Data analysis

Data analysis was conducted in several stages, informed by the RREAL methodology for rapid assessment procedures (RAP).²¹ The RAP process is considered an appropriate method to conduct rigorous and robust qualitative data analysis in a short amount of time. It is predominantly team based and collaborative in nature, and conducted as an iterative process alongside data collection. Data are independently reviewed by multiple team members, before being brought together in a group session for further analysis. RAP is particularly suitable for this type of research, which aims to synthesise diverse perspectives on a defined topic, and is focused on generating useful learning that can be actioned. While it was felt that ‘data saturation’²² was being reached at organisations, it is evident that results are intrinsically linked with the sample of people that we spoke to at each hospital.

Following the first stage of data collection, three further stages (multiple RAP sheet generation, individual consolidation and group confirmation) completed the RREAL process. After each interview, focus group, or debrief, a structured RAP sheet (online supplemental file 4) outlining the main themes was completed by the researcher leading the data collection. One or more members of the research team then listened to the recording, and independently completed a RAP sheet. RAP sheet generation occurred simultaneously with data collection, with discussion occurring after data collection had been completed. A minimum of two standardised RAP sheets was completed for each interview/focus group, and used to guide our reflexive analysis process. In total, 171 RAP sheets were generated. Researchers then read and reviewed all the RAP sheets generated, and completed independent informal notes to support synthesis. Two members of the team independently generated further ‘site-specific’ RAP sheets, providing an overview of key themes at each organisation. This synthesis process focused on considering the specific context of each organisation as well as the broader project research questions.

bmjopen-2023-081056supp004.pdf

We then conducted several day-long group debriefing and analysis sessions, in which we focused on identifying key themes from RAP sheets. This process led to the creation of brief data summaries that were then used iteratively for cross-case comparison. The whole team contributed to these sessions, which aimed to establish consensus, while still being sensitive to diversity of views and experiences in different organisations. Each analysis session ended with confirmation and agreement of key themes.