strengths and weaknesses of this approach. Secondly we assessed the magnitude of violence and the characteristics of the victims using register-based data. METHODS: We used three nationwide registers to identify victims of violence: The National Patient Register, the Victim Statistics, and the Causes of Death Register. We merged these data and assessed the degree of overlap between data sources. We identified a reference population by selecting all individuals in Denmark over 15 years of age that had not been exposed to violence. For the study population and the reference population, socioeconomic and demographic information were retrieved from Statistics Denmark. We used logistic regression models in a cross-sectional analysis to identify characteristics of victims of violence. RESULTS: In 2006, 22,000 individuals were registered as having been exposed to violence. About 70% of these victims were men. Most victims were identified from emergency room contacts and police records, and few from the Causes of Death Register. There was some overlap between the two large data sources. We found significant differences between victims and non-victims according to socio-economic status, education, marital status, and ethnic origin, and also between victims by source of identification. CONCLUSIONS: We have identified a study population consisting of individual victims of violence that opens for further studies on violence. The use of different data sources is a strength but also a potential weakness to epidemiological, health economic, and other analyses using these data.
AIMS: The aim of this study was twofold. Firstly we identified victims of violence in national registers and discussed strengths and weaknesses of this approach. Secondly we assessed the magnitude of violence and the characteristics of the victims using register-based data. METHODS: We used three nationwide registers to identify victims of violence: The National Patient Register, the Victim Statistics, and the Causes of Death Register. We merged these data and assessed the degree of overlap between data sources. We identified a reference population by selecting all individuals in Denmark over 15 years of age that had not been exposed to violence. For the study population and the reference population, socioeconomic and demographic information were retrieved from Statistics Denmark. We used logistic regression models in a cross-sectional analysis to identify characteristics of victims of violence. RESULTS: In 2006, 22,000 individuals were registered as having been exposed to violence. About 70% of these victims were men. Most victims were identified from emergency room contacts and police records, and few from the Causes of Death Register. There was some overlap between the two large data sources. We found significant differences between victims and non-victims according to socio-economic status, education, marital status, and ethnic origin, and also between victims by source of identification. CONCLUSIONS: We have identified a study population consisting of individual victims of violence that opens for further studies on violence. The use of different data sources is a strength but also a potential weakness to epidemiological, health economic, and other analyses using these data.
Authors: Federico Fortugno; Christina Katsakou; Stephen Bremner; Andrzej Kiejna; Lars Kjellin; Petr Nawka; Jiri Raboch; Thomas Kallert; Stefan Priebe Journal: PLoS One Date: 2013-03-19 Impact factor: 3.240