Abstract
Within Great Ormond Street Hospital for Children, (GOSH), valuable patient feedback is collected via the Friends and Family Test (FFT). GOSH receives 2,000 comments per month from inpatient and outpatient areas and 85% of the feedback contains unstructured free text.
The GOSH Patient Experience Team (PET) processes the feedback, manually applying sentiment and themes to each qualitative comment. Trends and themes are analysed and reported on from all wards and departments. This process is extremely time consuming, resulting in feedback trends being shared in the Trust retrospectively.
Consequently, GOSH PET applied as the only Paediatric Trust to be involved in the ‘Scale Spread & Embed’ project, led by Imperial College London, and funded by the Health Foundation. The project objective is to use Natural Language Processing (NLP), a branch of Artificial Intelligence (AI), to transform unstructured text into a structured format, enabling faster analysis. The PET felt it was vital to be involved to improve the resolution period for any negative feedback, as quicker resolution of any issues would result in a better experience for GOSH patients and their families.
One of the project elements is the creation of bespoke dashboards that are uncomplicated and user-friendly containing ‘real-time’ feedback data. These will be accessible to all GOSH staff and enable quick identification of any negative themes allowing for remedial actions to be carried out as necessary. Another important component of the dashboards is a spotlight on the positive feedback, which constitutes 95% of the feedback received, cultivating staff morale. Additionally, feedback will be utilised to identify any Quality Improvement projects, specifically within the new Q-TEAMS meetings.
This project has brought together experts from teams across GOSH, Patient Experience, Quality Improvement, Information Services, DRIVE, Ward and Department Managers and ICT who have all played a key role in the project’s advancement.
Acknowledgements for funding or support This work is supported by Imperial College, London and the Health Foundation. The views expressed are those of the author(s) and not necessarily those of the Imperial College London or the Health Foundation.
