Intimate Partner Violence. The gynaecologist's perspective.

Intimate partner violence (IPV) is an important public health problem, which has been extensively studied all over the world, yet Belgian data are limited. IPV remains a taboo resulting in denial and underreporting. For an obstetrician-gynaecologist (OB/GYN), IPV, committed by a male partner to a woman, is of particular interest, because of its negative impact on women's and children's health. In Belgium there are few data on IPV and guidelines for OB/GYN are -missing. In a multi-centered survey surveillance study which was carried out among pregnant women attending 5 large hospitals in the province of East Flanders, the lifetime prevalence of IPV was estimated to be 10.1% and the period prevalence during pregnancy and/or in the year preceding pregnancy 3.4%. In our highly medicalised society, only 19.2% and 6.6% of the victims of physical and sexual abuse respectively sought medical care. Routine screening for IPV by a general practitioner or OB/GYN was found to be largely acceptable. In a questionnaire-based Knowledge, Attitude, and Practice survey among OB/GYN in Flanders, OB/GYN prove -unfamiliar with IPV and largely underestimate the extent of the problem. Merely 6.8% of the respondents ever received any education on IPV. They refute the incentive of universal screening, even during pregnancy and one of the major barriers is fear of offending patients. Physician education was found to be the strongest predictor of a positive attitude towards screening and of current screening practices. Hence, there is a definite need to improve women's awareness regarding abuse and to endorse physician training on IPV.

Introduction

1.1. Definition and Prevalence

Intimate partner violence (IPV) is defined as physical, psychological, economic or sexual coercion of one partner in a relationship by the other (Landa et al., 2002; Krug et al., 2002; CDC & ACOG, 2000). Lifetime prevalence estimates of partner-inflicted harm to women range from 10 up to 69% (Krug EG et al., 2002), while partner abuse during pregnancy has consistently been found to occur in 3-8% of pregnancies (Gazmararian et al., 1996; Coker et al., 2004; Janssen et al., 2003; Neggers et al., 2004; Johnson et al., 2003; Hedin et al., 1999; Saltzman et al., 2003; Silverman et al., 2006; Bacchus et al., 2004; Parker B et al., 1994). Comparison between studies is difficult because of large variations in definitions used, populations studied, study designs, and other methodological issues.

1.2. Health consequences

Women who experience IPV are at increased risk of injury and death, as well as of physical, emotional and social problems (Eisenstat et al., 1999). Physical health consequences include a 50 to 70% increase in gynaecological, central nervous system and stress-related problems (Campbell et al., 2002). Mental and emotional health impairments associated with exposure to IPV, include depression, anxiety, suicidality, posttraumatic stress disorder, mood and eating disorders, substance dependence, antisocial personality disorders, and non-affective psychosis (Wathen et al., 2003; Danielson et al., 1998; Heise et al., 2002).

Violence in pregnancy increases the risk of late entry into prenatal care (Plichta et al., 2004), perinatal death (Coker et al., 2004; Janssen et al., 2003; El Kady et al., 2005), preterm birth (Coker et al., 2004; Neggers et al., 2004; El Kady et al., 2005; Rachana et al., 2002; Saurel-Cubizolles et al., 2005), intra-uterine growth retardation and/or low birth weight (Coker et al., 2004; Janssen et al., 2003; Neggers et al., 2004; El Kady et al., 2005), antepartum hemorrhage (Janssen et al., 2003) and/or abruptio placentae (Rachana et al., 2002), uterine rupture (El Kady et al., 2005), and fetal distress (Rachana et al., 2002). Abuse can even result in femicide to pregnant women. Femicide is an important, but often unreported, cause of maternal mortality (El Kady et al., 2005; McFarlane et al. 2002; Plichta et al., 2004; RCOG, 2004). IPV during pregnancy represents a risk marker for more severe IPV, including IPV-related homicide (Campbell et al., 2003).

Moreover an emerging body of evidence shows that childhood exposure to IPV is related to risk-taking behaviours during adolescence and adulthood (Horon et al., 2001; Bair-Merritt et al., 2006; Dube et al., 2002), is significantly associated with the occurrence of learning problems and school health concerns (Kernic et al., 2002), and child behavioural problems both in the presence and absence of co-occurring child maltreatment (Kernic et al., 2003). Having witnessed IPV as a child is supposed to increase the risk of becoming victim according to some (Bensley et al., 2003; Coker et al., 2000) but not all articles (Ernst et al., 2007) and also to augment the risk of becoming a perpetrator of IPV later in life (Ernst et al., 2006), thus feeding the intergenerational cycle of violence.

1.3. Role of the health sector and of the gynaecologist

Currently, care of victims of IPV is largely confined to those who consult a physician or emergency department with abuse-related physical trauma. However, most victims of IPV do not present as such to the health care worker. The crux of IPV is that most women who encounter some kind of coercion will not present with overt signs of abuse, but rather with a wide variety of vague and non-specific symptoms, if any (Elliott et al., 1995; Muelleman et al., 1998; American Medical Association, 1992). Hence there seems to exist a window of opportunities to detect women suffering from IPV through screening in the health care sector. As a straightforward corollary, systematic screening for IPV by health care workers has been advocated by several authoritative bodies in the UK and the USA (CDC et al., 2000; ACOG, 2006; American Medical Association, 1992; American Academy of Family Physicians (AAFP) Commission on Special Issues and Clinical Interests, 1994; 2004; American College of Emergency Physicians, 1995; American Academy of Pediatrics Committee on Child Abuse and Neglect, 1998; RCOG, 2004; RCOG Study Group, 1997). On the other hand, the US Preventive Services Task Force (USPSTF) and the Canadian Task Force on Preventive Health Care (CTFPHC) could not determine the balance between benefits and harms of screening for IPV among women and hence found insufficient evidence to recommend for or against universal screening (U.S.Preventive Services Task Force, 2004; Wathen et al., 2003).

For an obstetrician-gynaecologist (OB/GYN), IPV, committed by a male partner to a woman, is of particular interest, because of its impact on women’s and children’s health, and because of medical liability in taking responsibility. IPV may affect each and every aspect of women’s health and well-being and also complicate women’s reproductive life and their offspring. In Belgium, there are few data on IPV and guidelines on IPV in reproductive health settings are missing. Hence, the aim at the very outset of this work was to make an assessment of the problem of IPV in Flanders, Belgium, and to explore this health problem from a OB/GYN’s perspective.

2. Objectives

2.1. General objective

The general objective of this work is to describe the problem of intimate partner violence in Flanders, Belgium and the role of the health care worker in dealing with this problem.

To estimate the prevalence of physical and sexual IPV among a regional sample of the general obstetric population as the lifetime prevalence, as the 1-year period prevalence before pregnancy, and as the prevalence during the index pregnancy

To assess the rates of disclosure and help-seeking behaviour with IPV among the same regional sample of the general obstetric population

To determine the acceptability of screening for IPV among the same regional sample of the obstetric population

To assess barriers of health care workers towards IPV screening. To assess the knowledge, attitude and practice of gynaecologists in relation to IPV

To describe the barriers that victims encounter when looking for professional help

To develop a protocol of care for victims of violence

3. Methodology

3.1. Multi-centre study of pregnant women

We performed a multi-centred, survey surveillance study on IPV among pregnant women attending five large hospitals in the province of East Flanders, Belgium as a regional probability sample of the general obstetric population, between September 1, 2003 and October 31, 2003 (Roelens et al., 2008). In Flanders over 98% of pregnant women visit a gynaecologist for a prenatal follow-up and over 95% of gynaecologists are affiliated to a hospital. Initially, participating obstetricians were asked to screen for IPV with the Abuse Assessment Screen (AAS) according to CDC and ACOG guidelines (CDC et al., 2000; Norton et al.,1995; Weiss et al., 2003; McFarlane et al.,1992). However, as most gynaecologists were reluctant to launch direct questions about abuse and hence were unwilling to perform such AAS-based screening, the study design was shifted towards a written questionnaire-based survey. The study was approved by the Ethics Committee of the Ghent University Hospital. The questionnaire was developed by adopting the AAS in full and further extended with a series of questions about (1) the most recent episode of physical/sexual assault to assess coping behaviour, health-seeking behaviour and disclosure of abuse to legal services, and (2) about attitudes towards IPV screening and hence the patient’s preparedness to disclose abuse. The questionnaire was developed in Dutch, and included an English-to-Dutch translation of the AAS tool without back translation. During the study period, all Dutch-speaking women who visited a gynaecologist for a prenatal follow-up at one of the five collaborating centres were offered a stamped envelope with the questionnaire and an informed consent form. Women were asked to complete the questionnaire at home and to return it anonymously by mail to the principal investigator. To distinguish between women who attended with or without a partner, the envelopes were marked differently.

A pilot study was performed in one of the participating hospitals (Ghent University Hospital) over a 2-week period, resulting in a response rate of 52.3% (56 out of 107). The overall study set-up was considered feasible and acceptable, and therefore the methodology was left unchanged in the eventual study and the pre-test results were included in the final analysis.

This work was supported by a research grant from the Province of East Flanders, Belgium and from the Flemish College of Obstetricians and Gynaecologists.

With this study we could respond to the first 3 specific objectives.

3.2. Knowledge Attitude Practice Study of Flemish gynaecologists

This was a state wide survey among board-certified obstetrician-gynaecologists in Flanders (n = 478) (Roelens et al., 2006). The questionnaire was basically designed as a knowledge, attitude, and screening and referral practices assessment tool with regard to IPV and consisted of 69 items, including 60 items with forced-choice answers (Likert-scale or yes/no) and 9 open-ended questions. The questionnaire has been approved by the Ghent University Hospital Ethical Board and by the Flemish College of Obstetricians and Gynaecologists and was sent to all members of the Flemish College of Obstetricians and Gynaecologists (n = 478). A total of 249 completed questionnaires were returned and available for further analysis, corresponding to a response rate of 52.1%. The questionnaire was divided into seven sections; i.e. a) physician and practice characteristics (n = 10), b) prevalence of IPV (n = 2), c) current IPV screening practices (n = 7), d) attitude towards IPV screening (n = 15), e) recent assessment, treatment and/or referral of patients in case of IPV (n = 19), f) knowledge and attitude towards referral possibilities and facilities in case of IPV (n = 14), and g) willingness and motivation to screen and to participate in future research with regard to IPV(n = 2). The questions and survey statements were principally developed to fit a conceptual framework, which was developed by Woolf and subsequently modified by Cabana et al. in a systematic review of barriers to physician adherence to clinical practice guidelines (Woolf, 1993; Cabana et al., 1999). We modified the model, which was derived from a retrospective literature review and applied it as a conceptual framework to model expected barriers to future guideline implementation with regard to screening for intimate partner abuse. Barriers identified fit into one of three major groups depending on whether they affect physician’s knowledge, behaviour, or practice. The model assumes a behavioural framework, i.e. before specific physician-targeted information on a health-related issue modifies patient outcome, it first affects physician’s knowledge, then physician’s attitude, and eventually physician’s behaviour and practice. Though other pathways may be involved, this algorithm is believed to underlie the most sustainable behavioural modification.

This work was supported through a research grant from the Province of East Flanders, Belgium.

With this study we tried to respond to specific objective 4.

3.3. In depth interviews of victims

The objectives of this study were to assess the barriers victims encounter when seeking professional help and to investigate experiences with professional help. General practitioners, gynaecologists of the Ghent University Hospital, psychosocial services and shelters looked for potential candidates for the interview. Candidates for the interview were given an informative letter about the project and its objectives including the structured questionnaire. They signed an informed consent before participating.

As it appeared that health care workers were embarrassed to propose the study to (potential) victims, it was decided to put pamphlets in the waiting rooms, to allow victims to present themselves for the study, without the intermediate of the health care worker. The main topics raised in the interviews related to the barriers that victims encounter when looking for professional help with health care services, police, juridical and psychosocial services. Victims were also asked about experiences with professional help resources and to formulate suggestions for recommendations.

All the interviews were done by a professional health care worker, recorded, typed and statistically handled according the principles of the program NUD*IST (Version QSR N5). Data incorporation was anonymous. Though only seven victims were interviewed, important information was gathered.

This work was supported through a research grant from the Province of East Flanders, Belgium.

The results of this study helped us to understand which barriers victims encounter when looking for professional help (specific objective 5).

3.4. Multidisciplinary elaboration of protocol of care for victims of violence

Several meetings with different departments of the Ghent University Hospital (Emergency Medicine, Psychiatry, Psychosocial service, Paediatrics, Forensic Medicine, Gynaecology, Geriatrics, Infectiology, ICT) and with authorities outside the University Hospital (delegate of the organization of general practitioners, paediatrician from the ‘Vertrouwenartsencentrum’, violence specialist from the Leuven University Hospital) took place. Based on international literature and on the expert opinions of the working group, a protocol for care of victims of violence, including victims of IPV, was developed (Buylaert W et al., 2004). This protocol is now widely distributed to other hospitals and health care worker organizations.

For the development of this protocol, we received a grant from the Province of East Flanders, Belgium.

This protocol deals with all victims of acute violence, including partner violence (specific objective 6).

4. Summary of key findings

4.1. Prevalence of IPV

From a questionnaire-based surveillance study among pregnant women (Roelens et al., 2008) constituting a regional probability sample of East-Flanders, we estimated that one in five women (22.3%) had a lifetime history of physical violence, whereas one in ten women (11.2%) had a history of sexual abuse, with more than half of the cases of sexual assault being rape or something bad with pain. Intimate partner violence in particular occurred overall with one in ten women (10.1%, 95% CI 7.7–13.0%) and with about one in 30 women (3.4%, 95% CI 2.1–5.4%) during pregnancy and/or in the year preceding pregnancy. Hence, in accordance with international data, IPV as a whole as well as in relation to pregnancy in particular, proved a particularly common threat in our community.

4.2. Rates of disclosure and help-seeking behaviour with IPV

Despite the tremendous burden of violence-related harm, we found that merely 19.2% (23 out of 120) and as few as 6.6% (4 out of 61) of the victims of physical and sexual abuse respectively sought medical care by consulting a general practitioner, a gynaecologist, or an emergency department. Similarly, less than 1 in 5 women (17.5%) reported physical violence and less than 1 in 10 (6.6%) reported sexual assault to the police (Roelens et al., 2008).

Hence, as a major finding of concern, we revealed that even in our highly medicalised society, women experiencing IPV rarely disclose abuse spontaneously to the widely available health care services and providers and therefore fail to pursue for the most part medical care following a violent attempt.

4.3. The acceptability of screening for IPV

In the context of a high prevalence of IPV not disclosed spontaneously, routine screening for violence by health care workers seems a straightforward solution in the secondary prevention of this common affliction. However, such a screening incentive can only be considered if screening for violent experiences would also prove acceptable to a general gynaecologic or obstetric population. In order to assess the attitude of women towards being asked about exposure to violence as part of routine medical history-taking, survey participants completed three distinct questions, i.e. their attitude towards being directly asked about a history of abuse by their general practitioner, by their gynaecologist, or through a questionnaire similar to the survey instrument (Roelens et al., 2008). As shown in table I, the vast majority of women approved or, if not, did at least not disapprove routine questioning regarding abuse by their GP or gynaecologist. Women with a history of abuse were overall slightly but significantly less in favour of direct questioning by their GP or by their gynaecologist, though no such difference occurred with questionnaire-based screening.

Table I.

Attitude towards being asked about exposure to violence.

Asking questions about violence	Considered as negative		Considered as neutral		Considered as positive
	Ever assaulted	Never assaulted	Ever assaulted	Never assaulted	Ever assaulted	Never assaulted
By general practitioner	12.9% (19/147)	9.1% (35/383)	36.7% (54/147)	30.3% (116/383)	50.3%• (74/147)	60.6%• (232/383)
By gynaecologist	16.3%° (24/147)	7.6%° (29/382)	29.9% (44/147)	28.3% (108/382)	53.7%‡ (79/147)	64.1%‡ (245/382)
In this questionnaire	1.4% (2/147)	1.0% (4/382)	12.9% (19/147)	16.2% (62/382)	85.7% (126/147)	82.7% (316/382)

For the items indicated, there was a statistically significant difference in attitude between women with and without a history of abuse (•p = 0.04, °p = 0.004, ‡p = 0.04). For all other items displayed in the table there was no such difference.

4.4. Barriers of health care workers towards IPV screening. Knowledge, attitudes and practices of gynaecologists in relation to IPV

In the KAP study (Roelens et al., 2006), we used a conceptual framework to model expected barriers to future guideline implementation with regard to screening for IPV. The model (Figure 1) assumes a behavioural framework, i.e. before specific physician-targeted information on a health-related issue modifies patient outcome, it first affects physician’s knowledge, then physician’s attitude and eventually physician’s behaviour and practice.

Fig. 1

Screening for partner violence: Knowledge, Attitude and Practice (Roelens et al., 2006).

Accordingly, we aimed to assess various self-perceived barriers among a community-wide sample of OB/GYNs, within these distinct segments of knowledge, attitude, and behaviour with regard to IPV.

Knowledge: familiarity and awareness

Familiarity with IPV

Merely 6.8% of the respondents in the survey (17/249) stated having received or pursued any kind of education or information on IPV. Over two thirds (67.9%) of the participating gynaecologists (169/249) acknowledge that there is a defined need to incorporate such education during medical training.

We did not further assess physician’s familiarity with risk factors, signs, symptoms, and co-morbidity patterns relating to IPV as an issue of knowledge. However, the gynaecologists stated that they would screen in case of suspicion of IPV: most of them would screen in case of ‘overt physical lesions’ and not in case of ‘’psychological complaints’. As we know that only a minority of victims will present with physical lesions, it seems that gynaecologists are rather unfamiliar with the signs and symptoms of IPV.

Neither did we make an attempt to assess obstetrician’s knowledge of screening strategies as an indicator of their relevant knowledge.

Awareness of intimate partner abuse

Intimate partner abuse is deemed a rather rare phenomenon by most survey participants, i.e. affecting less than 1 in 100 or even less than 1 in 1000 patients attending. Merely one in four gynaecologists estimated the prevalence of abuse among non-pregnant women to be at least one percent. Similarly, intimate partner abuse is thought to occur with at least one percent of pregnant women by only one in five gynaecologists.

Attitude: incentive agreement, motivation, perceived self efficacy, and outcome expectancy

Agreement with the incentive to screen

In the main, survey participants decline universal and systematic screening and also refute the common view of pregnancy as a window of opportunity to screen. Rather, obstetrician-gynaecologists do favour direct questioning of the patient by means of the Abuse Assessment Screen (AAS) in case of suspected abuse.

Even in the context of a prevalence study (Roelens et al., 2008), the gynaecologists did not want to screen directly using the AAS. This is the reason why we changed the initial study design of the prevalence study, switching from the AAS to an anonymous written questionnaire.

Motivation

Most physicians surveyed consider directed screening though not universal screening as an issue of medical liability. Respondents largely contradict common beliefs about partner abuse and in particular they do not consider it a family affair, for which partners should take responsibility, nor a phenomenon pertaining to lower social classes or an affliction for which the victim itself is to blame.

Perceived self-efficacy

The preponderance of survey participants feels insufficiently skilled to discuss partner abuse in a straightforward manner and to manage abuse-related issues with putative victims of domestic violence. Similarly, physicians surveyed feel insufficiently acquainted with referral practices in case of partner abuse.

Outcome expectancy

A majority of obstetrician-gynaecologists believes that screening for IPV may be an effective means to counteract such abusive behaviours. Yet, about half of all survey respondents are also convinced that there is a defined lack of referral services and specialised care facilities for women suffering from IPV.

Practice and behaviour: screening practices and perceived barriers

When asked for current screening practices at the time of the survey very few obstetricians appear to adhere to a universal screening policy. Merely 8.4 % of all participants (21/249) stated to screen each patient at least once during pregnancy. Lack of time and fear of offending or insulting patients were the most frequently cited barriers towards implementation of screening.

Survey respondents feel confident in relying on their clinical index of suspicion in their screening practice and stated to launch direct questions most of the time in case of suspected abuse. Partner abuse is however only suspected in case of overt physical lesions, whereas psychological distress or complaints more rarely prompt direct questioning about potential abusive behaviours.

An indirect estimate of screening sensitivity was obtained from a series of survey questions regarding the most recent case of physical and/or sexual abuse treated by the responding physicians. The median time span elapsed at the time of the survey since the last victim encounter was 6.3 months (interquartile range 3.1 to 12.8 months) and 5.9 months (interquartile range 2.3 to 12.4 months) for sexual abuse and physical abuse respectively. Overall, one in three obstetricians stated not to have encountered sexual coercion and two in three not to have confronted physical abuse among their patients over the past five years.

Determinants of attitudes, practice and behaviour

Physician education was found to be the strongest predictor of a positive attitude towards screening and of current screening practices.

4.5. Barriers that victims encounter when looking for professional help

We interviewed seven victims of violence to assess barriers that victims encounter when looking for professional help. We also asked them about experiences with professional help resources and to make suggestions for recommendations. We tried to find candidates for the interview through different ways: the gynaecologists of the University Hospital, the general practitioners in Ghent, the social service of the Ghent University Hospital and the shelter CAW Artevelde in Ghent.

But again health care workers found it difficult to actively procure victims and on the other hand victims could not (language barrier) or did not want to be interviewed. Hence, only 7 victims were interviewed.

The main ‘internal barriers’ when looking for professional help were formulated as: ‘looking for help implicates that you admit that something is wrong’, ‘not being aware that psychological violence is violence’ and ‘fear of escalation of violence’.

On the other hand, the barriers related to care providers were lack of accessibility (‘you have to start talking, they never ask about violence’), money (especially for paying the lawyers) and lack of coordination between the different care providers. The victims also complained about a lack of action of the care providers: ‘they only listen’, ‘no action on the perpetrator’, ‘late intervention’ or no intervention based on ‘professional secrecy’. The victims also recommended that care workers should be trained and sensibilised, that professional help should be widely available and accessible and that the different care workers should coordinate their actions.

4.6. A protocol of care for victims of violence

Through multidisciplinary collaboration and based on international literature, a protocol of acute care for victims of violence including partner violence during and outside pregnancy, was developed (Buylaert et al., 2004). This protocol gives tools for acute medical care for victims, for referral to other facilities and for attestation.

The aims of the protocol are:

Immediate care to the victim, including psychological help.

Prevention of consequences of violence such as unwanted pregnancy, sexually transmitted infections …

Adequate referral to psychosocial care facilities.

Correct documentation of lesions, including forensic evidence collection. Correct attestation.

Referral to legal aid.

Education and training of health care workers.

For each victim a coordinator (a medical doctor of the main discipline involved) is assigned who has to ensure that:

a checklist with his/her name is in the patient’s file

the patient has been seen by all necessary disciplines

a complete file for each patient is made

the social service is proposed to the patient and contacted if necessary

a contact person of the patient is contacted if he/she wants so

the family doctor is contacted

attestations are written

follow-up is ensured.

In case of violence, according to our protocol the duty of the doctor goes much further than pure medical care and should include referral to other facilities.

Together with the implementation of the protocol, trainings of health care workers (emergency ward, gynaecology, surgery, internal medicine, psychiatry) and of social workers were organized. Violence was also incorporated in the medical curriculum in the Ghent University. The protocol has been widely used and spread to other hospitals in Belgium and to professional organizations.

5. Conclusions and recommendations

IPV is a common affliction of women’s health in our community and may therefore well be considered an important public health issue. Partner abuse may affect each and every aspect of women’s health and well-being and also complicate women’s reproductive life and their offspring.

In Flanders, women experiencing partner violence rarely disclose abuse to the widely available health care services, unless they are directly asked about it, which appears an acceptable practice.

Systematic screening for IPV by health care workers has been advocated by several authoritative bodies in the UK and the USA. This has been challenged by others, mainly because of safety concerns and lack of evidence in support of effectiveness. However, from recent literature positive health outcome following screening and intervention for violence becomes more and more evident (Wathen et al., 2003; Gerbert et al., 1999; Gerbert et al., 1999; Rhodes et al., 2003; Drossman et al., 1996; Rosenberg et al., 2000; Glowa et al., 2003; Chang et al., 2005; McFarlane et al., 2002; McFarlane et al., 1997; Parker et al., 1999; McFarlane et al., 2000; Krasnoff et al., 2002; Tiwari et al., 2005; Melendez et al., 2003; McFarlane et al., 2006; Sullivan et al., 1992; Sullivan et al., 1999; Eckenrode et al., 2000; Olds et al., 1997; Huefner et al., 2007). Recently publishes articles also give some evidence that screening is rather safe (Liebschutz et al., 2008; Houry et al., 2008). Safety problems on the other hand can partly be avoided by screening in a private confidential environment.

In our setting, gynaecologists not only account for gynaecologic and obstetric pathology but also act as the primary care physicians to the general female population, e.g. in providing primary obstetric care and in offering preventive women’s health care, yet appear to feel uncomfortable with a routine IPV screening policy. This in turn was shown to relate to several intrinsic and extrinsic barriers. The main internal barriers identified are related to a lack of knowledge - and hence underestimation of prevalence and unfamiliarity with signs and symptoms- and perceived lack of self-efficacy in dealing with the problem and properly referring patients. The main external barriers of the gynaecologists pertain to lack of time and fear of offending patients. On the other hand it may also be acknowledged that gynaecologists actually see it as their medical responsibility to act in case of suspected abuse and they believe that with interventions the outcome of the victims might be improved. In our survey, OB/GYNs recognize that their knowledge and self-efficacy is inadequate and acknowledge that there is a need for training on violence. It therefore appears that most barriers should be remediable through proper physician training and education.

In Belgium, it was not until 1997 that IPV is considered as a crime by the Belgian law, and it took until 2000 for concrete actions to combat violence. The National Action Plan (NAP) to combat partner violence was elaborated by different Ministries (The Ministry of the Civil Service, Social Integration, Cities Policy and Equal Opportunities, together with the Ministry of Justice, the Ministry of Home Affairs and the Ministry of Social Affairs and Public Health) and was launched as a holistic plan with actions targeted at increasing awareness, prevention, training, assistance, repression and other measures at different levels of society.

At the health care level, the National organization of family physicians developed a consensus on the role of the family physician in detecting and dealing with IPV. Very recently this consensus has been updated and now it includes the recommendation that all pregnant women should be screened for IPV. However, OB/GYNs are not involved in this action plan as yet and no recommendations regarding pregnant women and IPV have been developed or endorsed for OB/GYNs. As a matter of fact, until now there is no consensus on violence for gynaecologists.

Hence we formulate the following recommendations:

1. Development of guidelines and training for OB/GYN practitioners on the issue of IPV

Gynaecologists need formal guidelines including

information on prevalence, health consequences and signs and symptoms of IPV;

information on attitudes of patients towards being screened for IPV;

transferable skills and screening tools;

consensus statement and recommendation on screening;

guidelines for acute care for victims of violence;

knowledge of formal referral pathways

These guidelines serve a number of valuable functions: they can be a useful educational and training tool helping gynaecologists to improve the detection rate and the quality of treatment and support provided to victims of violence. Clear guidelines with formal referral pathways make it possible for the gynaecologist to deal with the problem in a time-efficient manner.

2. Development/testing of an acceptable screening tool in our setting

The Abuse Assessment Screen – the screening tool recommended by the CDC, ACOG among others – does not seem to be an acceptable screening tool in our context. No other screening tools have been tested however; hence there is a need to study what the best screening tool would be. Our study group (ICRH/Department of Gynaecology & Obstetrics of the Ghent University Hospital) is currently applying for funds with one of the specific objectives being the development of the best screening tool in the Belgian context.

3. Development of a screening policy, in and outside the context of pregnancy

We would recommend universal screening for IPV during pregnancy, and opportunistic screening for violence beyond pregnancy, i.e. in case of suspicion of abuse (gynaecological complaints not explained by any evident organic cause).

Pregnancy offers a unique window of opportunity for screening: there is a regular contact with the woman, and questions about family situation, work, smoking and drinking habits etc are asked on a routine basis, allowing one to integrate some direct questioning about violent experiences. On the other hand, during gynaecological consultations, especially when women have obvious organic pathology or come with a clear question there is less room for systematic screening and hence a universal screening policy seems less feasible and possibly less acceptable in this setting, while it does, however, not prevent an opportunistic screening strategy.

The development of a consensus on screening is not easy and we admit that in our context a universal screening policy might be too ambitious or not realistic, but we are convinced that we should at least augment the detection rate of IPV by sensitive opportunistic screening. Screening based on risks factors (e.g. teenage pregnancy, history of abuse, drug addiction) or on symptoms is second choice as the crux of IPV is that victims will not present with overt signs of abuse, but rather with a variety of vague and non-specific symptoms. The physician’s eye is therefore even in the presence of a high index of clinical suspicion unlikely to grasp most victims and their potential signs in a general obstetric or gynaecologic population.

4. More studies on the safety of screening and the outcome after screening and intervention for violence

There is a need to know more about the safety of screening and the outcome after screening and intervention for violence. These important barriers are used as arguments against universal screening by some authorities.

5. Resources for confidential reporting of intimate partner violence

In the past when confronted with (the suspicion) of child abuse, doctors were in conflict between two opposing principles: the duty of professional confidentiality and the duty to provide assistance. Resources for confidential reporting (in Belgium the so-called ‘vertrouwenartsencentra’) were founded and made it possible for the medical doctor to report the (suspicion of) violence and to refer the patient, without breaking the professional confidentiality.

Much alike it would be of great help to have resources for confidential reporting for intimate partner violence too, especially in the case the woman does not want to report the IPV to other health care workers or referral facilities.

6. Reinforcement of the role of ‘Kind en Gezin’/‘Office de la Naissance et de l’Enfance’

There is a strong link between IPV and child abuse. There is also good evidence that follow-up and home visits by social workers can prevent child abuse and neglect.

Since more than a century, we have child welfare, a social organization in Belgium, involved in vaccination and follow-up of newborns including home visits. They also offer social care to deprived pregnant women. The population coverage approximates 100%.

Hence, we would call for a reinforcement of the role of ‘Kind en Gezin’/’Office de la Naissance et de l’Enfance´ in the prevention of child abuse and neglect in families with IPV.

7. Continuation of the National Action Plan

We strongly believe that societal-level interventions, such as policing and legislative policies, might affect the incidence of violence against women (primary prevention), but above all improves the help for victims (and perpetrators) of intimate partner violence (secondary and tertiary prevention). Accordingly, we would plea for a continuation of the National Action Plan as to make the battle against violence a continuous effort.