Another limitation of this analysis is that it does not provide estimates of violence for subpopulations that other literature has identified as particularly vulnerable to violence, including disabled girls and women10, and girls and women living with HIV. 11, 12 A focus on these vulnerable groups is also an important direction for future research; although data limitations and variations in how this information is collected across contexts might render more systematic estimates impossible at this stage, and therefore incorporating these dimensions into data surveillance systems might be a useful step moving forward.

The literature analysing the effectiveness of strategies to prevent and reduce intimate partner violence in lowincome and middle-income countries has expanded rapidly in previous years. Much of this literature is now encapsulated in the RESPECT framework13 and its implementation plan. 14 Clearly, given the scale of intimate partner violence in 2018, as summarised in this analysis, and the fact that evidence suggests that violence has increased further during the COVID-19 pandemic, it is crucial to focus on exploring methods by which prevention interventions for intimate partner violence can be deployed at scale using available human resources, and integrated into existing health, educational, and social protection systems. This comprehensive new data should only re-emphasise the urgency of developing, evaluating, and scaling strategies for the prevention and reduction of intimate partner violence for women around the world, particularly the most vulnerable.