Table 1.

Description of the rapid analysis and thematic analysis

		Rapid analysis	Thematic analysis
The researchers		Clinical background.	Not clinical.
		Embedded in the field.	No prior exposure to field.
		First time using RA.	Experienced in TA – no need to ‘learn’.
		Shared office, opportunity to interact.	No informal interaction/reflection.
		Evenly shared workload.	IL Conducted most of the analysis.
		Main focus of work.	Conducted alongside other core work.
		Conducted over short period.	Project delivered over a longer period.
		Focused on producing and ‘crafting’* outputs for known stakeholders.	Less focused on the needs and expectations of stakeholders.
Epistemological position		Theoretically interpretive, generic qualitative approach.	Theoretically interpretive, generic qualitative approach.
Data collection		Interviews, focus groups conducted, documents gathered from participants by BT and CH facilitating focus group.	Provided with pregathered dataset.
Transcription		Audio recordings transcribed by third party. Transcripts checked for accuracy by researcher. Participant name retained in transcript.	Pseudoanonymised transcripts and documents provided.
Timing		During and following data collection process.	Analysis conducted postdata collection.
Ordering		Interviews, then focus groups, then documents. Strategic participant data analysed first.	Documents, then interviews with strategic participants first, finally the focus groups.
‘Data management and review’ stage	Early analysis	[researchers 1 and 2] dual analysed one interview transcript, inserting them into a ‘summary template’, organised according to the research objectives (see online supplementary file 1). They compared template content. The process and the template structure were reviewed and amended (some subheadings applied). A second transcript was subsequently dual analysed in the same way. The ‘one hour per transcript’ rule was not applied here, spending 1.5–2 hours on each.	An identical sample of three transcripts reflecting a range of job title and seniority were analysed independently by [researchers 4 and 5] and the themes that emerged discussed and finalised. These themes were arranged into analytical hierarchies, that is, consisting of the key themes and associated subthemes, and these formed the basis for the codification of the remaining data.
	Main analysis	Remaining data items allocated equally to [researchers 1 and 2], following the same process, limiting time to 1 hour maximum per data item (less for some less complex documents). Researcher entered information directly into a matrix, structured as the template, using individual templates duplicated work.	[Researcher 4] independently analysed the remainder of the transcripts and the resulting themes and subthemes were agreed with SG and formed the analytical hierarchy for the remaining data.
‘Interpretation’ stage	Data interpretation	[Researchers 1 and 2] reviewed content in one another’s matrices, and combined them. Data were allocated equally to [researchers 1 and 2] for interpretation and write up, organised according to the template, for example, facilitators to implementation. The ‘barriers’ section was more complex, and this was subdivided into themes, which were allocated to [researchers 1 or 2], for example, training, promotion and recruitment. A summary of findings and a set of recommendations were produced for each. Summaries were reorganised thematically.	IL undertook interpretation and write-up of the findings according to the thematic headings. For each theme and subtheme, an explanatory sentence was produced, and an exemplar quote or quotes was selected. These themes and subthemes were used to create a list of findings specific to each overarching theme.
	Final report writing	Summaries of findings and recommendations were combined and checked by [Researchers 1, 2 and 3] to eliminate duplication and reach consensus regarding interpretation, revisiting the primary data where necessary.	These findings were used to inform a final report, populating the template provided by BT. The report template included the following headings: Participants and data (not written up in secondary analysis). Timeline for development of service. Service design (logic models developed for intended and actual service design). Achievements. Challenges Barriers to implementing the model as intended. Barriers to delivering specific service outcomes. Service data Data being gathered. Responsibility for data collection/entry/analysis. What is going well in Home Birth Service (HBS) data capture and use. Data-related challenges. Recommendations.
Researcher interaction		[Researchers 1, 2 and 3] reflected and discussed the data and interpretation on a regular, iterative basis.	[Researchers 4 and 5] had several telephone and one face-to-face discussion.

*‘Crafting’ refers to the writing and editing of findings and recommendations to present content and language deemed to be appropriate to the service stakeholders by the rapid analysis team.

HB, Home Birth Service.