Authors' objectives: Health surveys often include comment boxes, called freetext questions. There is no easy way to analyse or summarise these. So NHS Trusts use them in ad hoc ways and they are not analysed at a national level. We plan to get the most from freetext with a novel approach. We use the cancer patient experience survey (CPES) to test this. CPES is sent over a 3 month period each year to all UK cancer patients receiving NHS inpatient or day case treatment, and its findings used to improve cancer care. CPES includes 3 comment boxes which get 70,000+ answers. Conventionally researchers analyse text by reading it and grouping it into common themes, which takes months and is labour intensive. A more automated approach, text mining , uses a computer to search or mine for themes, like a Google search. But it is designed for general use so the computer has to be taught to recognise specialised text (e.g. about cancer care); researchers must still group most of the data into themes as examples for the computer to compare with new text. So neither approach is suited to routine use. We plan to use an alternative called text engineering with a dashboard . Our study has 3 parts: the text engineering the dashboard a proof of concept study. 1: TEXT ENGINEERING is similar to text mining but it has benefits because the researcher modifies the search at the start by adding to the computer programme terms and rules relevant to the survey topic. We will ask a mix of stakeholders (patients, partners and carers, NHS staff) to send us ideas, e.g. outcomes of healthcare we might have overlooked, to incorporate. This means the programme looks for health-related themes from the start that moreover reflect the perspectives of patients and other stakeholders. This is not practical with other approaches, which are also slower. We then use our programme to analyse 2014 CPES freetext comments. We will check results against the researcher s common themes approach, and text mining, which we have already completed. We want to make sure that text engineering works effectively. 2. Like car dashboards our DASHBOARD will give a visual overview of complex data. We will use visuals to summarise themes from our novel analysis, and relevant data that NHS Trusts routinely collect. Each visual can be clicked on for more detail, including original freetext comments. To ensure our dashboard shows the right information in the right way to help improve service quality, we will ask our stakeholders to help us design it in structured workshop groups. This codesign means the dashboard will include what both patients and staff agree is important and relevant, so that it can be used effectively by both groups. We will include special questions about possible barriers to use, and interview up to 15 stakeholders 1-2 weeks later for further feedback. 3. PROOF OF CONCEPT: We will try our approach out on a new cancer survey and ask staff in 3 UK NHS Trusts (including London, Wessex) to talk about what they do as they try out our dashboard (a common technique in technology development). We will ask these staff if they would have issues using our approach at work and discuss solutions. We will report costs of our approach for cancer surveys and for adapting to other surveys and write instruction manuals so other people can use our process. We plan 4 dissemination events and articles for a wide range of audiences. Our team includes NHS managers to help with further evaluation and adoption of our approach.

Project Status: Completed

URL for project: http://www.nets.nihr.ac.uk/projects/hsdr/1415615

Year Published: 2019

URL for published report: http://www.nets.nihr.ac.uk/projects/hsdr/1415615

English language abstract: An English language summary is available

Publication Type: Not Assigned

Country: England, United Kingdom

Organisation Name: NIHR Health Services and Delivery Research programme

Contact Address: NIHR Journals Library, National Institute for Health and Care Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK

Contact Name: journals.library@nihr.ac.uk

Contact Email: journals.library@nihr.ac.uk

Copyright: Queen's Printer and Controller of HMSO