Optimising data sharing whilst protecting participant privacy: a data note describing processed data from a qualitative study of healthcare professionals' experiences of caring for women with false positive screening test results

Abstract

Introduction: The present article describes the processed data generated in a qualitative interview study and template analysis. Many women find the experience of being recalled and receiving a false-positive breast screening test result to be distressing. The interview study aimed to understand breast screening healthcare professionals' (HCPs) experiences of providing care during the recall process and when receiving false-positive screening test results, including their communication with women around false-positive screening test results.

Methods: Twelve HCPs from a single screening unit in the English National Health Service Breast Screening Programme participated in semi-structured interviews in 2020. All participants were female. A range of HCPs roles were recruited, including advanced radiographer practitioners, breast radiographers, breast radiologists, clinical nurse specialists, and radiology healthcare assistants. The data were analysed thematically using template analysis from a limited realist perspective.

Results: A total of 20 data files are described, reflecting the iterative nature of template analysis. The files report various versions of codes, subthemes, themes, and every template produced during analysis. The files are publicly available on the Open Science Framework and UK Data Service (ReShare).

Discussion: This data note outlines our approach to conducting a template analysis of qualitative data while protecting highly identifiable data, which is stored in a non-public archive and only available to the study team. It offers a practical, worked example of the template analysis process, thereby providing a detailed illustration beyond the concise summaries typically found in published reports, and complementing methodological papers of template analysis.

Keywords: Qualitative data; breast screening; false-positive test results; open data; template analysis.

Figure 1.

An example of preliminary coding on an extended excerpt of anonymised interview data from the first subset of interview transcripts. These codes form the basis of parts of Theme 1 within the developing template. The full list of preliminary codes can be found in archived file 1 Preliminary coding 1 on the OSF and ReShare. This Figure has been retrospectively created for the purposes of illustrating coding within the present Data Note.

Figure 2.

An example section of Theme 1 from the entire initial template (which includes all preliminary overarching themes and all lower-level subthemes and codes). This figure indicates initial clusters of codes and provides the names given to represent the theme and subthemes at this stage in the analysis. The full (10 page) version can be found in archived file 8 Template 3 longer_revised on the OSF and ReShare.

Figure 3.

An example section of the first near-final Theme 1, with the overarching theme name, subthemes, and all relevant individual codes. The full version can be found in archived file 18 Template 5_longer on the OSF and ReShare. While it may seem that the template structure has shortened and simplified over time (e.g. since Figure 2), it is not the case. With close examination and analysis of the data, and continued discussion among the study team, the coding increased in depth, representing increased discrimination and clarity in our thinking behind the data. Template 5 shows the first ‘top level’ theme (‘Gauging and navigating women’s anxiety’) for the final version of our template.

Figure 4.

An example section of Theme 1 in the final template. The full version can be found in archived file 20 Template 7 on the OSF and ReShare. This template structure is again shorter than the previous structures. Depth (rather than breadth) of coding allows fine distinctions to be made in key areas; having too many themes and subthemes makes it difficult for the researcher to draw together the analysis as a whole (Brooks et al., 2015).