Over the past decade, the co-occurrence of HIV and intimate partner violence (IPV) has emerged as a significant public health problem.^{1,2,3,4,5,6,7,8,9,10,11} This co-occurrence is particularly salient in inner city emergency departments (EDs), which function as primary sources of care for women infected with HIV,^12–14 those at risk of infection¹⁰ and for women experiencing IPV.¹⁵

Past studies have found that 20–50% of the women seeking care in EDs reported lifetime IPV, and 11–14% reported IPV in the past 12 months,^15–17 with the variability in estimates appearing to depend on methods of assessing IPV as well as on size and geographical location of the ED.¹⁶ Every year, one to four million women seek ED care for IPV-related injuries,¹⁸ a staggering finding that underscores the fact that, for many abused women, the ED is the first, and sometimes, only link to the healthcare system.¹⁵ There is growing evidence that urban women with low income who experience IPV and are at high risk for HIV transmission often lack access to healthcare and use EDs as their predominant source of medical treatment.^12,19 Research has shown that ED staff often fail to assess for and document IPV, especially if no visible injuries are presented.^16,20,21 This failure is of particularly concern in the light of the aforementioned research indicating that women who have a history of IPV are at a high risk of engaging in HIV sexual risk behaviours or of being HIV-infected. Anonymous HIV prevalence surveys in EDs have found that 1–13% of ED patients do not know their HIV status.^14,22,23

Moreover, research suggests that IPV is associated with a number of sexual HIV risk factors, including: (1) engaging in unprotected sex^{1–3,6,8,10,24,25}; (2) higher rates of sexually transmitted infections (STIs)¹²; (3) sex with multiple partners^12,26; (4) disclosure of a STI or a positive HIV status^27,28; (5) trading sex for money or drugs^5,29; (6) having a risky sexual partner (eg, a person who injects drugs, is HIV positive, has had an STI and/or has had sex with multiple partners)^10,25,29,30; and (7) injecting drugs.³² It has also been shown that women who attempt to protect themselves from HIV—for example, by requesting condoms or refusing to have sex without a condom—experience higher rates of IPV.¹²

To date, research on the relationship between HIV risks and IPV among women receiving care in EDs has been conducted exclusively with small and non-representative samples. This sampling limitation is significant in the light of the prevalence of co-occurring HIV and IPV in women seen in EDs. To address this gap, we examine: (1) the prevalence rates of physical, injurious and sexual IPV among a random sample of 799 women receiving care in an ED located in a low-income, urban neighbourhood in the Bronx, New York; and (2) the associations between HIV risk behaviours and experiencing of physical, injurious and sexual IPV among this sample of women.

METHODS

Design, sampling and participants

Data for this study were collected from a New York City Hospital ED in the South Bronx that serves a catchment area of 1.1 million residents. From August 2001 to April 2003, participants were recruited during randomly selected 6 h time blocks. In 2002, during the enrolment period, a total of 18 045 unduplicated female patients aged >18 years visited the ED (unpublished data from the ED 2002 records). A total of 215 time blocks were selected, including 03:00–09:00 (7 blocks, 3%), 09:00– 15:00 (118 blocks, 55%), 15:00–21:00 (77 blocks, 36%) and 21:00–03:00 (13 time blocks, 6%). Among those selected time blocks, 29% occurred on a weekend. The probability of a specific time block being selected was adjusted to match the proportion of patients seen in the ED (according to ED census data from the previous year) based on the day of the week and time of day.

Recruitment procedures selected based on prior success with similar studies in other ED settings^16,32,33 were as follows: female research assistants (RAs), fluent in Spanish and English, approached every female patient who was admitted to the ED during the designated time blocks before or immediately after medical care. The RAs introduced the study to potential participants who, if interested, were asked to provide informed consent before data collection. On obtaining informed consent, the RA conducted the survey interview with the female participant in a private room in the ED. Arrangements were made with the doctors and nurses on duty so that participating women would not lose their positions in the treatment queue while being interviewed.

Eligibility criteria for this study included being a female aged >18 years and being admitted to the ED during a time block selected for inclusion in the study. Women who were admitted for psychiatric emergencies and women who showed severe cognitive or psychological impairment were excluded from the study because of their inability to give informed consent. For female patients in the ED who were designated as severe triage or moderate triage which required hospitalisation, interviews were conducted within 2 weeks’ of their ED visit in the patient’s hospital room before discharge, or in another private setting after discharge. All participants were compensated US$5 for their time and information. Study protocols were approved by the institutional review boards of the research institution and the study site.

Measurement

The RAs conducted a 10 min, structured interview. The assessment interview was designed to elicit self-reported data on the variables listed below.

Sociodemographics

Sociodemographic variables included: age, race/ethnicity and education, marital status, having children aged <18 years, experience of homelessness in the past 12 months, current employment status, and the reason for admission to the ED.

Sexual behaviours

Participants provided self-reported data on the number, type and gender of intimate partner(s) in the past 12 months, and frequency of vaginal sex in the past 6 months.

HIV risk indicators

HIV risks in the past 6 months that were assessed included: having had an STI or STI symptoms, injecting drugs, frequency of condom use during vaginal sex and having had sex with an injecting drug user (IDU) or a HIV-infected partner. Participants were asked whether they had been tested for HIV and their HIV serostatus.

Intimate partner violence

Experiencing IPV was assessed using the physical, sexual and injurious subscales from the Revised Conflict Tactics Scale (CTS2).³⁴ The CTS2 also classifies behaviours as minor or severe. Examples of items from each of the three subscales are: minor (eg, “Has a partner ever twisted your arm or hair?”) and severe physical IPV (eg, “Has a partner ever choked you?”); minor (eg, “Have you ever had a sprain, bruise, or small cut because of a fight with your partner?”) and severe injurious IPV (eg, “Have you ever passed out from being hit on the head by your partner in a fight?”); and minor (eg, “Has a partner ever insisted on having vaginal, anal or oral sex [but didn’t use physical force]?”) and severe sexual IPV (ie, rape, eg, “Has a partner ever used threat of force [like hitting, holding down, or using a weapon] to make you have vaginal, oral or anal sex?”). Participants were asked about IPV that occurred in their lifetime and in the past 6 months.

Statistical analysis

Univariate statistics were used to describe the sociodemographic characteristics, prevalence of HIV-risk behaviours and different types of IPV. Multivariate logistic regression analysis was used to obtain estimates regarding the association between HIV-risk-related behaviours and experiencing IPV in the past 6 months. For multivariate analyses, IPV items were combined as follows: (1) any physical and/or injurious IPV (ie, combining across severe and minor subscale items); (2) any severe physical and/or injurious IPV; (3) any sexual IPV (ie, combining minor and severe subscale sexual IPV items); and (4) severe sexual IPV. The CTS2 sexual IPV item, “Has a partner ever made you have vaginal, anal or oral sex without a condom?” was removed from the multivariate analysis to eliminate overlap between the construct of sexual IPV and HIV risk behaviours. In order to have meaningful comparisons for HIV sexual risk behaviours, those women who did not have any vaginal sex in the past 6 months were excluded from those analyses. Multivariate analyses also included covariance adjustments for age, ethnicity, high-school diploma/General Equivalency Diploma (GED; which is the high-school equivalency diploma), marital status, having children aged <18 years, employment and homelessness. Adjusted odds ratios (ORs) with their 95% confidence intervals (CIs) are reported.

RESULTS

Of the 1251 female patients approached by interviewers during the selected time blocks, 452 refused to participate, yielding a final study sample of 799 (65%) women; 6% of the interviews were conducted in Spanish. The majority of the women who refused to participate stated that they felt too ill to take part in the study.

Sociodemographics

Table 1 shows the descriptive statistics for the study sample. The majority of participants were identified as Latina, followed by African–American. Slightly more than half had a high-school diploma or GED. More than half were single or never married and about a quarter were divorced, widowed or separated. About a half reported having children aged <18 years. Two-fifths were currently employed. More than 1 in 10 reported having experienced homelessness in the past 12 months. About 1 in 40 women reported that the reason they visited the ED was due to IPV-related injuries.

Sexual behaviours

The prevalence rates of sexual behaviour among the sample are presented in table 2, part A. About 68% of the women had had vaginal intercourse in the past 6 months. The majority (87.6%) of female participants had only male sexual partners, 3.5% had only female partners, 3.3% had both male and female partners and 5.6% had no sexual partners during the past 12 months.

HIV risk

As shown in table 2, part B, about half of the women reported that they had never or sometimes used condoms during vaginal sex in the past 6 months. 3.4% had had vaginal sex with an IDU or a HIV-positive man in the past 6 months and 6.5% had had more than one intimate partner in the past 12 months. Moreover, almost 3% reported a diagnosis of STI or exhibited STI symptom(s) in the past 6 months. About 4 in 5 women had been tested for HIV, with 2.5% of women reporting HIV-positive status. 20.9% indicated that they did not know their HIV status. Of the total sample, 1.3% of women reported intravenous drug use in the past 6 months.

Intimate partner violence

As shown in table 3, almost half (49.6%) of the women reported a history of any form (minor and severe) of physical, injurious and/or sexual IPV, while 38.7% experienced severe physical, injurious and/or sexual IPV. Furthermore, 11.8% of the women reported any form of physical, injurious and/or sexual IPV in the past 6 months and 8% experienced severe physical, injurious and/or sexual IPV in the past 6 months. With respect to sexual IPV lifetime prevalence rates, about one-fifth (22%) of the sample reported minor sexual IPV and 15% experienced severe sexual IPV (ie, rape). In addition, 5.6% of the women stated that they had been forced to have sex by an intimate partner in the past 6 months, and 2.4% reported that they were raped by an intimate partner in the past 6 months.

Association between HIV risk and experiencing IPV

Several HIV risk behaviours were significantly associated with experiencing physical and/or injurious IPV, sexual IPV and all types of IPV in the past 6 months (table 4). Women who reported engaging in sex with an HIV-infected partner or an IDU in the past 6 months were significantly more likely to have experienced any form of physical/injurious IPV, severe physical/injurious IPV (adjusted OR = 3.1 and 4, respectively) and any form of sexual IPV (adjusted OR = 3.6) in the past 6 months than women who did not have sex with such partners. Furthermore, women who reported having had more than one intimate partner in the past 12 months were significantly more likely to indicate experiencing any form of physical/injurious IPV and severe physical/injurious IPV in the past 6 months (adjusted OR = 5.2 and 4.7, respectively) as well as any form and severe form of sexual IPV in the past 6 months (adjusted OR = 6.5 and 9.9, respectively) than women who did not have multiple partners. In addition, women who reported injecting drugs in the past 6 months were significantly more likely to experience any form of physical/injurious IPV and severe physical/injurious IPV (adjusted OR = 4.5 and 5.6, respectively) and any form of sexual IPV and severe sexual IPV (adjusted OR = 6.7 and 8.7, respectively) in the past 6 months than those who had not.

Furthermore, one of the socioeconomic status indices, homelessness, was found to be a significant confounding variable in almost all of the multivariate models (except in 2 out of 28 models).

DISCUSSION

To our knowledge, this is the first study to examine the associations between HIV risk behaviours and experiencing IPV among a random sample of predominantly low income, ethnic minority women receiving care at an urban ED. Consistent with rates found in other studies,^1,15–17 the findings show that a substantial number of female patients in the ED reported experiencing sexual, physical and/or injurious IPV both in the past 6 months and over their lifetime. These IPV rates underscore the severity of the problem among female patients in the ED, in particular the high rates (15%) of severe sexual coercion (rape). Women who are sexually abused clearly are at high risk for HIV transmission because safe sex and use of condoms are not practised during forced sex. The finding that the majority of the sample was sexually active, but that less than one-fifth always used condoms also highlights these women’s significant risk for HIV/STI transmission. Consistent with the findings of other studies in ED settings, HIV seropositivity was high in our sample of women.^14,22,23

Several HIV risk behaviours were significantly associated with experiencing IPV, including: having had sex with a HIV-positive partner or an IDU, having had sex with more than one partner and injection drug use. Dovetailing with several previous studies,^{5,10,25,35,36} this study found that women seen in an inner city ED who have had sex with a HIV positive partner or an IDU have an increased likelihood of experiencing IPV. Recent research findings have indicated that the relationship stress created over a partner’s injection drug use or HIV status may escalate into IPV.^29,30 Moreover, the significant relationship between having multiple sexual relationships and experiencing IPV, a finding which is consistent with previous studies,^7,10,12 suggests that a partner’s perception of a woman’s additional sexual affairs may trigger IPV, or, conversely, that a woman’s experience of IPV may lead her to engage in outside relationships. Finally, the significant relationship between a woman’s IDU and experiencing any type of IPV and any severe IPV is consistent with previous studies that propose that gender-based inequalities often pervade the practice of injecting drugs.^1,4,5,37,38 According to our research, women are forced to use and in some cases inject drugs with their male partner for several reasons: (1) to obtain financial support from him for household expenses, (2) to obtain protection against other drug users in her social network, (3) out of fear of losing the relationship and (4) for other financial and social dependencies.^4,5 This finding could also indicate that psychological distress and physical pain from IPV may lead women to inject drugs as a means to cope with IPV.

This study is limited by the following factors:

1. The non-response rate in this study limits the generaliseability of the study findings. For example, the exclusion of women who were admitted because of psychiatric emergencies, which has been found to be associated with a range of mental health problems may underestimate the rate of IPV.^39–41

2. The cross-sectional design of this study prohibits drawing conclusions about the causal relationships between HIV risks and IPV.

3. Research shows that IPV is associated with a number of health problems not adjusted for in the analysis.

4. Collecting sensitive, person-identifiable data and paying study participants may potentially bias results.

5. The study would have been strengthened by the use of a validated scale for socioeconomic status, rather than the use of education, employment and homelessness as markers of socioeconomic status.

Despite these limitations, the study findings underscore that IPV is a health-related risk factor that needs to be addressed as a public health problem by health professionals, in particular ED staff. Furthermore, our findings build on what are known clinical indicators of IPV (eg, multiple injuries, drunkenness, depression and post-traumatic stress disorder (PTSD))⁴² by identifying HIV risk behaviours as potential “abuse indicators” that should prompt inquiry into abuse, particularly when consultation time with patients is limited. While some researchers argue that there is insufficient empirical evidence to show that identification of abused women contributes to either reduction of IPV or improvement in quality of life,^43,44 today, many professional associations in the US and England^45–50 recommend routine inquiry for IPV for all patients who visit EDs and other healthcare settings—a recommendation also supported by epidemiological data.^51–55 Research has also recognised a number of challenges that include time constraints, professional discomfort with raising the topic, the patient’s fear of disclosure, staff attitudes toward women who experience IPV, lack of staff training on the topic and lack of evidence on effective IPV interventions that can be used to help women. These challenges, while significant, should not obviate the goal of routine inquiry for IPV, but should lead to identifying effective ways to safely conduct inquiry, and to develop more effective interventions to address IPV.^50,56–59 HIV testing and IPV routine inquiry for women in ED settings, in particular for women who disclose IPV or report HIV risk behaviours or positive HIV serostatus, would offer an important window of opportunity to initiate treatment and referral for services for women affected by these overlapping epidemics.