Xantura's Children’s Predictive Safeguarding Model brings together data from multiple agencies to identify children who are most at risk of neglect or abuse, but were not previously known to the local authority, to help social workers intervene early.
The model brings together data from multiple agencies and applies risks scores to predict, for children under two years old, the likelihood of them being neglected or abused by the age of five. The model includes an information sharing platform, a secure alert system which sends escalated risk scores to social work teams, and tools to support the engagement of families as risks begin to escalate.
The effectiveness of the model is enhanced by timely data from the vulnerable families and data from others with whom they are in contact, using mobile phone and web technology. The entire system is used to support the professional judgement of social workers – not to override it.
In the longer term this model should strengthen safeguarding across London’s local authorities, help more families have greater independence from intervention, and improve families’ quality of life whilst cutting costs.
Predictive analytics have already been used by the NHS, Department for Work and Pensions, London Fire Brigade, and a number of London local authorities. The benefits of the venture are:
- Savings of circa £122k from increased efficiency in Trouble Families (TF) teams.
- Increased identification of Troubled Families. One LA has already identified almost 400 additional families to receive support through their TF programme.
- Improved access to multi-agency data, leading to increased efficiency in safeguarding teams, equating to circa £148K.
- Identification of families and children at an earlier point than currently, leading to more targeted, effective interventions and a potential reduction in the number of safeguarding cases. This could amount to cost avoidance of over £700K.
Thomas Man Head of Capital Ambition (firstname.lastname@example.org)
Lisa Henry Programme Manager, Capital Ambition (email@example.com)