Factors Associated with Disclosure of Intimate Partner Violence among Women in Lagos, Nigeria

Abstract: Background: Though the prevalence of Intimate Partner Violence (IPV) remains high in less developed countries, data suggest that these figures may represent an underestimation considering that many women are unwilling to disclose abuse. This paper aims to determine women's willingness to report abuse, factors determining willingness to disclose IPV, and to whom such disclosure is made. Methods: A total of 911 women visiting reproductive health facility responded to the questionnaire, and the collected data was analyzed using multivariate analysis. Results: About 54% (n=443) of the participating women reported that would not disclose IPV. Among those willing to disclose abuse, 68% (n=221) would opt to disclose to close relatives in contrast to 32% (n=103) who would disclose to some form of institutions (i.e. religious leaders, law enforcement officers). Ethnicity, woman's own use of alcohol and autonomy in decision making such as having a say on household purchases, money use and visitation, independently predicted willingness to disclose IPV. conclusions: The role of family is still important in the Nigeria context and the implications for research and intervention are discussed.


Original Article
Factors Associated with Disclosure of Intimate Partner Violence among Women in Lagos, Nigeria Leah Okenwa,P a P Stephen Lawoko,P a,* P Bjarne Jansson P a Introduction espite the inaction of laws and regulations against Intimate Partner Violence (IPV), the prevalence of IPV remains alarmingly high. Globally, the one-year prevalence ranges between 15 -71%,P 1,2 P with variations depending on factors such cultural norm, laws and other local conditions that favor/disfavor gender inequity. Yet, these figures may represent an under-estimation considering that significant proportions of women are unwilling to disclose abuseP 3,4 P Disclosure of abuse is a vital step in the process of finding a lasting solution and breaking the abuse chain. Thus, unless victims are willing to disclose abuse and make use of available resources, screening for and eventual management of IPV may be heavily constrained.
Intimate Partner Violence is defined as a pattern of assaultive and coercive behaviors, including physical, sexual and psychological attacks, as well as economic coercion that adults or adolescents use against their intimate partners.P 5 P Though women believe that screening gives victims support and information they need, they acknowledge that they have never disclosed disclose abuse in health care setting P 6,7,8 P The reasons for this discrepancy remain unclear but a likely explanation is that healthcare providers may lack adequate skills in promoting disclosure of abuse.P 8 P In family planning and antenatal clinics three categories of women have been identified; women who will disclose abuse or fear of it; women who will not openly disclose abuse, but present with abuse-related physical symptoms (e.g. bruises) as well as reproductive health complications (e.g. lacerations and history

Injury & Violence 38
journal homepage : http://www.jivresearch.org of unexplained pregnancy complications); and finally women who live in an abusive relationship but do not report or show any signs associated with the abuse.P 9 P These distinctions raise questions as to why some women disclose abuse while others do not.
A number of factors both at the organizational, household and individual level have been identified to explain why women may choose not to disclose abuse. Within the clinical setting constraints to disclosure reported by women interviewed, are the perceptions that clinicians lack interest in IPV and a lack of trust in the health care provider. This is further compounded by threats of more violence in retaliation from the partner and embarrassment.P 10 P At the household level, women refrain from reporting abuse depending on the economic alternatives they have in the event of having to leave an abusive relationship. Lack of alternative economic resources may prevent women from reporting abuse.P 11 P Moreover, potential for child abuse may hinder women from reporting abuse,P 12,13 P suggesting that disclosure of abuse may fuel the intergenerational circle of violence from intimate partner violence to child abuse. At the individual level, there is evidence that women's ability and willingness to disclose abuse is influenced by; her emotional strengths, her level of adherence to gender roles, decision-making autonomy, being asked about it, social empowerment and her perception of available social supportP 8,10,14 P It is however likely that these individual level factors are fueled by gender and social inequities at the societal level. Corroborating this argument, it is noted that factors such as ethnicity, culture, gender role definitions, kin and friendship networks may influence a woman's perception of her options, the help she seeks, as well as the nature and scope of violence she experiences in an intimate relationship.P 15 P Ethnicity and culture on their part have significant impact on women's attitude to IPV such that an ethnic group that is more gender restrictive is more likely to condition women to agree or consent to wife beating.P 16 P Normalization of IPV plays out significantly in Sub-Saharan African context. Recent data suggest that over 75% of the women believed that wife beating was justified when a woman does not leave up to her traditional normative roles (e.g. cooking and taking care of children).P 2,17,18 Societal, cultural and religious factors are not only important in determining whether women will report abuse or not, but also to whom such abuse will be reported. In many parts of Sub-Sahara Africa, marriage is considered a family and community affair rather than a private one. The role of the extended family therefore includes arbitrating in marital conflicts and finding ways to resolve them. Disclosure of abuse to some institutions such as law enforcement agencies is viewed as disrespect for the family. Indeed, authorities such as the police themselves condone such activity as women who dare to report are usually advised to go and settle with their husbands, denying women the opportunity to press charges and ultimately reducing their interest in seeking justice.P 11,19,20 P Despite evidence that the major religions practiced in Nigeria i.e. Christianity, Islam and traditional religion all have teachings of female submission and obedience to the man as the head, findings reveal that some women are willing to disclose to religious leaders.P 11 P However, distinctions between the categories of women who would make such reports are not yet clear.
Few studies coming from the African context have systematically examined the extent, nature and determinants of IPV disclosure. The objective of this study is: 1) to determine how willing women visiting an out-patient clinic in Lagos, Nigeria are to disclose abuse; 2) to whom such disclosure would be made; and 3) to determine factors which influence both willingness to disclose and the choice of to whom disclosure is made.

Study design and setting
This cross-sectional study was conducted on woman attending out-patient clinic of the Obstetrics and Gynecology department of the Lagos University Teaching Hospital (LUTH), Nigeria. The hospital, as its name implies, is a teaching hospital affiliated with the Lagos University which is one of the oldest and largest institution in Nigeria. The hospital is a feepaying federal government owned tertiary institution known for conducting quality research.

Sampling procedure and participants
A convenient sample of 934 women aged 15-49 years was obtained while they were visiting the Obstetrics and Gynecology clinic of LUTH. Women were included in the study if they were 15-49 years of age. The sample size needed for the study was established using a power analysis, assuming a binomial distribution. To obtain a statistical power over 90% which is considered as very good, a sample size of about 900 was necessary based on a statistical significance level of alpha=0.05, and an estimated average yearly probability of IPV occurring in developing countries of 0.125.P 1 P Each participant responded to a questionnaire comprising of previously validated questions under the guidance of trained personal.

Questionnaire
A structured questionnaire covering demographic and health issues was administered verbally to the eligible women by trained field workers and voluntary participation emphasized. Study questionnaire was adapted from those commonly used by the World health organization and the demographic and health surveys and translated into the three major Nigerian languages, i.e. Hausa, Igbo and Yoruba. It was later back translated for validity. It covered women and their spouse's background (such as education, use of alcohol), their reproductive history, utility of family planning methods, fertility preferences, child mortality, awareness of and precaution against sexually transmitted diseases, marriage and sexual behaviour, attitudes towards IPV, disclosure of IPV, psychosocial health outcomes, demographic, social and empowerment indicators as well as exposure to domestic violence. For the current paper, the questions of primary interest were those on attitudes towards IPV, exposure to IPV, demographic, social and empowerment indicators and willingness to disclose IPV.

Dependent variable
Disclosure of IPV: participants responded to a hypothetical question posed as "would you disclose abuse?" The response options were "yes", "no", "don't know". Those responding "yes" to this question were probed further to find out to whom they would report. The response options to this follow-up question were: woman's family only, husband's family and close friends only, both families, religious leaders, the police and other types of social institutions. Subsequently, responses to this questions were dichotomized into two broader categories: 1) families and close friends (comprising those who would report to the woman's family only, husband's family and close friends only and those who would report to both families); and 2) the institutions (comprising of religious leaders, the police and other types of social institutions).

Independent variables
Attitudes to IPV was assessed using commonly used questions assessing IPV attitudes in the African context.P 20, 21, 22 P The questions assess whether participants would justify wife beating in five hypothetical situations: if the wife goes out with another man, neglects the children, argues with her partner, refuses to have sex with partner or cooks bad food/or food is served late. Answer options were "yes", "no" or "don't know". An affirmative response to one or several of these questions was considered having a tolerant attitude towards IPV, while a "no" response on all five situations denoted a non tolerant attitude.
Exposure to IPV was assessed using a modified version of the Conflict Tactic Scale (CTS).P 23 P Physical abuse was operationalized as being slapped, pushed, punched, choked, burnt on purpose, kicked and assaulted using knife or other weapons. Psychological abuse included being insulted, made to feel bad about self, belittled in front of other people, done things to scare or intimidate, and threatened to hurt respondent or someone she cares about. Sexual abuse included being physically forced to have sexual intercourse when she did not want to; having intercourse out of fear or forced to do sexually degrading or humiliating act. In this study, a victim of IPV was a woman who has experienced at least one of the forms of abuse described above. Socio-demographic variables included: age; literacy (1=can read little or nothing, 2= can read whole sentences); religion (1=Catholic, 2=Protestant, 3=Muslim, 4=others); ethnicity (1=Yoruba, 2=Ibo, 3=others); earning income (1=yes, 2=no).

Empowerment indicators included
Access to information, assessed using frequency of reading newspaper, listening to radio, and watching TV all with response alternatives (1=almost everyday, 2=at least once weekly, 3=less than once weekly, 4=almost never/not at all); Decision autonomy, assessed by asking respondents whether they had a say on household expenditure, health care and household purchases with the following response options (1=complete say, 2=partial say, 3=no say). Women's and households economic position, assessed by inquiring whether the woman contributes to household purchase, whether the household has problems making ends meet, or problems managing monthly expenditures with the following response options (1=yes, 2= no).

Behavioural variables included
respondent's and partners use of alcohol and smoking habits (1=yes, 2=no) and polygamy.

Ethical considerations
National and local ethical clearance was granted by the Nigerian Institute of Medical Research, NIMR and the department of obstetrics and gynecology, LUTH before the questionnaire were administered. Ethical and safety recommendations set by the World Health Organization (WHO), which include training of and support to field workers, obtaining informed consent from participants, emphasis on voluntary participation as well as securing of anonymity were strictly followed.P 25 journal homepage : http://www.jivresearch.org Review Board of the Nigeria Institute Medical Research approved the procedures, methodology and questionnaire content.

Statistical analyses
Data from the questionnaire were first entered into Microsoft excel and later transferred to SPSS program version 15.0, where analysis was done. Chi-square test was used to assess associations between willingness to disclose IPV and the independent variables. The significance level was set at p<0.05 for all statistical analysis. Logistics regressions analyses were used in the multivariable analysis to assess the independent contribution of the explanatory variables while adjusting for possible confounding. The direction and magnitude of associations were expressed as odds ratio. The significance level was set at p<0.05 for all statistical analysis.

Willingness to disclose abuse and to whom
The majority of women in the study (54%) were unwilling to disclose IPV (Table 1). Of those willing to report abuse, barely 1% indicated willingness to report to the police compared with over 28% willing to report to the man's family, and 26% to religious leaders. In general, about 32% were willing to report to institutions contrasting with 68% willing to report to families and close friends (Table 1)."

Socio-demographic and behavioral factors vs. willingness to disclose abuse and to whom
As exhibited in Table 2, religion impacted significantly with willingness to disclose abuse with women of Catholic and "other" denomination most willing to disclose abuse (χP 2 P (3) = 7.9; p<0.05). Willingness to report abuse was more common among women who used alcohol (χP 2 P (1)= 16.5; p<0.001) and those whose partners used alcohol (χP 2 P (1)= 5.1; p<0.05) in contrast with their peers who did not or whose partners did not use alcohol ( Table 2).
A similar trend was observed for experience of violence in the latest year. No association was found between having tolerant attitude to IPV and willingness to report abuse or to whom abuse would be reported (Table4).

Independent predictors of willingness to disclose IPV:
As expressed by the odds ratios in Table 5, ethnicity, alcohol use and some measures of autonomy remained significantly associated with willingness to report IPV when journal homepage : http://www.jivresearch.org adjusted for possible confounding variables in the logistic regression. Ibo ethnic group was less willing to report IPV than other ethnic groups. Women using alcohol, women who had say on household purchases, and say on visiting friends/relatives were more willing to disclose IPV than their peers who did not use alcohol and had no say on household purchases or visiting friends/relatives. All other variables did not reach statistical significance when possible confounding was adjusted for.

Discussion
This study examined willingness to disclose IPV among women aged 15-49 years in Lagos, Nigeria and identified factors associated with such disclosure as well as preferences regarding to whom disclosure would be made. The results revealed that majority of the interviewed women (54%), would choose not to disclose IPV. This figure seems higher than those reported previously where between 37% and 42% choose not to disclose violence.P 3,26 P These studies however addressed actual disclosure of abuse among abused women while our study addressed willingness to disclose abuse even among women never abused. Considering that willingness to disclose abuse may not directly translate to actual disclosure on the event of abuse, the higher figure observed in our study may have been expected. Among women willing to disclose abuse, almost twice as many opted for disclosure to close relatives (68%) in contrasted with disclosure to the institutions (37%), where only a modest 1% were willing to disclose to the police. These findings are in agreement with other journal homepage : http://www.jivresearch.org  N=Number within category, n= number within category that is willing to disclose, % = N/n * 100 (i.e. proportion willing to disclose within category, p is the significance level for associations between independent variables and willingness to disclose constituting internal barriers. On the other hand, women refraining from disclosing IPV to the institutions could also be an indication that they lack trust in such institutions or that such institutions lack interest in domestic problems. Data from developed and other non-African context suggest that this may be the case.P 3,30,31 P Further researches are warranted to investigate institutional readiness to assist abused women within African culture in Nigeria.
A number of the demographic variables were significantly associated with willingness to disclose abuse. Catholic women were most willing to disclose abuse when compared with other denominations, though they were, together with Muslim women, less willing to disclose to the institutions when compared with Protestants. Though these findings add to the literature suggesting that ethnicity and religion may affect women's choices in terms of disclosure and acceptability of journal homepage : http://www.jivresearch.org they may also be suggestive that institutional readiness to assist abused women may vary depending on their religious and ethnic affiliations. Further research is warranted to test the later hypothesis.
Our findings show that after the ethnic Igbo women, ethnic Yoruba women were more likely than women from "other" ethnic groups to disclose to families, (although this did not reach statistical significance). The reason for this might be that It can thus be concluded that Yoruba women tend to report more to families bearing in mind their higher status as sisters and daughters.

Continue of
Women having some form of autonomy in household decisions (i.e. say on expenditure, purchases, number of children to have and visiting friends) were more willing to disclose abuse than their peers lacking such autonomy. These results were confirmed in the multivariable analysis. It is suggested that women's social and economic empowerment is likely to lessen her dependence on her partner.P 35 P This independence is often reflected in her ability to speak out. Our findings are in line with other studies indicating the role of education in the empowerment of women to denounce intimate partner violence.P 36,37,29 P Empowering factors such as education and access to information were also significant factors in our study regarding IPV disclosure. Women with little or no education preferred reporting to families and were less willing to disclose to institutions. A likely explanation is that education enlightens women on their options and thus empowering them to challenge traditional norms on gender inequality. Lack of access to information may also be another reason why women remain bound to tradition. Our findings seem to point in this direction as women without access to radio or television preferred to disclose to families more so than to institutions.
One of the factors influencing willingness to report IPV in our study is the experience of IPV in itself. Women who have experienced physical, psychological and sexual violence in general were more willing to report abuse when contrasted with non-abused peers, corroborating previous work where actual disclosure other than willingness to disclose have been studied.P 29,38 P These findings could not however be confirmed in the regressions analysis suggesting a possible confounding effect warranting further investigation. Contrary to our expectations, women with tolerant attitudes towards IPV in our study did not differ from their peers with intolerant attitudes to IPV regarding willingness to disclose. This appears contradictory to theories linking exposure to intimate partner violence with tolerant attitudes towards violence itself among women.P 39 P Capitalizing on these previous works, we had expected to observe higher willingness to disclose IPV among women with intolerant attitudes to IPV. Thus, the role of attitudes in disclosure of IPV deserves further investigation before firm conclusions can be drawn.
In practice, the implications for intervention/prevention program are enormous. The extended family remains a respected authority in resolving marital issues in the Nigerian culture. Prevention programs can capitalize on this by empowering the family unit by providing IPV related educational workshops, and improving their access to IPV prevention information, including information related to gender role issues. The importance of involving family in IPV prevention cannot be overemphasized. It is indeed suggested that lack of family support could be a barrier for victims of IPV, preventing them from taking steps towards ending their ordeal.P 11 P Lack of willingness of women to disclose IPV to the institutions also has important implications for training of law enforcement as well as religious leaders to become more proactive in handling and dealing with reports of IPV. Studies also point to the important role of health providers in screening for IPV and suggest that women are more likely to disclose IPV if probed by their health care providers.P 40, 41 P To the best of our knowledge, this is the first time that data on underlying factors determining women choice to disclose IPV to family/relatives or various institutions is being presented. However, more research is warranted to validate this finding. There are few limitations to this study that should be noted. The cross sectional design does not allow for causal interpretation of the results. It is also important to note that willingness to disclose abuse does not directly translate to actual disclosure on the event of abuse. Caution is therefore warranted in interpretation of our findings. Furthermore, this study was conducted in one site using convenient sampling journal homepage : http://www.jivresearch.org which limits the generalizability of the findings to other hospital settings or ethnic communities in Nigeria. Larger study is needed to assess determinants of IPV disclosure among women using a random sample that is representative of multiethnic, multicultural and multi-religious society like Nigeria. It is also important to add that though our study has identified a number of factors that may affect IPV disclosure, other prominent factors such as threat of increased violence in retaliation of a report have not been included in the analysis. Future research may need to incorporate such measures. Another limitation of the findings has to do with the lack of sample power to assess the independent role of spouse, family members, friends, and institutions, as a separate entity, in associations with IPV disclosures. Finally, the study sample was self-selected in that only women willing to participate were included until the required sample size was reached. Even though the interviewers reported that there were only a few women opting not to participate, the characteristics of these women remain unknown. Whether this non-response was systematic or not remains therefore unclear.