Background
Intimate partner violence (IPV) is a serious social issue in Japan. A national survey conducted in 2008 revealed that approximately 25% of adult women, an estimated eight million women, reported being subjected to physical violence by a male partner, over their lifetime: approximately 16% reported sexual violence, and approximately 17% reported psychological abuse [
1]. This survey revealed that the 5-year prevalence of IPV was 14% among women. Furthermore the number of women experiencing physical, sexual or psychological violence from a partner has increased gradually, approximately 19% to 33%, over the ten-year period of 1998-2008 [
1,
2].
The law for the prevention of spousal violence and the protection of victims, enacted in Japan in 2001, called for the identification of abused women and for providing adequate resources such as support centres or hotlines. The language used in the law indicates that health care providers may report IPV to the police and support centres, taking into consideration the intention of the victim, however reporting is not mandated. Providing women with access to counselling centres is a clear intent of the law. Despite the law, most health care providers are not adequately trained to carry out this role [
3], nor is abuse screening, or even inquiry, routine in the healthcare settings, therefore safety planning and referral to community resources are not provided. It is also likely that many health professionals do not recognize most incidents of IPV and therefore do not contribute to the woman's health recovery.
According to a recent survey in Japan, the prevalence of IPV during pregnancy was around 5% [
4]. IPV affects women and infant's health and pregnancy outcomes: gynecological symptoms [
5], sexually transmitted diseases, abortion, and complications of pregnancy: diabetes, hypertension, infection, anemia [
6‐
8] and low weight infants [
9]. Pregnancy is one of the few opportunities where healthy women come into contact with their own health care providers. Therefore, IPV screening and assessment in prenatal settings should be standard care for all pregnant women. Although there is insufficient evidence to support the effectiveness of interventions after universal screening [
10,
11], screening, carefully orchestrated early intervention attempts during pregnancy for IPV are still imperative because of the high risk of substantial damage to the health of women and children. In order to start effective interventions for abused women in Japan, the appropriate timing and method of IPV screening in healthcare settings needs to be better understood.
Even though researchers developed valid and reliable screening instruments [
12], their effective implementation is still controversial. Researchers have conducted some randomised controlled trials (RCT) to compare screening methods. MacMillan et al. examined optimal screening methods (computer, written, face-to-face) for IPV in emergency departments, family practice and women's health clinics. They concluded that no statistically significant differences were found for screening methods, however, women preferred the self-administered method [
13]. Another RCT in family practice also identified no statistically significant difference between interviews and self-administered questionnaires [
14]. Of the two RCTs conducted in emergency departments, one demonstrated that computerized screening increased the identification rate of abused women compared to a questionnaire [
15]; the other found that an audiotape questionnaire detected more abused women than a questionnaire [
16].
All RCTs were conducted in North America; no studies were undertaken in Asian countries or in prenatal settings. The issue of cultural sensitivity remains unexplored. It is not known whether approaches used in North American are relevant in Asian countries. While the names of the cultural values may be similar, such as respect, trust and caring, the order and meaning of the values and the way they are enacted may differ, thus requiring different approaches in screening for domestic violence. Therefore we conducted a randomised controlled trial in a prenatal clinic in Japan to compare the rate of IPV disclosure based on face-to-face interviews and a self-administered questionnaire and to identify responses to screening experiences.
Methods
This was a parallel group, randomized controlled treatment study. The study took place from February through November 2003 at the prenatal clinic of a general hospital, in a typical urban area of Tokyo, which had approximately 1500 deliveries each year.
Participants
Eligible pregnant women were consecutively selected between February 22 and May 30, 2003. At the time of recruitment most were at 14-15 weeks of pregnancy, although a few were close to 25 weeks, which was the cut-off point for study eligibility. Research assistants, who had master's degrees in nursing, or the researcher (YK) invited eligible women to privately discuss their participation in the study. They were informed in detail (following the Declaration of Helsinki) regarding their right to confidentiality, to withdraw from the study at any time without jeopardizing their care and protection of anonymity if results were published or presented. The women were asked if they were Japanese speaking, had any serious mental diseases and if their partner would agree to leave the room during the interview or while completing the questionnaire. Children under two years old were allowed to stay with their mother, but older children, who might report the discussion about IPV to the male partner or family members [
17], were not allowed to stay and instead were provided with toys and a separate place to play. All participants gave written informed consent.
Randomization
After informed consent, the participants were randomised, by means of numbered sealed envelopes, to either the interview or the questionnaire group. We used a random number table in blocks of four to ensure that approximately equal numbers of women were allocated to each group. Because of the nature of the screening methods participants could not be blinded to the group assignment. The same researcher performed the allocation procedure and data analysis.
Instruments
Two methods of implementing the screening tool, Violence Against Women Screen (VAWS), were compared in this trial: self-administered and face-to-face interview. VAWS is a 7-item Japanese screening instrument. Items include information about: couple's relationship, perceived behavior of the partner (i.e. difficultly settling arguments by talking it through; becoming frightened, being yelled at, partner hitting the wall or throwing objects; being forced into sex; being pushed or slapped; being hit or kicked in that order). Item responses are on a 3-point Likert scale:
often (2),
sometimes (1) or
never (0). The scores range from 0 to 14 and a score of 2 or above indicates IPV. The sensitivity of VAWS is 86.7% and specificity is 80.2%, based on the Japanese version of the index of spouse abuse [
18].
Secondary outcomes were comfort level and the need to consult with the nurse after screening, which were used as indicators of possible adverse effects of IPV screening. The comfort level question referred to their sense of ease or discomfort. We asked women who had used either method: self-administered questionnaire or interview: "How did you feel being questioned about your relationship with your partner". They indicated their comfort level using a 4-point Likert rating: 1 (not at all comfortable) to 4 (very comfortable). In order to measure their desire for consultation about their relationship, we asked women: "Would you like to consult with a nurse when you have a problem with your husband?" The response options were: yes, no or cannot decide.
Prior to administrating the first screening, all participants responded to a written questionnaire covering demographic information such as age, education, employment, marital status, alcohol and smoking habits before pregnancy, and lifetime experience of physical violence from a former partner.
Screening and consultation protocol
All participants received three screenings: the first screening was upon entry to the study at 14-25 weeks of pregnancy; the second screening was at 20-30 weeks and the third screening was at 35 weeks or more. In the first screening, women reported about abuse for the "past 12 months", but at screenings two and three, they reported about abuse: "since the last time you were interviewed" [
19]. Three assessments were made because disclosure about abuse could occur at any time during pregnancy especially after establishing a trusting relationship with a nurse.
The interviewers were three female nurses working at the hospital outpatient department. The researcher provided three training sessions in order to assure interviews proceeded in accordance with the study protocol and with an awareness of IPV as an issue for women. The first training session was a 2-hour lecture about the prevalence, dynamics and the physical and psychological impact of IPV, and appropriate care for abused women. A survivor of IPV who told her story of recovery conducted the second session. Theories (i.e. cycle of violence and learned helplessness) were discussed to explain why she did not escape from her partner. In the third session, the researcher trained the nurses on the use of the protocol for screening and intervention for abused women in the prenatal setting.
Interviews were conducted at the prenatal clinic in a private room that was partitioned into three cubicles or in a private area. The nurse sought to establish a partnership with the woman, based on the women centered care model [
20] of respect and empowerment. Half of the women were interviewed by the same nurse at all three screenings. The nurse read the VAWS questions to the woman, while the woman also read the questionnaire.
If the woman screened positive and wanted the nurse to report the abuse to a support centre or to the police, the nurse would make the report. In addition, the nurse discussed with the woman how to maintain her safety and attempted to aid the woman in decision-making. The nurse also gave the woman a community resource card with information about a crisis hotline and local shelters. Women in the questionnaire group completed the VAWS in the official prenatal clinic interview room where the same type of resource cards were available. The resource cards were also available in the waiting area. The abuse consultants' contact information was listed in the questionnaire for easy access. Thus, although all the women had access to IPV resources the women in the interview group may have received additional assistance compared to the women in the questionnaire group.
Data Collection
The primary outcome was the disclosure rate of abuse by women who were identified by the VAWS tool during the three screening sessions compared to those in the interview group. We counted single disclosures only once, at the first or second or third screenings. We also included the pattern of multiple disclosures at screenings: first and second; first and third; second and third and first, second and third. This provided us with unduplicated accounts of abuse disclosures.
Sample Size and Statistical Analysis
From the survey, we estimated the baseline proportion of physical violence cases during pregnancy to be 3% [
4], so we aimed to recruit 140 women per group in order to detect a difference between groups at a 7% effect size, based on previous research [
19,
21] with 80% power. Extra women were recruited to account for attrition, which was expected to be 10-20%.
All analyses were carried out using SPSS for Windows 11.5 J (Chicago, IL, USA). Categorical differences between the two groups were tested with the χ2 test or Fisher's exact test. Multiple logistic regression analysis was performed to adjust for potential baseline differences in the characteristics between the groups. For each group we calculated the IPV positive screen rate using a 95% confidential interval (CI). All statistical tests were performed with a two-sided 5% level of significance. All analyses were done using an intention-to-treat basis, which included all randomized participants. The Ethics Committee of St.Luke's College of Nursing approved the protocol.
Acknowledgements
This study was supported by Grant-in-Aid for COE Research, provided by the Ministry of Education, Culture, Sports, Science and Technology of Japan. The authors thank Caroline White RN PhD and Sarah Porter RN PhD who provided English editing services.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YK carried out the data collection and drafted the manuscript. HE participated in the sequence alignment. YY participated in the design of the study and performed the statistical analysis. SH conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.