Intimate partner violence (IPV) against women is considered an important public health problem.^1,2 In a review of 48 population based surveys carried out worldwide, between 10% and 69% of women reported having been physically assaulted by their partner at some time in their lives.³ The literature shows that physical IPV is often accompanied by psychological and sexual violence.^4,5 However, prevalence studies have traditionally placed more emphasis on physical and sexual abuse than on psychological abuse, and therefore our knowledge about this type of violence is limited.^6–9 In Spain, a population based survey carried out in 2002 by the Women’s Institute, found that 11.1% of surveyed women were abused (physically, psychologically, or/and sexually) by their male partner.¹⁰

Qualitative studies have shown that most abused women try a variety of strategies to overcome an abusive relationship.¹¹ Women generally find three alternatives: seeking outside help, taking legal action against their abusers, or ending the violent relationship. Until very recently these options were socially repressed, as IPV was considered a phenomenon that had to be dealt with privately within the family. In recent years however, abused women have been encouraged to report the violence and, in fact, IPV cases reported to the police in Spain have increased by 32.8% from 2002 to 2004.¹² However, data also show an even greater rise in the number of female homicides in recent years.¹³ The new situation therefore, demands legal, social, and financial resources to provide victims with adequate protection.¹⁴

The factors conditioning abused women to remain in an abusive relationship have been widely studied. In 1988, Strube suggested that women who remain with their abusers experience less severe abuse, are more likely to be financially dependent on their male partners, and are less likely to report that their children have been abused.¹⁵ More recently, it has also been shown that age and severity of abuse seem to be good predictors when it comes to seeking formal help.¹⁶ Moreover, the fear of punishment, the lack of financial resources, the concern for the children, emotional dependence, the lack of support from family and friends, and the constant hope that the partner will change, have been identified as barriers when leaving abusive relationships.^17–21 Another point worth considering is that the abused woman’s option to leave this type of situation is often conditioned by factors that go beyond her control, such as the community’s attitude towards IPV or available resources for battered women.^22,23

The purpose of this study was to: (1) describe the different responses adopted by women in Spain who are victims of IPV; (2) identify the different sociodemographic characteristics of the abused women who presented each type of response; (3) identify the factors contributing to adopting a response, and (4) study the association between the different types of response and the different types of IPV (physical, psychological, and sexual).

METHODS

Participants and setting

In this cross sectional study participants were recruited by using a convenience sample of 23 volunteer general practices in three regions of Spain (Andalusia, Madrid, and Valencia). All female patients aged 18 to 65 years and seeking medical care in these practices from May to October 2003 were eligible for the study. Women were considered non-eligible if they were illiterate, did not understand Spanish, or had severe cognitive disabilities. Following the Ethical and Safety Recommendations for Research on Domestic Violence Against Women, women who attended the practice with a male partner were excluded.²⁴

Each general practitioner randomly recruited a maximum of two women a day. All practitioners were to include the eligible female patient from those who entered the surgery at two previously set times; one at the beginning of the shift and another one at the end. If the woman selected did not meet the inclusion criteria, the following one was selected. The rationale for choosing this recruitment process was that it caused minimum inconvenience to the practitioners and guaranteed a random selection of women.

Survey measures and instruments

A self administered structured questionnaire was specifically developed for this study. It consisted of 21 closed-ended questions on IPV, sociodemographic variables, and health status, taking 15 minutes to complete. Measures not considered in this study have been described elsewhere.^25–27

The outcome measure was women’s response to IPV: none, partner separation, reporting the case to the police, seeking help from healthcare professionals, and seeking help from associations for battered women. Given that women usually respond to IPV with more than one response, four mutually exclusive categories were created. The first three categories included women who had taken one action*: partner separation, reporting the case to the police, and seeking help (from healthcare professionals and/or from associations for battered women), respectively. The fourth category included women who had taken two or three of the previously mentioned actions. A woman was considered to have “taken action to overcome IPV” if she gave an affirmative answer to at least one of the options and considered “not taken action to overcome IPV” if she answered negatively to all of the options.

The independent variables considered in this study were:

• Sociodemographics: age, number of children, marital status, nationality, employment status, education, monthly family income, and social support.

• Variables related to IPV:

  • Intensity: many times, sometimes, and never.

  • Combination of the types of lifetime abuse: physical, psychological, and sexual.

  • Duration of the abuse: <1 year, 1–5 years, and >5 years.

The IPV questions were taken from a questionnaire used in previous studies,²⁸ showing high comprehensibility and acceptability. They were adapted from scales used in other studies, such as the World Health Organisation multi-country study on women’s health and life events.^29–31

The woman was asked if her current intimate partner or any other partner in the past had abused her physically (hit, slapped, kicked, pushed…), psychologically (threatened, insulted, humiliated, been extremely jealous, scared her…), and/or sexually (forced her to have sexual activities against her will). These questions had three possible responses: “many times”, “sometimes”, and “never”. A woman was considered to have experienced IPV if she answered “many times” or “sometimes” to any of the questions in the three IPV categories.

Because of the considerable overlap that usually exists between IPV types, four mutually exclusive categories of lifetime abuse for the analysis of IPV and women’s response to violence were created. The first group included respondents who had only experienced psychological IPV. The second group included those who had experienced physical and psychological IPV but not sexual. The third group included those who had experienced psychological and sexual IPV. The fourth group included those who had experienced the three types of abuse. Other additional groups as categories of abuse (physical solely, sexual solely, and physical and sexual) were not included as the number of women in these categories (nine, three, and none, respectively) did not give permission for analyses to be carried out. Additionally, 18 women who reported having experienced IPV did not respond to all IPV questions and could not be placed under any IPV category.

Procedure

The information was gathered by the family practitioner at the end of the woman’s consultation. If the patient met the eligibility criteria, she was invited to participate in a study on women and health. After giving her consent, she was handed the questionnaire in a sealed envelope. The practitioner explained that the questionnaire was anonymous and confidential and the woman was offered help in filling out the questionnaire if necessary. Once completed, the woman deposited the questionnaire in a box prepared for this purpose. In the envelope, the woman received information on available community resources for battered women in the area.

Data analysis

An initial descriptive analysis was first conducted to describe the characteristics of the sample and the actions taken by abused women to overcome the violence. The association between the outcome measure and each of the independent variables was analysed using the χ² test for qualitative independent variables and the Student’s t test for quantitative ones. Significance was set at p<0.05. The magnitude of the association was estimated using the odds ratio (OR), with a confidence interval (CI) of 95%. Given the reduced number of women for some analyses, yielding very imprecise confidence intervals, non-mutually exclusive categories of responses to IPV were also used, specifically for the study of the association between the different categories of IPV and the different responses to IPV. Finally, a multivariate logistic regression analysis was conducted for the joint control of possible confounding factors. Included in the model were all the significant variables in the bivariate analysis and all those considered to be of interest for the study. It should be noted that some of the variables were recoded for the bivariate and multivariate analysis.

RESULTS

Overall, 1631 women were asked to participate. One hundred and eighteen refused and 72 had missing data on several response variables (88.35% response rate). We excluded 26 participants who had never been in an intimate relationship and 13 who did not meet the age inclusion criteria. Therefore, 1402 women were included in the analysis.

The average age of the total sample was 39 and the average number of children per woman was 1.5. Two thirds of women were married and around half of the participants were employed (51%). Forty one per cent had not completed secondary education, 34.6% had a university degree, and 49% had an income level between 600 and 1200 euros. Most of the sample reported having some kind of social support (94.2%).

As table 1 shows, almost one third of surveyed women reported having experienced some kind of lifetime IPV (n = 445). Sixty three per cent of abused women reported having taken some kind of action to overcome IPV, where separation from the partner was the most frequent action (82.6%).

The sociodemographic and IPV related characteristics of the 445 abused women, together with the association of these characteristics with the response to IPV, are presented in table 2. Single women and those who were separated/divorced/widowed, presented a significantly higher probability of taking an action to overcome IPV, compared with married women (OR = 1.92; 95% CI = 1.14 to 3.22 and OR = 4.38; 95% CI = 2.43 to 7.88, respectively). Moreover, those women who were employed were more likely to try to solve the situation than housewives (OR = 1.69; 95% CI = 1.03 to 2.79), as well as those who had a university degree, compared with those who had no education or primary education (OR = 1.90; 95% CI = 1.15 to 3.16). Women who reported having social support were also more likely to take an action (OR = 2.31; 95% CI = 1.28 to 4.15). The probability of responding to violence was greater among those women who had experienced abuse “many times”, compared with those who had experienced it “sometimes” (OR = 2.54; 95% CI = 1.60 to 4.02). Women who experienced physical and psychological abuse and those who experienced the three types of abuse also presented a greater probability of taking an action than those who experienced psychological abuse alone.

Table 3 shows the sociodemographic characteristics of the 224 abused women who took some kind of action to overcome IPV, differentiating between the various kinds of mutually exclusive actions. The average age of women who separated from their partner was 37 and the average number of children was 1.4. Sixty seven per cent were employed, and most of them had secondary or university education and a medium-high income level. Those who reported the case to the police were older, with more children, mostly separated/divorced/widowed (53.3%), housewives (46.7%), and with a lower education and family income level. Those who sought outside help were mostly married (72.7%), with an average age of 45, housewives (52.4%), and with a basic level of education. Those who took more than one action were, mainly, separated/divorced/widowed, employed (59%), and with a medium income level.

Table 4 shows the association between the different categories of IPV and the different responses to violence. Women who experienced physical and psychological abuse had the highest probability of separating and seeking help from healthcare professionals or associations for battered women. Women who experienced psychological and sexual IPV (with or without physical abuse), however, were the most likely to report the case to the police (OR = 8.66; 95% CI = 2.73 to 27.53). After adjusting for age, marital status, social support, and intensity of abuse, the association remained significant for separation and reporting the abuse to the police. Women who experienced physical and psychological abuse presented the highest probability of separation (OR = 2.28; 95% CI = 1.18 to 4.39) and reporting the case to the police (OR = 4.26; 95% CI = 1.10 to 16.48).

Finally, the independent factors associated with taking action to overcome IPV were being separated/divorced/widowed, having social support, having experienced physical and psychological abuse, and having experienced it frequently (Table 5).

DISCUSSION

Given its high prevalence, gender based violence has become a widely studied public health problem worldwide. However, the abused women’s response to IPV is an aspect that has hardly been addressed in the literature and, for this reason, our study aimed to contribute to the exploration of this issue.

The study has certain limitations that should be taken into account. This is a cross sectional survey and inferences regarding causality cannot be made. Moreover, this study has the usual bias of self report. It should also be noted the small sample size for some of the categories of analysis. Furthermore, women who attended the practice with their partners were excluded, and it could be that these women were more likely to be controlled or abused by their partners than women who attended the practice alone. Additionally, despite the association found in previous studies between IPV and low socioeconomic levels, given that participants in our study were asked to complete a written questionnaire, illiterate women were excluded. All these limitations however, could have only led to an underestimation of the real IPV associations.

In our sample, one third of women reported having experienced some kind of lifetime abuse. This result is consistent with findings from other international studies conducted in the healthcare setting.^32–36 Almost two thirds of abused women took an action to overcome IPV, mainly separation from the partner. This finding is also consistent with results from a previous Spanish study.³⁷ A longitudinal study carried out with couples in the United States also showed that domestic violence, besides the use of alcohol, was a good predictor of the separation of the couple.³⁸

In this study, those abused women who separated from their violent partners were, in general, young, single, comparatively independent, and with a medium-high educational and income level. Not being married to the abuser could make things easier for the victim who decides to leave an abusive relationship. Separation usually implies a long legal process, and it is often during this process when the most severe aggressions and violent episodes occur. In addition, a high educational and income level can be a guarantee of financial stability for the victim after she has decided to live without the income provided by the abusive partner.

Compared with women who separate from their partners, those who decide to seek outside help or report the case to the police, present different sociodemographic characteristics. It might be possible that these women, older and with more children, have to face more barriers before abandoning the man who normally provides the family income. They are usually unemployed women with a basic level of education, and this could constitute a very difficult barrier to overcome to have a successful independent life. This finding highlights the need to strengthen community support for those women who see that their financial and social stability will be endangered if they separate from their abuser.

Women with the highest income were more likely to seek outside help, while those with the lowest income were more likely to report the case to the police. This result shows the crucial part that service providers and legal institutions can play in enabling low income abused women take an action to overcome violence.

Only 7% of abused women who took action to overcome IPV sought help from associations for battered women. However, in recent years there has been an increase in the number of governmental and non-governmental services for battered women in Spain.³⁹ The results of this study seem to suggest that the use of these services is far below the expectations, perhaps because of the women’s lack of knowledge about the help they can provide, how to approach them, or lack of trust in their effectiveness. National media campaigns against IPV could include information on these services, to help abused women receive the appropriate support.

Social support is also a determining factor in women’s response to violence. While women who reported having support from their family or friends were more likely to separate or seek outside help, those who did not have social support were the most likely to report the case to the police, probably to receive the protection they do not find in their families and friends from the authorities.

The IPV characteristics that have been associated with taking action to overcome violence are the type of abuse and its intensity. Results seem to suggest that it is the physical component of IPV and the most intense abuse that makes women become aware of their abusive situation. This has also been seen in other studies, which show that women separate permanently from their partner when the violence becomes serious enough for them to realise that their partner is not going to change, or when the situation starts to have considerable negative consequences on their children. Similarly, this often happens when there is protection from family and friends, who offer the victim emotional and logistic support.^40–43

Identifying the factors that have an influence on the response adopted by abused women allows us to better understand the support needed by them to abandon an abusive relationship. The findings of this study highlight the need for political, social, legal, and health services to work together to provide abused women with the support and courage to overcome a violent relationship. This could also contribute to the development of policies and programmes that can become decisive in assisting abused women in their efforts to end with their situation of abuse.