Developing a Social Care Strengths and Vulnerability Index (SC-SVI) for older people

Being independent and managing well are important to older citizens and local authority social care organisations have a duty to actively encourage this. They try to do this by assessing the possibilities of harms to older people’s well-being and assets they have to overcome them. They then work out how these could be lessened by arranging services or professional help, keeping in mind older people’s wishes, choices and their own strengths.

This is not always easy.

Older people can under-report difficulties. Social care organisations may not pick up on older people’s circumstances that let them know in advance whether there might be threats to their wellbeing. Local authorities collect a lot of data: through early contacts with older people, social workers’ assessments and when they review the care they provide. But often this is not put to good use to judge potential future needs or risks. Data routinely collected by local authorities offers the potential of judging these threats to well-being, or picking up on strengths of older people that could lessen these threats.

This project aimed to discover if data routinely held by local authority social care organisations, could be used to help identify older people’s needs and potentially judge future threats to their well-being. 
 

• Aims

  Aim: to discover if and how a prediction model for social care outcomes, using routine data, could be developed, and used to identify older people’s needs and potentially judge threats to their wellbeing

  Objectives: to discover if:

  • the right data, to potentially predict older people’s vulnerability to adverse outcomes or evidence their strengths in lessening these, were present in local authority systems;
  • accurate, individual-level, datasets for large numbers of older people, representative of those receiving social care, could be constructed to allow robust statistical modelling;
  • robust predictive models, statistically predicting outcomes and tested for assumptions, missing data and model fit, could be developed and validated;
  • findings could be used to enable a prediction algorithm, to be made available, as an index for social care, that local authorities could use in planning more appropriate care around older people’s circumstances, resources, and strengths.

   

• Methods

  In order to meet the research objectives, we:

  • Unlocked the data held on older service users’ circumstances and needs and what happens to them two years from assessment.
  • Collated and analysed a range of data items – in a safe and secure way – from data held by the councils.
  • Explored the issues, challenges and opportunities of doing this – for council social care staff, professionals and service users and for creating an index from this analysis to inform practitioners and managers of the likelihood of older people either remaining at home or entering care homes.

   

• Findings and resources

  • Read the full findings report – Developing a Social Care Strengths and Vulnerability Index (SC-SVI) for older people
  • Academic poster – Developing a strengths and vulnerability index for older people using routine data in Adult Social Care
  • FOR DATA’S SAKE | An Interactive Event Exploring A New Social Care Index (including audio recordings of stakeholder feedback)

  We developed and validated statistical models from routine Adult Social Care datasets in two local authorities in Greater Manchester. From these, we provided robust, well-calibrated models for predicting which sorts of older people’s circumstances predicted the likelihood of them remaining at home two-years after initial assessment. We found circumstances predicting strongly that older people would not remain at home, and enter care homes, were that they belonged to ethnic groups other than Asian British or Black British, that their Primary Support Reason was for personal care, mental health, or support with memory and cognition, that they were privately rented tenants, and that they had no access to a carer. 

  We were able to develop a prototype index from these analyses that the councils and other local authorities nationally, could potentially use to help inform future professional decisions, service configuration and/or local planning. User, practitioner and manager views on such an index were that it was broadly acceptable but that it must be based on the right data, that the data and analysis was trusted, and that the data was secure.

  Adult Social Care practice in England does not presently capitalise enough on the wealth of routine data existing and generated within local authority Adult Social Care. Our study showed that much of the data most useful in predicting what might happen to older users in their care trajectories is currently collected. These data items, including those of Primary Support Reason, carer access, accommodation-type, and ethnicity, are those now mandated in the Client Level Data (CLD) set, and so will be collectable by local authority performance and intelligence teams. The CLD is now planned to be available nationally, also for researchers to access and use. Therefore, this can be used in future to undertake more sophisticated analyses on users’ journeys through the care system to inform evaluations of how councils are responding to their Care Act responsibilities.