AI system identifies domestic abuse risk 'three years' before victims seek help

Researchers in the United States have developed a pioneering artificial intelligence system capable of identifying patients at risk of intimate partner violence years before they officially enter intervention programmes.

Trained on comprehensive hospital data, including medical histories and written clinical notes, the technology recognises subtle patterns of trauma that mirror confirmed cases of domestic abuse.

By detecting these indicators, the system can flag potential victims who may otherwise remain silent due to fear or the social stigma associated with their situation.

The study, published in the journal Nature, demonstrated that the tool could identify risk more than three years before a patient typically seeks help.

Achieving an 88 per cent accuracy rate, the technology is designed to function as a clinical decision support system rather than a standalone diagnostic tool.

This provides a sensitive signal to healthcare professionals, allowing them to offer support without replacing professional medical judgement.

Regarding the breakthrough, Qi Duan of the US National Institute of Health said, “This clinical decision support tool could make a significant impact on prediction and prevention of intimate partner violence.”

Future implementation remains a priority for the research team, with scientists planning to integrate the AI directly into electronic medical record systems.

This will allow hospitals to receive real-time risk assessments during routine patient consultations, effectively bridging the gap between initial trauma and formal intervention.

This proactive approach represents a significant shift in preventative medicine, potentially saving lives by initiating care long before a crisis peaks and providing a vital safety net for vulnerable individuals.