ESRC Logo

Can adaptive language save lives? Improving uptake of cervical screening in ethnic minority populations

Start date

November 2022

End date

March 2023

Overview

Ethnic minority populations are currently underrepresented in their engagement with NHS screening appointments. For example, engagement with cervical screening has been shown to be lower for eligible patients from BAME backgrounds.  

This ESRC IAA project is building a partnership to increase engagement in cervical screening of ethnic minority patients in Islington, London, through an intelligent appointment scheduling system developed by SPRYT. By investigating patient perceptions of and patient interactions with the virtual receptionist (called ‘Asa’) who represents the front-end of the system, the project will make proposals for how Asa can be optimised to better engage ethnic minority patients through adaptive language.

Team

Partners

This ESRC IAA Project is a partnership between The University of Surrey and Spryt.

Impact

  • Increased engagement with cervical screening by 10%: Minority groups were consistently associated with higher risk of no-show, as were patients from lower socioeconomic backgrounds (Peng,et-al, 2016; Torres et-al, 2015; Kempny et-al, 2016).  In North Central London, where the trial takes place, lack of uptake for cervical screening is particularly acute in some population segments, e.g. patients from a South East Asian ethnic background.  Impact will be measured as a change in the percentage of women engaging with cervical screening, comparing patients who were invited through traditional means (phone, internet) vs. patients who were invited through the new tool. We anticipate raising engagement with cervical cancer screening appointments by 10%. More long-term, and as part of a wider roll-out, we hope that this figure can be improved for patients across all ethnic groups.   
  • Reduction in admin costs: By employing an AI driven conversational agent, administrative time spent by the health centre will be reduced, allowing them to focus on value-add tasks. Asa can respond to questions, alleviating any of the patient's concerns. Impact will be measured in terms of freed admin time and the costs for this time.    
  • Reduction in DNAs (Did NOT attend): Currently the cost of DNA's is over £236 million in primary care [NHS England,2021], even a 20% reduction in these costs would equate to ~£250 million a year. A reduction in DNAs, facilitated through behavioural nudges that optimally target specific patient segments, can be measured in percentages, comparing patients who were invited for screening through traditional means (phone, internet) vs patients who were invited through the new tool.