|Home | About | Journals | Submit | Contact Us | Français|
Hispanics, Blacks, and women are disproportionately burdened by intimate partner violence. Barriers to seeking medical care play an important role in victims accessing the full myriad of services they need. A secondary analysis of data collected over a 6-month period at a coordinated domestic violence social agency was completed to assess predictors of seeking medical care after experiencing intimate partner violence. A hierarchical logistic regression was conducted to assess the predictive ability of socioeconomic factors, type of abuse, and severity of abuse. Hispanic victims of intimate partner violence were less likely to seek medical attention compared to non-Hispanic Whites, even after controlling for socioeconomic factors, type of abuse, and severity of abuse, Adjusted Odds Ratio (AOR) = .40, p = .05, 95% CI [.164, .995]. Victims reporting physical abuse were over seven times more likely to seek medical attention, AOR = 8.02, p = .04, 95% CI [2.35, 27.34]. Medical care needs to be incorporated into coordinated social services offered to victims of intimate partner violence.
Intimate partner violence (IPV), the physical, sexual, and psychological abuse perpetrated by a current or former spouse or boyfriend/girlfriend, is a significant public health problem associated with numerous health consequences (Centers for Disease Control and Prevention [CDC], 2006). It is well documented that IPV is associated with an increased risk for numerous physical and mental health problems including injury, chronic pain, gastrointestinal and gynecological disorders, sexually transmitted infections (STIs), depression, and posttraumatic stress disorder (Campbell, 2002). Although it is clear that victims of IPV need to access medical care, there are numerous obstacles that interfere with their ability to do so. Nevertheless, little research exists exploring the factors associated with seeking medical care among ethnically diverse samples of women and men seeking IPV social services in urban centers.
The purpose of this study is to (a) identify if race and ethnicity is a predictor of seeking medical care when other socioeconomic factors are controlled for (i.e., age, financial dependence on partner and education), (b) explore if the type of abuse (i.e., physical, sexual, verbal, or psychological) is associated with seeking medical care, and (c) explore if the severity of abuse is associated with seeking medical care. Data for this study were collected as part of the Partnership for Domestic Violence Prevention (PDVP), a community-based participatory research study that aimed to assess the needs and preferences of IPV prevention programs targeting Hispanics in Miami-Dade County. PDVP received partial funding for this research project from the Center of Excellence for Health Disparities Research: EL CENTRO, National Center for Minority Health and Health Disparities Grant P60MD00266.
Findings from studies examining racial and ethnic differences in prevalence of intimate partner violence are conflicting although it appears that the overall rates of violence vary little by race or ethnicity. Some researchers have found that Hispanic (2.5 times greater) and Black (3.7 times greater) women are disproportionately affected by the occurrence and consequences of IPV when compared to White women (Caetano, Field, Ramisetty-Mikler, & McGrath, 2005). At least one population-based study has documented that Hispanics in cohabitating intimate relationships experience over twice the incidence and four times the recurrence of IPV when compared to their non-Hispanic White counterparts, even when socioeconomic factors are considered (Caetano et al., 2005). In a clinical sample, researchers have also documented that Hispanic women are more likely to report IPV within the past 5 years and within the past year than non-Hispanic White women (Bonomi, Anderson, Cannon, Slesnick, & Rodriguez, 2009). The findings from the National Violence Against Women Survey, however, indicate no difference in the prevalence of intimate partner violence by Hispanic versus non-Hispanic women when socioeconomic factors are controlled for (Tjaden & Theonnes, 2000).
Likewise, the risk of IPV among Black women appears to be about the same as that of White women, particularly when controlling for income (Rennison & Planty, 2003; Tjaden & Thoennes, 2000). Both Black and Hispanic women, however, appear to experience higher rates of lethal violence and higher rates of more severe forms of violence (West, 2004). For example, Hispanic females are more likely to report being pushed, shoved or grabbed; slapped; kicked, bitten, or hit; choked; forced into sex; or threatened with a gun when compared to non-Hispanic White women (Caetano, Schaffer, & Cunradi, 2001; Tjaden & Thoennes, 2000). Researchers have also found that Hispanic females are more likely to suffer from negative IPV health consequences such as depression (Bonomi et al., 2009), suicidal ideation and attempts (Krishnan, Hilbert, & VanLeeuwen, 2001) and homicide (Gonzalez-Guarda & Luke, 2009).
Despite the need that Hispanic and Black female victims of IPV have in accessing health related services, there are number of barriers that impede them from accessing quality, culturally sensitive care. These include not knowing where and how to access services, cultural and linguistic barriers, low insurance coverage rates, and a general sense of mistrust with health and social services (Bloom et al., 2009; Cuevas & Sabina, 2010; Gonzalez-Guarda, Vasquez, Urrutia, Villarruel, & Pergallo, 2011; West, 2004). Further contributing to barriers is the fact that domestic violence-related agencies tends to provide services that are fragmented and do not adequately address the physical and psychological health needs of clients and families (Warshaw, Gugenheim, Horoney, & Barnes, 2003). Consequently, in order for victims to access health care, they are often required to receive referrals, travel long distances, make multiple trips, and/or qualify for services. The combination of culturally specific barriers, system-related barriers and the isolation that is often experienced by Black and Hispanic women living with IPV create significant barriers to accessing the health care services that they need.
This study is a secondary data analysis of intake forms completed at the coordinated domestic violence agency in South Florida. Intake forms for all clients seen in a 6-month period (March–September 2010) were analyzed. This time frame was chosen because universal intake forms were adopted by the agency in March 2010, which eliminated inconsistencies in the type and quality of data gathered by the partners providing services on site. The universal intake forms contained information about the client-seeking services, their experiences with domestic violence, perpetrator characteristics, risk for safety, and needs for health and social services.
For this study, the dataset used was collected previously as part of a larger community-based participatory research study. This larger study included: focus group interviews with violence service providers and community members (Gonzalez-Guarda, Cummings, Becerra, & Fernandez, 2011), a community forum (Gonzalez-Guarda, Lipman, & Cummings, 2011), and a secondary data analysis of intake forms completed by IPV victims seeking services at a large county domestic violence service agency during a 6-month period. Preliminary analysis of the intake forms indicated that there were significant differences between Hispanic and non-Hispanic clients regarding likelihood of seeking medical care (Fernandez, Biacone, Lopez, Gonzalez-Guarda, & Mesa, 2011). This study expands upon these findings by exploring the predictors of seeking medical care in this sample. The original instrument was designed to identify the needs and preferences for IPV prevention among Hispanics in a South Florida community.
Approval from the Institutional Review Board (IRB) was received prior to conducting the secondary analysis of intake forms described in this study. The intake form, which is available in English, Spanish, and Creole, was completed when the victim arrived at the domestic violence agency. In order to facilitate the intake process, a victim advocate was present to assist the victim in completing this form. The victim advocate was also available to make recommendations and referrals for appropriate services and to obtain informed consent to share the collected information with agency partners. Victim advocates are counselors and social workers, many with master’s degrees, with many years of experience in serving victims of domestic violence. Victim advocates regularly participate in specific training to ensure sensitive and adequate responsiveness to victims’ needs. All intake forms were kept within the victim’s individual file and are kept in a secure, locked area to ensure confidentiality. All the intake forms completed at the coordinated domestic violence center were reviewed and assigned a unique number. The files containing the universal intake form that was implemented in March 2010 and a completed date that was between March and September 2010 were identified. Research assistants retrieved these files and entered the deidentified data into an Excel database according to a codebook. The quality of the data was ensured through multiple data-entry checks and the correction of discrepancies between these. Each volunteer participated in a training session to promote accuracy of data entry. As required by the IRB, all study personnel and volunteers completed the Collaborative Institutional Training Initiative (CITI) certification on research ethics and confidentiality. Data from the Excel sheet were imported to SPSS for statistical analysis.
Participants of this study included all clients seeking services at the coordinated domestic violence agency located in a large ethnically diverse metropolitan area in South Florida for the first time between March 2009 and September 2010 who reported IPV (n = 459). This agency includes more than 30 community partners that provide a variety of programs and services including legal aid, social support services, emergency financial assistance, and referral for shelters in one integrated location. Clients seen during the 6-month study period who did not have a completed intake form on file or were return visitors were excluded from the study. The intake form was designed for use by all of the partners and included information about the victim, perpetrator, children involved, and the abuse. Although the domestic violence agency accepts participants from many types of violence-related events, the analysis for this study was limited to clients who were identified as victims of IPV. IPV was defined as physical, sexual, or emotional/psychological violence or abuse perpetrated by a current or former boyfriend/girlfriend or spouse (CDC, 2006).
After the participants who were reporting IPV were identified (n = 459) and cases with missing values were removed (n = 184), there were 275 records included in the final analysis. Statistical differences between the groups (included in analysis or not included in the analysis) were compared. The only differences found were “Do you speak English?” (p = .033), “Are you a victim of psychological abuse?” (p = .034), and “Have you ever experienced sexual abuse?” (p = .015). The participants with missing data reported a higher percentage of sexual abuse (23.9%–14.1%) and psychological abuse (69.6%–59.4%), whereas participants without missing data reported a higher percentage of English speakers (71.1%–27.9 %).
There were no differences found in gender, age, race, relationship to abuser, verbal abuse, physical abuse, stalking, calling the police, abuser arrested, order of protection obtained, seeking medical attention after an assault, use of a weapon against the victim, or financial dependence of abuser (see Table 1).
The sample included 28 males (6.1%) and 431 females (93.9%). Ages ranged from 16 to 78 (33.6 ± 10.0) years. The sample consisted of 300 (65.6%) Hispanics, 102 (22.2%) Blacks, and 52 (11.3%) individuals reporting another race or ethnicity (White, Asian, multiracial, and Hawaiian/Pacific Islander). The majority (55.2%) of victims reported that they spoke English. In this sample, 63.5% reported psychological abuse and 17.6% reported sexual abuse.
The following variables were chosen for the analysis: (a) socioeconomic status (SES); (age, race, and financially dependent on the abuser), (b) type of abuse (physical, verbal, psychological, and stalking), and (c) severity of abuse (abuser was arrested and a weapon was used to abuse the victim).
Clients were asked to report their age. This was a continuous variable.
Clients were classified into groups based on self-identification: Hispanic, Black, White or Other (White, Asian, Native Hawaiian/Pacific Islander, and multiracial). We recognize that individuals can be both White and Hispanic or Black and Hispanic. However, in this case, individuals chose one option from the list. Due to the low number of participants whom identified as White (8%) or Other (3%), these two categories were combined (11%).
A series of yes/no questions assessed financial dependence on the abuser, type of abuse (physical, sexual, psychological, and/or stalking), and severity (if abuser was arrested, and if weapon was used).
Questions that ascertained socioeconomic status, race and ethnicity, type of abuse, severity of abuse, and seeking out medical attention on the intake form were selected for this study. Frequencies for these variables were calculated. In order to address the variables with missing values, a missing variable analysis was performed. Descriptive statistics were calculated for the participants. The two groups (those whose data were not included and those whose data were included in the final analysis) were compared using a chi-square analysis and t-test statistics. Group comparisons were performed to identify statistical differences between groups.
A hierarchical logistic regression was done to determine whether selected variables (SES, type of abuse, and severity of abuse) could predict seeking medical care after a violent episode. A hierarchical logistic regression analysis was performed to predict group membership between victims of intimate partner violence who had sought medical care to those who did not. Variables were entered into the model in three different blocks (see Table 2). To control for the group differences based on SES (age, race/ethnicity, and financially dependent) was entered into block 1. In block 2, the types of abuse were entered. In block 3, the severity of abuse was entered. By entering variables stepwise into the equation, variables entered in earlier blocks can be statistically controlled for while identifying significant blocks and individual predictors. Multivariate normalcy, linearity, and homoscedasticity were assessed and were not violated (Ottenbacher, Ottenbacher, Tooth, & Ostir, 2004). Based on distances and residuals (Mahalanobis, Cooks, leverage, studentized, and standardized), no multivariate outliers were present.
The model with SES predicting medical attention was not significant (p = .08). To control for SES, age and financial dependence of abuser were enter into block 1. Race and ethnicity were also entered. These variables explained 4.5% (Nagelkerke’s R2 = .045) of the variance. Hispanic participants when compared to non-Hispanic Whites were 63.1% less likely to seek medical attention after an incident of IPV, AOR = 0.37, p = .02, 95% CI [0.16, 0.84].
In block 2, the types of abuse (physical, verbal, psychological, and stalking) were added. Types of abuse were significant in predicting seeking medical attention (p < .001). The types of abuse alone explained an additional 12.7% of the variance in the model, and the overall model with SES and types of abuse variables explained 17.2% of the variance (Nagelkerke’s R2 =.172). Physical abuse was individually predictive. Participants who reported physical abuse were over eight times more likely to seek medical attention, AOR = 8.17, p = .001, 95% CI [2.43, 27.41]. Being Hispanic remained a significant individual predictor, AOR = 0.40, p = .04, 95% CI [.16, .96].
The variables in Block 3 describing the severity of abuse were not significant predictors of seeking health-care (p = .28). Nevertheless, the overall model remained significant when the severity of abuse indicators were added (p < .001). The overall model with SES, types of abuse, and severity of abuse explained 18% of the variance (Nagelkerke’s R2 = .18). Two variables remained individually predictive in block 3. Hispanics were 59.6% less likely to seek medical attention, AOR = .40, p = .05, 95% CI [.164, .998]. Participants who reported physical abuse were more than eight times more likely to seek medical attention, AOR = 8.02, p = .04, 95% CI [2.35, 27.34] (see Table 2). After controlling for SES and adding the types of abuse, 97.1% of the group that did not seek medical attention and 15.9% of the group that did access health care were correctly classified. The overall percentage correctly predicted group membership was 76.7%. (see Table 3)
This study provides insight about the help-seeking behaviors and needs of a racial/ethnically diverse sample of victims of IPV seeking social services in a large urban center. The study population that used the domestic violence agency mirrored the racial/ethnic composition of the area, which is 65% Hispanic, 18.9% Black, and 15.4% White non-Hispanic (U.S. Census Bureau, 2010). Nevertheless, as also documented in national studies indicating that women are 7 to 14 times more likely to report being physically assaulted (beaten up, choked, tried to be drowned, or threatened with a gun or knife) than men (Tjaden & Thoennes, 2000), most of the victims in this study were women (93.9%). The study of IPV and help-seeking behaviors among diverse samples is important because their needs for services are likely to be different.
This study identified differences between groups of IPV victims who sought medical attention and victims who did not. When controlling for age and financial dependency on the abuser, victims who self-identified as Hispanic were less likely to seek medical attention. This is consistent with qualitative components of the parent study, which identified the lack of knowledge about existing services and the poor quality of these services as major problems with the services provided to victims and their families (Gonzalez-Guarda, Cummings et al., 2011; Gonzalez-Guarda, Lipman et al., 2011). There are several additional obstacles that impede Hispanics from accessing health care, including low levels of health insurance, as also supported by this study, cultural and linguistic barriers, and distrust of services (Bloom et al., 2009; Cuevas & Sabina, 2010; Gonzalez-Guarda, Vasquez, et al., 2011). Nevertheless, the clients included in this study were, in fact, accessing social services. By incorporating medical care into the coordinated domestic violence agency where they were seeking help, access to the medical care they need would likely increase.
When controlling for SES and race and ethnicity, victims who reported physical abuse were more likely to access health care. The relationship between health care use among women experiencing physical violence found in this study is similar to previous studies that strongly support increased health care used with increasing severity of violence (Campbell, 2002).
The findings in this study have important implications for domestic violence services. Hispanic victims of IPV appear to be less likely to seek medical attention than their non-Hispanic counterparts. Socioeconomic factors, differences in the type of abuse (physical abuse) that they experience, and barriers in accessing care may explain some of these differences. These factors need to be further explored through research. Nevertheless, if Hispanic victims of IPV are willing to seek domestic violence services at a social agency, then opportunities for them to access medical care should be provided. If health care services were made widely available to Hispanics at the agencies where coordinated domestic violence services are being provided, then the evaluation of physical injuries and the consequences that IPV has had on their physical and mental health could be better assessed and addressed.
In addition to injuries directly related to the violence, numerous studies have supported increased rates of physical and mental illness in women experiencing IPV. Chronic physical conditions for which IPV is a significant risk factor include gastrointestinal disorders (loss of appetite, eating disorders), neurological problems (fainting or passing out, severe headaches, vision and hearing problems), urinary tract and other infections, and shortness of breath (Campbell, 2002; Campbell, Sharps, Gary, Campbell, & Lopez, 2002; Naumann, Langford, Torres, Campbell, & Glass, 1999). An increased risk for STIs among women who have experienced trauma related to IPV (Champion, Piper, Holden, Korte, & Shain, 2004; Cohen et al., 2000; Coker, 2007; Gielen, McDonnell, & O’Campo, 2002; Stockman, Campbell, & Celentano, 2010) has been well established by investigations with many different populations and settings and using both qualitative and quantitative data. Recent work has supported reproductive coercion (interference with a woman’s ability to make independent decisions about reproduction, including behaviors such as deliberately damaging condoms) as an important mechanism linking partner violence and STI risk/unintended pregnancy (Gee, Mitra, Wan, Chavkin, & Long, 2009; Gielen et al., 2002; Miller et al., 2010; Moore, Frohwirth, & Miller, 2010; Stockman et al., 2010). A recent study of more than 13,000 women in intimate relationships in the United States found that IPV was significantly associated with HIV infection, AOR = 3.44, 95% CI [1.28–9.22], and that 11.8% of HIV was attributed to IPV in the past year (Sareen, Pagura, & Grant, 2009).
Given the important risks associated with IPV and the low rate of health care use by this population, incorporating a primary health care clinic into agencies providing coordinated, comprehensive services could offer an important means for abused women, particularly Hispanic women, to access health care services by providers sensitive to the physical and mental health care needs of this population.
Further work is needed to better understand the relationship between decreased health care use and race/ethnicity found in this group. More specifically, additional predictors of seeking medical care and other health and social services should be investigated among both samples of clients already seeking some type of service as well as community samples, as these are likely to vary. Special attention should be given to the effects that race and ethnicity have on accessing health-related services. Consequently, research needs to explore both inter- and intra-ethnic variations in health care usage. By identifying barriers to healthcare use, interventions and services can be developed and tested. Further, the health care outcomes of IPV victims seeking health care and potential health disparities related to these should also be explored. These insights are fundamental to improving the health and well-being of victims affected by IPV.
This study had a number of limitations. First, the data were self-reported. Consequently, the information that is given may or may not accurately depict the reporter’s characteristics, experiences with IPV and help-seeking behaviors. For example, victims may not report all types of abuse due to fear of reprisal and shame. Research has found that victims who are perpetrated by strangers are much more likely to make a report with law enforcement than those perpetrated by intimate partners (Tjaden & Thoennes, 2000). Nevertheless, many of the items that are collected on the intake form are ones that could only be collected through self-report because victims are often alone with the perpetrator during incidents of IPV. Additionally, even though the intake forms were provided in English, Spanish, and Creole, the English-speaking victims completed more items on the forms than the victims who did not speak English. With fewer non-English speaking victims completing the forms, differences between the groups may not have been appreciated. Further, the population in South Florida is unique when compared to the rest of the country. Even though the population in South Florida is similar to the victims seeking services, the information collected is strongly influenced by the high number of Hispanic victims in the area making generalizations of findings difficult to compare to other populations in the United States.
Among this group of Hispanic women seeking services related to IPV, only a small percentage have accessed health care despite a large body of evidence suggesting that they are at risk for a wide range of mental and physical health problems. Accessible, culturally relevant and trauma-informed health care services are urgently needed for this at-risk population.
This article was funded by the National Center on Minority Health and Health Disparities of the National Institutes of Health (1P60MD002266-01; Nilda Peragallo, principal investigator). We would also like to thank the Coordinated Victim Assistance Center (CVAC), Miami-Dade County Department of Human Services, the Community Advisory Board for the Partnership for Domestic Violence Prevention (PDVP), University of Miami student volunteers, Ivon Mesa (Director of CVAC), and Maria Becerra (President, Entre Nosotras Foundation) for their important contributions to this research.
Sarah L. Lawson, University of Miami School of Nursing and Health Studies, Coral Gables, Florida.
Kathryn Laughon, University of Virginia School of Nursing, Charlottesville, Virginia.
Rosa M. Gonzalez-Guarda, University of Miami School of Nursing and Health Studies, Coral Gables, Florida.