User feedback on the NHS Test & Trace Service during COVID-19: the use of machine learning to analyse free-text data from 37,914 UK adults (preprint)

Objectives The UK government's approach to the pandemic relies on a test, trace and isolate strategy, mainly implemented via the digital NHS Test & Trace Service. Feedback on user experience is central to the successful development of public-facing services. As the situation dynamically changes and data accumulate, interpretation of feedback by humans becomes time-consuming and unreliable. The specific objectives were to 1) evaluate a human-in-the-loop machine learning technique based on structural topic modelling in terms of its serviceability in the analysis of vast volumes of free-text data, 2) generate actionable themes that can be used to increase user satisfaction of the Service. Methods We evaluated an unsupervised Topic Modelling approach, testing models with 5-40 topics and differing covariates. Two human coders conducted thematic analysis to interpret the topics. We identified a Structural Topic Model with 25 topics and metadata as covariates as the most appropriate for acquiring insights. Results Results from analysis of feedback by 37,914 users from May 2020 to March 2021 highlighted issues with the Service falling within three major themes: multiple contacts and incompatible contact method and incompatible contact method, confusion around isolation dates and tracing delays, complex and rigid system. Conclusions Structural Topic Modelling coupled with thematic analysis was found to be an effective technique to rapidly acquire user insights. Topic modelling can be a quick and cost-effective method to provide high quality, actionable insights from free-text feedback to optimize public health services.

Applying machine-learning to rapidly analyse large qualitative text datasets to inform the COVID-19 pandemic response: Comparing human and machine-assisted topic analysis techniques (preprint)

Background: Machine-assisted topic analysis (MATA) uses artificial intelligence methods to assist qualitative researchers to analyse large amounts of textual data. This could allow qualitative researchers to inform and update public health interventions 'in real-time', to ensure they remain acceptable and effective during rapidly changing contexts (such as a pandemic). Objective: We aimed to understand the potential for such approaches to support intervention implementation, by directly comparing MATA and 'human-only' thematic analysis techniques when applied to the same dataset (1472 free-text responses from users of the COVID-19 infection control intervention 'Germ Defence'). Methods: In MATA, the analysis process included an unsupervised topic modelling approach to identify latent topics in the text. The human research team then described the topics and identified broad themes. In human-only codebook analysis, an initial codebook was developed by an experienced qualitative researcher and applied to the dataset by a well-trained research team, who met regularly to critique and refine the codes. To understand similarities and difference, formal triangulation using a 'convergence coding matrix' compared the findings from both methods, categorising them as 'agreement', 'complementary', 'dissonant', or 'silent'. Results: Human analysis took much longer (147.5 hours) than MATA (40 hours). Both human-only and MATA identified key themes about what users found helpful and unhelpful (e.g. Helpful: Boosting confidence in how to perform the behaviours. Unhelpful: Lack of personally relevant content). Formal triangulation of the codes created showed high similarity between the findings. All codes developed from the MATA were classified as in agreement or complementary to the human themes. Where the findings were classified as complementary, this was typically due to slightly differing interpretations or nuance present in the human-only analysis. Conclusions: Overall, the quality of MATA was as high as the human-only thematic analysis, with substantial time savings. For simple analyses that do not require an in-depth or subtle understanding of the data, MATA is a useful tool that can support qualitative researchers to interpret and analyse large datasets quickly. These findings have practical implications for intervention development and implementation, such as enabling rapid optimisation during public health emergencies.