Article Text

Download PDFPDF
057 Utilizing the cardiff model to understand characteristics of individuals reporting violence
  1. Gretchen Baas1,2,
  2. Jasmine C Moore3,4,
  3. Daniel Wu2,4
  1. 1George Mason University, Fairfax, USA
  2. 2Emory University, Atlanta, USA
  3. 3Georgia Institute of Technology, Atlanta, USA
  4. 4Grady Health System, Atlanta, USA


Statement of Purpose To identify the characteristics of those that report violent crimes to law enforcement versus those that do not, explore how specific types of violent injury influences reporting decisions, and the association between encounter data points and reporting violent injuries to law enforcement.

Methods/Approach Secondary data analyses of the original Cardiff Model Pilot Study dataset from May 2015-November 2017 from Grady Memorial Hospital were conducted. An additional chart review was first conducted and merged to the main dataset. During the pilot study, participants were screened in the emergency room with the Information Sharing to Tackle Violence screen; of those screened, 300 were mappable violent injuries. Emergency department data was matched with law enforcement data at three location sensitivities: 100m, 500m, and 1000m. Descriptive statistics, chi-square, simple logistic regressions, and multivariable logistic regressions were conducted at each location sensitivity.

Results Associations between means of arrival, mechanism of injury, acuity, gender, chief complaint, and financial class at various location sensitivities. Multivariable logistic regressions revealed predictors between means of arrival (walk-in; 100m, p=0.044; 500m, p=0.028), location of injury (street) (500m, p=0.031), and gender (500m, p=0.015; 1000m, p=0.010). There are associations between demographic characteristics, violent injury type, encounter data points and reporting status at all three location sensitivities.

Conclusions Means of arrival, location of injury, and gender are significant predictors of reporting status. This study provides initial understanding of the associations and predictors of the characteristics of individuals when reporting violence. Additional research is needed with larger sample sizes to understand how these associations can differ between races, genders, and socioeconomic status.

Significance Our results provide law enforcement a deeper understanding of whom is impacted by violence and how that differs from their reported violent crimes.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.