Using language analysis on patient experience feedback

  • Research type

    Research Study

  • Full title

    Listen, Learn & Improve: Using language analysis to interpret and act on written patient experience feedback for near real-time patient benefit

  • IRAS ID

    234218

  • Contact name

    Ruth Nicholson

  • Contact email

    jrco@imperial.ac.uk

  • Sponsor organisation

    Imperial College London

  • Duration of Study in the UK

    1 years, 2 months, 2 days

  • Research summary

    This project will be conducted in three key phases. The first phase will make use of natural language processing (NLP) along with machine learning (ML) to extract themes from free-text field data provided by patient comments in patient experience surveys. Although these free-text (unstructured) fields contain rich data in terms of experience, the data is often under-utilised due to the effort required to manually review the content. NLP provides a tool for translating human language into coded data so that it can be analysed using traditional and more advanced techniques. Using ML on this coded data will provide a source of data to further inform business intelligence and quality improvement based on direct patient feedback to help prevent future negative patient experience.

    Given that utilisation of NLP is a relatively new technique in clinical research it is important that the accuracy of results is directly validated. The second phase will explore views of healthcare staff about the usefulness of NLP, how the data should be presented and discuss quality improvement ideas from the existing data. The third phase will be a focus group involving patients, their family and/or carers, to determine patients opinions of providing feedback using free-text. Only those that would have access to the identifiable data as standard will do so before it is anonymised

  • REC name

    North East - Tyne & Wear South Research Ethics Committee

  • REC reference

    17/NE/0306

  • Date of REC Opinion

    14 Sep 2017

  • REC opinion

    Favourable Opinion