Skip to content

2016 Conference Presentation

Outcomes and qualitySafeguarding EnglandUnited Kingdom

5 September 2016

Models of safeguarding: costs and outcome

Martin Stevens, King’s College London, UK


Objective: Adult safeguarding is the subject of increasing attention in England and internationally. The aim of this paper is to explore the implications of the different models of adult safeguarding, explored in the previous papers in this session, in relation to outcomes and costs and to compare the importance of the model of safeguarding with other variables.

Data and method: The paper describes the findings of secondary analysis of local data in five councils that have adopted different models of adult safeguarding. Each of these five councils provided anonymous data on safeguarding referrals, which formed their Abuse of Vulnerable Adults (AVA, now known as the Safeguarding Adults Return or SAR) return to the Department of Health, and results of the Adult Social Care Survey (ASCS). The ASCS is run annually by councils and provides a measure of social care outcomes, using the Adult Social Care Outcomes Tool (ASCOT), and a single item indicator of Quality of Life. A tentative analysis of the possible different costs of different models is also presented. Finally the paper includes some of the results of an online survey of safeguarding practitioners undertaken as part of the Models of Safeguarding Study. Associations between the model of safeguarding and other important variables (numbers of referrals, kinds of alleged abuse and characteristics of adults at risk) and outcomes were explored. Associations are also explored between model of safeguarding and social workers’ job satisfaction and views about the effectiveness of safeguarding.

Results: Dispersed-Specialist sites appeared to have a higher rate of substantiating alleged abuse and lower costs compared with other models. However staff in the two sites operating this model reported lower job satisfaction compared with those in the Dispersed-Generic site. Statistical correlations with outcomes were found with types of victim profiles and the perpetrator/victim relationship.


Skip to toolbar