ABSTRACT
Domestic violence (DV) is a global prevalent health problem leading to adverse health consequences, yet health systems are often unprepared to address it. This article presents a comparative synthesis of the health system's pre-conditions necessary to enable integration of DV in health services in Brazil, Nepal, Sri Lanka and occupied Palestinian Territories (oPT). A cross-country, comparative analysis was conducted using a health systems readiness framework. Data collection involved multiple data sources, including qualitative interviews with various stakeholders; focus-group discussions with women; structured facility observations; and a survey with providers. Our findings highlight deficiencies in policy and practice that need to be addressed for an effective DV response. Common readiness gaps include unclear and limited guidance on DV, unsupportive leadership coupled with limited training and resources. Most providers felt unprepared, lacked guidance and felt unsupported and unprotected by managers and their health system. While in Brazil most providers felt they should respond to DV cases, many in Sri Lanka preferred not to. Such organizational and service delivery challenges, in turn, also affected how health providers responded to DV cases leaving them not confident, uncertain about their knowledge and unsure about their role. Furthermore, providers' personal beliefs and values on DV and gender norms also impacted their motivation and ability to respond, prompting some to become 'activists' while others were reluctant to intervene and prone to blame women. Our synthesis also pointed to a gap in women's use of health services for DV as they had low trust in providers. Our conceptual framework demonstrates the importance of having clear policies and highlights the need to engage leadership across every level of the system to reframe challenges and strengthen routine practices. Future research should also determine the ways in which women's understanding and needs related to DV help-seeking are addressed.