Data processing was limited to anonymisation and preparation for sharing. Participant names were replaced with numeric participant identifiers throughout the dataset. An initial find-and-replace process was followed by a manual review of every transcript to identify and replace any remaining references to participant names with the corresponding participant number. The transcripts were also checked for other direct name references that could disclose an individual’s identity. Questionnaire data was provided as an Excel export generated directly from Microsoft Forms. Other than replacing participant names with the corresponding numeric identifiers, the questionnaire data has not been reformatted, recoded, imputed or otherwise cleaned. Original question wording, response values, column headings and blank or missing responses have been retained as captured by Microsoft Forms. No substantive changes were made to participants’ responses or to the content of the transcripts. The numeric identifiers are used consistently across the questionnaire data and transcripts to allow records relating to the same participant to be associated without disclosing their name. The document linking participant names to these identifiers is not included in the shared dataset.