Bulletin of the World Health Organization

Health-sector responses to intimate partner violence in low- and middle-income settings: a review of current models, challenges and opportunities

Manuela Colombini a, Susannah Mayhew a, Charlotte Watts a


Over the past 10 years, violence against women has become recognized as a serious public-health issue.1 Research, initially in North America and Europe, but increasingly from other settings, has helped demonstrate the high prevalence and wide range of health consequences of intimate partner violence (IPV). The WHO multicountry study on women’s health and domestic violence showed that the lifetime prevalence of physical or sexual partner violence, or both, varied between 15% and 71% in 10 countries.2 Abused women are more likely to have poorer health than women who have never been abused3 and may suffer health consequences of violence long after the abuse has ended.4 The physical health consequences include both injury and a broader range of impacts,4,5 including: (i) nutritional status, digestive problems and hypertension;6 (ii) sexual and reproductive health, including fertility, contraceptive use, and HIV and sexually transmitted infections (STIs); (iii) maternal health, including increased risk for high blood pressure, risk of antepartum haemorrhage and of miscarriage;79 and (iv) mental health, including risk of depression and suicide.10 IPV has also been shown to affect the health and well-being of children in violent families, e.g. by decreasing vaccination status and increasing the risk of behavioural and psychological problems.7,8

Need for health-sector responses

Given that IPV is an important risk factor for a range of health problems, there has been growing awareness of the need for health providers to be able to respond better to cases of violence that they encounter, and to help identify women experiencing violence and refer them to specialized services. This referral is very important, as many women experiencing violence will never seek help from a legal or stand-alone service, but will probably go to a health service during their adult life. Women may access the health system at a range of potential entry points for service provision and may have a range of presenting health needs. Some women experiencing partner violence will present at primary care,11,12 while women experiencing serious injuries may present to hospital emergency services.1315 Given that coerced sex and violence in pregnancy is widespread, ante- and postnatal care, family planning or post-abortion care are also potentially important entry points.1618 Therefore, it is important that the health sector ensures not only the efficient delivery of health-related services to victims of violence, but also facilitates these women’s access to non-health services.

Some women may disclose violence without being questioned, while others may not openly disclose the cause of their presenting problem. Much of the debate regarding the health-sector response has focused upon whether women should be “screened” for violence, and whether such interventions impact on women’s future risk of violence.19,20 There has been much less debate about what may be the most important entry points for health-sector involvement in different settings, or consideration of what may be the most feasible ways for health services in low- and middle-income countries to integrate responses to violence into the health sector.

After briefly summarizing the evolution of literature on integration and its integrated service models, this paper reviews promising health-sector responses to violence currently being integrated into existing services in low- and middle-income countries. We present a broad framework to help conceptualize the potential entry points for care, and the required systems of referral both within and outside the health sector. Challenges and opportunities, and future research priorities, are then discussed.

Models of service integration

Debates on “integration” of services, and what constitutes “integrated” services at different levels of the health system have been ongoing since the 1970s. During the 1980s a (somewhat false) dichotomy emerged between “selected” (issue-specific, more vertically organized) services and “comprehensive” (more linked or integrated) services. During the 1990s, much attention was given to expanding the remit of family-planning programmes to encompass a broader range of reproductive and sexual health services, notably the management of STIs, including HIV.21,22 Research focused on how providers, facilities and health systems (including policies and programmes guidelines) could – or should – respond to the challenge of adding, or integrating, new services into existing ones.

An extensive literature from various fields, but particularly sexual and reproductive health, highlights a range of issues associated with integration.21,2327 What emerges is the lack of consensus on a definition, although “integrated services” tend to be equated at some level with the notion of “holistic” service delivery.27 In general, this literature describes integration at three different levels: at the level of the provider, the facility and the system. “Provider-level integration” means the same provider offers a range of services during the same consultation, e.g. a nurse in accident and emergency is trained and resourced to screen for domestic violence, treat her client’s injury, provide counselling and refer her to external sources of legal advice. “Facility-level integration” means a range of services is available at one facility but not necessarily from the same provider, e.g. a nurse in accident and emergency may be able to treat a woman’s injury, but may not be able to counsel a woman who discloses domestic violence, and may need instead to refer the woman to the hospital medical social worker for counselling. “Systems-level integration” means that there is a coherent referral system between facilities so that, for instance, a family-planning client who discloses violence can be referred to a different facility (possibly at a different level) for counselling and treatment. Unlike provider- and/or facility-level integration, which usually happen within the same site, system-level integration is multisite. Within the health sector, most services involve a combination of provider- and facility-level integration; full systems-level integration is rare.

Integration literature highlights challenges at each of the three levels. At the provider level, entrenched medical hierarchies may impede putting training on integrated service provision into practice.22 At the facility level, many issues have been identified, including poor management, shortage of personnel and supplies, lack of appropriate equipment for expanding services, and poor physical infrastructure.2123,26 At a systems level, there is a lack of coordination among various actors and departments involved in planning integrated services, lack of clear guidelines for training staff, underfunding and poor, or no, legislative systems for supporting integration (e.g. if nurses are to be enabled to prescribe drugs for STI management, the law or health policy may require changing).2124,27

Health-sector responses

For this paper, a detailed literature review of the published and grey literature (in English, French and Spanish) on promising health-sector interventions responding to violence in low- and middle-income countries between 1995 and 2005 was conducted. Sources of research evidence for the review included electronic bibliographic databases (African Healthline, Cochrane Library, ELDIS, Isis Web of knowledge, LILACS, Popline and PubMed®); web sites from key organizations/nongovernmental organizations (NGOs) in the field; reference lists from primary and review articles; peer-reviewed journals, grey literature and conference proceedings. Programmes were selected based on the following criteria: (i) implemented between 1995 and 2005; (ii) focused specifically on health-service provision to address violence against women in developing countries; and (iii) where possible, being evaluated or measured.

The literature review identified 17 promising programmes that, after further appraisal, were reduced to nine: five programmes in central and Latin America, three in Asia and one in Africa.25,2835 Of the nine models reviewed, seven were implemented in middle-income countries, and two in low-income countries. Four were implemented at a primary level, and five at secondary or tertiary level. Drawing on the integration models discussion above, these programmes have been characterized into three models of integration: (i) provider- and/or facility-level integration of selected services at the same site (i.e. a few selected services are integrated into existing services by the same provider and/or on one site); (ii) provider- and/or facility-level integration of comprehensive services at the same site (i.e. a wide range of services are integrated into existing services by the same provider and/or on the same site); and (iii) systems-level integration involving multisite linkage in addition to provider- and/or facility-level integration. Table 1 summarizes the programme models and they are discussed below.

Provider- and/or facility-level integration at same site

Selective integration

Implemented at both primary and secondary level of health care, the first type of model is characterized by the integration of one or two service components for abused women (e.g. counselling or psychological therapy) in vertical programmes. For instance, in Honduras, regional family-counselling centres, based at regional mental-health clinics, provide individual and group counselling for abused women;33 there are no external referrals. Another example comes from a pilot project in Brazil, where a dedicated counselling programme – CONFAD – has been integrated at a medical-school health centre and provides basic and therapeutic counselling.34,36

Comprehensive integration

An alternative model is the provision of a comprehensive range of IPV services delivered in one setting. This is found most commonly in industrialized settings (particularly in the United States of America) and primarily at secondary or tertiary levels of care.43 Developing country examples include the One-Stop Crisis Centre (OSCC) model, initially developed in Malaysia (based on a Canadian model) for battered women and later extended to rape and sexual assault,28 and now implemented at a national level in Bangladesh, Malaysia, Namibia and Thailand.16,32 These operational centres offer a wide range of integrated services to address IPV, including health, legal, welfare and counselling services, in one location – usually the accident and emergency departments of urban public hospitals. Some of these centres have dedicated staff manning the centres at all times, others have core staff members and a list of contacts, such as psychologists and medical social workers, who can be called upon to provide specialized services on site when needed.

Systems-level integration (multisite linkage)

Though still offering a comprehensive package of services for abused women, the third type of model differs from the two previous ones because services are not all provided at the same site. A range of basic services, such as screening and medical care, is delivered at one facility, with external referrals to other facilities for specialized services. The examples reviewed are from both primary- and secondary-care levels, with reproductive-health services being the main entry point.

At the primary-care level, the three-site regional International Planned Parenthood Federation (IPPF) project in Latin America44 integrated violence screening and related support services (counselling, legal advice) into existing sexual and reproductive health services.35,45 Affiliates’ clinics also strengthened their off-site referrals and established a partnership with local NGOs to refer abused women for psychological support and shelter.45 For instance, Plafam, in Venezuela, established external referrals to specialized psychological, legal and social services.40 Over the course of the project, the affiliates’ clinics (especially in the Dominican Republic) evolved into a more comprehensive model, trying to offer all the services at one facility, though challenges remained, including a scarcity of financial and human resources, and staff time constraints.

At secondary and tertiary levels, a range of medical IPV services has been integrated, mainly into maternity hospitals, with external referrals for other specialized services.25,30,31 For instance, in Bangladesh, the Women’s Friendly Hospital Initiative aimed to reduce maternal mortality and violence rates, and included treatment and referral for abused women among its services. On-site services included medical care, documentation of injuries, and external referral for social and legal support to other agencies or higher-level hospitals.30 In Armenia, Prime II project, has integrated IPV services in a polyclinic’s reproductive health services, and used a coordinated approach to strengthen external links to counselling, legal aid social support, hot-line services and shelter.31 A similar approach was used in a women’s hospital in Kenya, with referrals to NGOs for legal and economic support, shelter and police investigations.25

Challenges and opportunities

Several publications have discussed the challenges associated with integrating responses to gender-based violence into the health sector.16,35,46 The models described above illustrate how each may face common problems, but also that each model has both strengths and limitations. For instance, the OSCC model offers a broad range of health and legal services in one setting and is based in a non-stigmatized department. However, its location at a tertiary level may result in a more limited coverage than interventions implemented at a primary-care level. Being integrated within a reproductive-health clinic, and incorporating routine screening, the IPPF programme has the potential for broader outreach, but is dependent upon referring women externally to legal and other support services. Despite this, screening and detection rates increased in the three IPPF affiliates’ clinics after on-site integration and staff training,35 though some barriers, such as lack of time and of referrals to community services, remained.44

Human resources, training and management protocols

A major concern in the provision of services for violence is to ensure that women are not further victimized by the health sector, but are treated sensitively.28 Related to this, a common issue of concern among the various models is the challenge of ensuring that health personnel are appropriately trained to provide support services. Findings from the IPPF regional initiative in Latin America, for example, show that initially some providers discounted women’s stories, seemed uninterested, ignored the situation and focused on physical symptoms.45 However, over the course of the project, low knowledge and negative attitudes of providers towards screening for abuse seemed to decline, as staff, once trained, became more aware of the links between IPV and sexual and reproductive health,47,48 and felt more empowered and committed to raise the issue of violence with their clients.44

The models reviewed had different approaches to staff training. In the Dominican Republic, all staff of six family planning clinics offering screening, free-standing counselling and legal services, were trained, including receptionists and security staff,29 though the level of training was given strategically, as not all staff were expected to respond to IPV with equal competence and dedication. In Malaysia, only doctors and staff nurses from the accident and emergency and the gynaecology departments were trained, as they were the most involved in the provision of services.

Several models complemented training initiatives with the development of clear procedures and guidelines for providers that stated their required roles and competencies, and that established systems for supervision and ongoing monitoring. Studies of the Latin American and the Malaysian OSCC models suggest that the integration of policies, protocols and other tools and procedures for IPV response is important to help institutionalize IPV services as part of delivery care, and contributed to the improvement of their implementation.28,29,44

As an aspect of this, the sustainability of training in the long term is a common challenge for integration. For instance, lack of record-keeping, high staff turnover, and variations in local training programmes were constraints faced in the Malaysian programme.28,49

Financial, structural and health-system issues

Financial constraints appeared to be a challenge in all integrated models, but especially of the stand-alone, hospital-based OSCCs, where funds depend mainly on local hospital boards.49 Poor infrastructure, non-existent or poor documentation systems, and lack of private examination and counselling rooms were some of the challenges health services typically faced. The programmes reviewed illustrate good practices in these regards, especially around confidentiality and privacy, as private spaces were created for screening and treating abused women, and policies safeguarding confidentiality of medical records were reinforced.45 In Malaysia, colour coding or stamping on registration files were two systems developed to protect clients’ confidentiality.28,29,44 In the Dominican Republic, sound-proof clinic rooms were created, though the final evaluation showed that staff still entered consultation rooms while providers were with abused women.50

Partnerships with other agencies and organizations

It is important that health systems are able to facilitate women’s access to health and non-health services if needed. The stand-alone model of OSCCs aims to partially address this issue by providing a range of services within accident and emergency and other units of the hospital, as well as off-site referral for specialist non-health services, although referral to other settings may be limited by the options available. In Malaysia, despite most OSCCs providing temporary shelters to victims for the night, there was a lack of emergency shelters to which women could be referred.28 Generally, partnership with local women’s NGOs, whenever available, proves to be a crucial element for providing support services to abused women once discharged. In Venezuela, the multisite-linked model also experienced a paucity of referral sites and additionally had difficulties in following up cases that had been externally referred.40,44 On the other hand, in the Dominican Republic, Profamilia strengthened its off-site referrals, and established a partnership with local NGOs to refer abused women for psychological support and shelter.29,45


There is growing recognition of the public-health burden of IPV and the potential for the health sector to identify and support abused women. Drawing upon models of health-sector integration, this paper has reviewed current initiatives to integrate responses to IPV into the health sector in low- and middle-income settings.

The review is limited in some ways because very few of the identified programmes have been evaluated systematically. Available publications offered a descriptive analysis of some evaluated health-setting approaches, but little mention of the processes, or contextual factors, that influence an organization’s integration of IPV services. Therefore, it has been difficult to analyse some of the selected programmes, especially in low-income countries (Bangladesh and Kenya). Nevertheless, our review gives an overview of the range of responses being implemented, and illustrates the degree to which health systems in low- and middle-income settings are starting to engage with the issue of violence.

Our paper shows that many countries are actively seeking to respond to the issue. It appears that a few key models of integration are replicated in many settings, which can be characterized by their level and type of integration: (i) provider/facility-level integration providing selected or comprehensive services; and (ii) systems-level integration providing referral to services across multiple sites. The models provided services at primary, secondary and tertiary levels of care. Based on our findings, Fig. 1 presents a summary model of the entry points identified for integrating IPV services into existing health services (which see a range of presenting conditions) and the referrals necessary to ensure full systems-level integration.

Fig. 1. Potential entry points for delivery of health care to abused women and systems of referral for effective integration
NGOs, nongovernmental organizations; STIs, sexually transmitted infections.

This paper highlights the multiple challenges faced at different levels of integration and in different country contexts. These range from individual service providers’ attitudes and lack of knowledge about violence to managerial and health systems’ challenges, such as insufficient staff training, lack of inclusion of violence-response training in national medical curricula, no clear policies on IPV, and lack of coordination among various actors and departments involved in planning integrated services. Furthermore, given the variety of locations where women may present and the range and potential severity of presenting health problems, there is an urgent need for coherent, effective referral within the health sector, and for strong local partnerships to facilitate effective referral to external, non-health services. The influence of more external structural and political issues (including laws on IPV and the availability of external sources of support for women experiencing violence) is also important.

Further research is needed to understand the successes and operational lessons to be learned for the scale-up in different settings. These need to be addressed if the quality and appropriateness of services provided is to improve. ■

Funding: World Health Organization, DFID and Sigrid Rausing Trust.

Competing interests: None declared.



  • London School of Hygiene and Tropical Medicine, 49-51 Bedford Square, London, England.