Rapid prioritisation of topics for rapid evaluation: the case of innovations in adult social care and social work.

Background: Prioritisation processes are widely used in healthcare research and increasingly in social care research. Previous research has recommended using consensus development methods for inclusive research agenda setting. This research has highlighted the need for transparent and systematic methods for priority setting. Yet there has been little research on how to conduct prioritisation processes using rapid methods. This is a particular concern when prioritisation needs to happen rapidly. This paper aims to describe and discuss a process of rapidly identifying and prioritising a shortlist of innovations for rapid evaluation applied in the field of adult social care and social work.

Method: We adapted the James Lind Alliance approach to priority setting for rapid use. We followed four stages: (1) Identified a long list of innovations, (2) Developed shortlisting criteria, (3) Grouped and sifted innovations, and (4) Prioritised innovations in a multi-stakeholder workshop (n = 23). Project initiation through to completion of the final report took four months.

Results: Twenty innovations were included in the final shortlist (out of 158 suggested innovations). The top five innovations for evaluation were identified and findings highlighted key themes which influenced prioritisation. The top five priorities (listed here in alphabetical order) were: Care coordination for dementia in the community, family group conferencing, Greenwich prisons social care, local area coordination and MySense.Ai. Feedback from workshop participants (n = 15) highlighted tensions from using a rapid process (e.g. challenges of reaching consensus in one workshop).

Conclusion: The method outlined in this manuscript can be used to rapidly prioritise innovations for evaluation in a feasible and robust way. We outline some implications and compromises of rapid prioritisation processes for future users of this approach to consider.