Table 1 Method of data analysis, modified from Braun and Clarke [19]

1. Familiarisation with the data

Check accuracy, transcripts read twice with and without concurrent audio file and notes made regarding potential coding ideas (JG)

2. Generating initial codes

Inductive, data driven, coding of transcripts, (JG, OM). No interpretation made, initial codes discussed with research team and attached to relevant sections of text to ensure context and facilitate emergence of semantic and latent codes. In cases of ambiguity, audio file was replayed and field notes consulted to better understand context

3. Searching for themes

Themes developed inductively to incorporate analysis across interviews, leading to preliminary themes and sub-themes. Hand sorting of individual codes to key areas of interest, produced preliminary thematic map. Iterative review clarified the relationships between themes and sub-themes

4. Reviewing themes

Preliminary themes checked (FJ) with text to ensure accuracy, refined by consensus discussion with research team and considered internal homogeneity and external heterogeneity, which shared analogous data whilst retaining independence [17]. Consensus amongst research team resolved incongruous findings (MH, FJ, AM), developed formal themes and sub-themes. Produced thematic map (JG). Employed a 2-phase approach, initially linking original text to themes and subsequently ensuring thematic map accurately reflects entire data set. Cognisant of thematic map 10 transcripts were re-read to ensure refinement process produced themes accurately reflecting data

5. Defining and naming themes

Final refinement of themes and sub-themes undertaken (JG) and discussed amongst research team to achieve consensus (MH, FJ, AM). Themes and sub-themes described and named. Achieved an accurate, clear, and balanced interpretation of the data [19,20,21].

6. Producing a report

Select compelling text examples to illustrate themes. Relate analysis to data and literature to produce report (JG)