| Abstract|| |
Background: Assault is a common mechanism of injury among female trauma victims. This paper identifies risk factors for assault in female victims and explores the interplay between identified predictors of assault and their combined contribution to female violent victimization. Materials and Methods: A retrospective analysis of all female trauma patients was performed using the Illinois Department of Public Health Trauma Registry from 1999-2003. Patients with assault listed as their mechanism of injury were compared to patients with other mechanisms of injury. Bivariate and multivariate analyses were performed using STATA statistical software to identify independent risk factors for assault. Finally, interaction affects were studied among these identified risk factors. Results: Female victims of assault were more likely to be African American (OR 1.32, P < 0.001), lack insurance (OR 1.79, P < 0.001), and to have tested positive for drugs (OR 1.32, P < 0.001) than women with other mechanisms of injury. In addition to the independent effects of these variables, patient drug use and lack of insurance demonstrated interaction effects (OR 1.67, P = 0.02). Conclusion: In this study, women of color, the uninsured, and those using drugs were disproportionately represented among assault victims, highlighting further evidence of trauma disparities. Most significantly, this study demonstrates that predictors of assault in women frequently coexist and both independently and in combination may increase the risk for female violent victimization.
Keywords: Assault, disparities, female trauma patients, risk factors
|How to cite this article:|
Periyanayagam U, Shaheen AW, Crandall M. Predictors of assault among urban female trauma patients. J Emerg Trauma Shock 2012;5:299-303
| Introduction|| |
Approximately 1.7 million persons were treated in U.S. hospital emergency departments for non-fatal physical assaults in 2000.  Assault is defined as any physical contact with another person without their consent. Breslau and associates estimated the lifetime prevalence of physical assault in the United States to be 8.3% for both genders.  Other studies have reported lifetime rates of physical assault from 7% to 12% for women and from 10% to 19% in males. ,,, Although the risk for assault exists in all ethnic, racial, and socioeconomic classes, certain demographic groups experience disproportionate rates of physical violence. ,
Several studies show racial disparities exist among victims of assault. ,,,  Amstadter and associates showed African Americans were two-fold as likely to have experienced a new physical assault compared to their Caucasian counterparts.  Other studies have confirmed the finding that African Americans are at the greatest risk of violence. , In his assessment of lifetime rates of violence among men and women, Kilpatrick et al. found 28% of African Americans reported that they had been assault victims, compared with 19% of Caucasians.  Low income is also reported as a predictive factor for assault. A study of income and risk of assault by Kessler et al., found persons of both sexes with annual household incomes below $ 15,000 are one and a half times more likely to be physically assaulted as persons with annual household incomes above $ 15,000.  The inverse relationship between socioeconomic status and susceptibility to assault was confirmed again in a study in Milwaukee, Wisconsin study adolescent assault victims which found subjects with lower median household incomes experienced higher rates of interpersonal violence. 
The costs of nonfatal physical assault are significant and entail more than simply the physical injury sustained by the victim. Victims of violent assault also experience psychological injury, lost productivity from missed work and school days, and loss of quality of life due to pain and suffering. Lifetime costs for intentional injury for patients twelve and older were estimated at $ 187 billion during the period from 1987-1990, with $ 96 billion dollars attributed to injury from assault.  The estimate $ 96 billion dollar cost associated with injury from assault included approximately $ 8 billion dollars for medical costs, $ 53 billion for mental health care costs and $ 35 billion for lost quality of life.  This breaks down to a per physical injury cost for assault victims of $ 22,000.
Although racial and socioeconomic disparities and the costs of these disparities to assault victims is well documented, the predictive factors that lead to these disparities in female assault victims remains an open area of research. , In this paper, predictors of assault in female victims are identified. More significantly the interplay between the identified predictors of assault and their combined contribution to violent victimization in women is better elucidated. Existing literature fails to show the influence that one predictor of assault has on the co-existence of other predictors of assault. This interplay suggests that even though individual risk factors may increase the likelihood of assault, the combination of multiple risk factors, creates a susceptible patient population that is significantly more likely to experience assault. The identification of this interaction between two different risk factors could be useful to select a population more susceptible to the risk for assault and trauma than any other individual risk factor alone would be able to predict. Selecting for a susceptible population would allow public health efforts to be focused appropriately on the highest risk population. Additionally, existing literature largely draws conclusions from all emergency room visits and fails to specifically analyze "severe assaults," or assaults serious enough in nature to necessitate trauma activation. The purpose of this study is to determine predictors of assault and predictors of high-risk behaviors in women who sustain injuries significant enough to require State of Illinois trauma activation. Identifying the predictive factors that make women a susceptible population to assault requiring trauma activation will help inform injury prevention efforts and ideally lead to a reduction in violent victimization among women.
| Materials and Methods|| |
To identify both predictors of severe assault and the high-risk behaviors of severely assaulted women, a retrospective analysis of all female trauma patients was performed using the Illinois Department of Public Health Trauma Registry. The registry is a state-wide, mandatory reporting system that contains de-identified information for all trauma patients from all trauma centers in Illinois. Permission to use the database was obtained from the Illinois Department of Public Health and this study was approved by the University Institutional Review Board.
For the period 1999-2003, all female trauma patients aged 15-50 were identified. The following E-codes for assault were used to identify the female victims: 960 (0.0-0.1), 961, 962 (0.0-0.9), 963-964, 965 (0.0-0.9), 966, 967 (0.0-0.9), 968 (0.0-0.9) and victims of gunshot and stabbing wounds were included in this group. The remaining women (n = 23,841) who had other, non-assault mechanisms of injury served as the control group.
Demographic data was collected on both the study and control groups to include age, race, pregnancy status, alcohol and drug use, insurance status, residence in a rural area [Table 1]. The cut off for pregnancy in the Illinois State Trauma Registry is 24 weeks. No other indicators of pregnancy are reported to the state. Alcohol use was determined based on blood alcohol levels. Drug use was based on the presence of one or more of the following seven substances: amphetamine, barbiturates, benzodiazepines, cocaine, marijuana, opiates, and phencyclidine. Patients were considered uninsured if they had Medicaid or were listed as having no insurance. Medicaid patients were considered uninsured because many of these patients entered the hospital without insurance and were enrolled in Medicaid only during their hospital stay. Any other forms of insurance granted patients an insured status. Scene of accident was used to determine rural status in the registry. Chicago, Springfield, and East Saint Louis are considered urban; all other areas fall under the rural category.
|Table 1: Demographic characteristics and predictors of assault in female trauma victims by mechanism of injury|
Click here to view
Data was analyzed using STATA statistical software. A logistic regression model was used to compare the two groups of female victims (those who were assaulted against those who had other mechanisms of injury) on variables including age, race, pregnancy status, alcohol and drug use, rural residence and lack of insurance, in order to identify factors predictive of assault and high-risk behaviors. Once factors predictive of assault were identified, a separate analysis was conducted to study interaction affects among these identified risk factors. This additional analysis examined the combined effect of the identified predictors of assault on end outcome- all possible variations of combined risk factors were studied. A P value less than 0.05 was considered statistically significant.
| Results|| |
In a retrospective review of all trauma patients from all trauma centers in the State of Illinois for the time period 1999-2003, 26,602 female victims were identified. Of these 26,602 female victims, 2,761 (10.4%) reported assault as their mechanism of injury, and the remaining 23,841 women had other listed mechanisms of injury.
In comparing the two groups, the age distribution in the assaulted women group was younger than in the other mechanism of injury group. About 53.4% of the assaulted women group was in the ages 15-30, while in the other mechanism of injury group, only 48.1% fell in this age range. In regards to race, 75.6% of assaults were against non-white women, in contrast to the female victims of non-assaultive traumas, of whom only 35.6% were non-white. This disparity was largely due to the significant percentage (64.3%) of assaulted women who were of African American descent. Other races did not factor significantly in the racial disparity between the assaulted and other mechanism of injury groups, largely due to small total numbers in the study population. In addition to being non-white, women who were assaulted were also more likely to use alcohol (54.7% versus 39.6%) and drugs (25.0% versus 16.2%, P < 0.001) than women in the other mechanisms of injury group [Table 1].
After comparing demographic differences between female trauma victims who were assaulted and female trauma victims who reported other mechanisms of injury, attention turned to identifying specific predictors of assault among the female assault victims.
On multivariate regression analysis, three variables were noted to be independent predictors of assault in the female victims including African American race (OR 3.88, P < 0.001), lack of insurance (OR 1.79, P < 0.001), and drug use (OR 1.32, P < 0.001) [Table 2]. Age under 30 also predicted assault (OR = 1.17, P < 0.001). Neither pregnancy nor rural residence appeared to play a role in female victims' likelihood of being assaulted.
In this data set, female victims with one identified predictor of assault often exhibited other predictive factors for assault. That is, risk taking behaviors in female assault victims often existed in tandem. Multivariate analysis to predict the likelihood of certain risk factors among female assault victims was undertaken.
In [Table 3], predictors of alcohol use in female assault victims are analyzed. The data shows assaulted women who use alcohol are also significantly more likely to be uninsured (OR 3.5, P < 0.001) and also more likely to use drugs (OR = 1.7, P < 0.001).
These findings were confirmed again when predictors of drug use and uninsured status were analyzed individually. Female assault victims who used drugs were also more likely to be uninsured (OR = 1.8, P < 0.001) and to use alcohol (OR = 1.7, P < 0.001). African-American race also appeared to be more prevalent among female assault victims who used drugs (OR = 1.5, P = 0.009). In contrast, neither Latina race nor age appeared statistically significant predictors of drug use in female assault victims [Table 4].
Predictors of uninsured status in female assault victims included African American race (OR 4.6, P < 0.001), alcohol use (OR 3.3, P < 0.001) and drug use (OR 1.7, P < 0.001). Additionally, Latina race was also a statistically significant predictor of being uninsured for female trauma patients who reported assault as their mechanism of injury [Table 5].
Clearly, female assault victims often exhibit more than one high-risk behavior or risk factor for assault. These risk factors may not just exist in tandem but may also interact to exert a shared and therefore increased influence on the overall susceptibility of female victims to assault. The combined influence, or interaction affects, of these risk factors on violent victimization was tested in a regression analysis [Table 6]. The regression model showed that drug use and being uninsured had a combined effect of predicting assault (OR 1.67, P < 0.02).
| Discussion|| |
Characteristics of female assault victims and predictors of physical assault in female trauma patients were examined in this study. The study showed women of African-American race, women without insurance, and women who used alcohol and/or drugs are disproportionately represented among assault victims. A significant finding of this study was that female assault victims with one predictor of assault were more likely to have multiple predictors of assault and when drug use is combined with lack of insurance the likelihood of being assaulted increases. While there are known disparities in trauma, this paper highlights the interaction between these risk factors, which has not previously been studied in this patient population. Lack of insurance and drug abuse are more frequent among people with low socioeconomic status, however, having both of these factors increase the risk more that each individual factor would predict. Therefore, this susceptible population bears more of the risk for assault and trauma than any individual risk factor would predict. This underscores the observation that there are racial and socioeconomic disparities with respect to violent victimization, and makes an argument for specific interventions to this high risk population.
This study demonstrates that predictors of assault in women frequently coexist and both independently and in combination may increase risk. The underlying causation for this phenomenon is yet unexplained. One possibility is that one of the predictors of assault identified in this study, for example, lack of insurance, is the unifying link that increases the likelihood of a victim exhibiting all other predictors of assault. Another possibility is that an unnamed risk factor such as poverty or education level is the link that explains the co-existence of all of these high-risk taking behaviors in one setting. There is also the possibility that the temporal order between victimization and the identified predictors of assault is inverse to what is interpreted in this paper. That is, instead of drug use and lack of insurance being factors predicting the likelihood of women to experience assault, females may be more likely to use drugs and alcohol and not have insurance precisely because they have been assaulted.
The significance of the co-existence of multiple predictors of assaults in female victims is relevant both to the understanding of what leads to gender-based violent victimization and how healthcare providers and society can most effectively intervene. However, this study is not without limitations, including defining a rural/urban dichotomy on the basis of population, without considering boundaries or suburbs, and possible lack of generalizability as this study is restricted to urban patients from a single state in the U.S. It also would have been beneficial to have data on a wider range of predictors of assault including socioeconomic status, education level, U.S. birth and history of previous abuse (physical, sexual, or psychological). The database does not include those variables.
| Conclusion|| |
Despite its limitations, this is a powerful study that demonstrates the interdependence of risk factors for assault in a population-based sample of women in a large Midwestern state. This study helps us gain an understanding of how predictors of assault in female victims interact and lead to an increase susceptibility to violent victimization. Although there is no acceptable justification for violence against women, understanding the factors that predict assault in women can serve as a way to increase awareness of the victimization of women and to help better direct intervention and prevention efforts, especially in areas such as multimodality drug or treatment programs or comprehensive domestic violence shelters offering drug and alcohol services. Future research should also be directed at identifying protective factors against female violent victimization, which remains a yet unexplored area of study.
| References|| |
|1.||Centers for Disease Control and Prevention (CDC). National estimates of nonfatal injuries treated in hospital emergency departments, United States, 2000. MMWR Morb Mortal Wkly Rep 2001;50:340-6. |
|2.||Breslau N, Davis GC, Andreski P, Petersen E. Traumatic events and posttraumatic stress disorder in an urban population of young adults. Arch Gen Psychiatry 1991;48:216-22. |
|3.||Norris FH. Epidemiology of trauma: Frequency and impact of different potentially traumatic events on different demographic events. J Consult Clin Psychol 1992;60:409-18. |
|4.||Kilpatrick DG, Seymour AK, Boyle J. America Speaks Out: Citizens Attitudes About Victim's Rights and Violence. Arlington, VA: National Victim Center; 1991. |
|5.||Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB. Posttraumatic stress disorder in the National Comorbidity Survey. Arch Gen Psychiatry 1995;52:1048-60. |
|6.||Resnick HS, Kilpatrick DG, Dansky BS, Saunders BE, Best CL. Prevalence of civilian trauma and PTSD in a representative sample of women. J Consult Clin Psychol 1993;61:985-91. |
|7.||Amstadter AB, Elwood LS, Begle AM, Gudmundsdottir B, Smith DW, Resnick HS, et al. Predictors of Physical Assault Victimization: Findings from the National Survey of Adolescents. Addict Behav 2011;36:814-20. |
|8.||Melzer-Lange MD, van Thatcher CD, Liu J, Zhu S. Urban community characteristics and adolescent assault victims. WMJ 2007;106:394-6. |
|9.||Miller TR, Cohen MA, Rossman SB. Victim costs of violent crime and resulting injuries. Health Aff (Millwood) 1993;12:186-97. |
|10.||Ringel C. Criminal Victimization 1996: Changes 1995-96 With Trends 1993-96, Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics; 1997, NCJ 165812. |
|11.||Tjaden PG, Thoennes N. National Institute of Justice, Centers for Disease Control and Prevention. U.S. Department of Justice, National Institute of Justice and Centers for Disease Control and Prevention Research Report, 2000, NCJ 183781. |
Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
Source of Support: None, Conflict of Interest: None
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]