A Systematic Approach to Improving the Transparency of Interview Refinements and Data Sufficiency in Semi-Structured Interviews: An Illustration Using the Mass Atrial Fibrillation Screening Study

Abstract

Background: Semi-structured interviews allow progressive interview refinements to gain an in-depth understanding of a research question, but interview refinements affect the stopping point. This paper aims to describe a method that provides transparency in interview refinements and helps researchers decide on a stopping point. Methods: We established systematic and practical methods to document interview refinements and visualize novel and duplicate interview codes. We illustrated the methods using a case study called the Mass Atrial Fibrillation Screening Study, a real-world clinical trial. Interviews were audio recorded and transcribed progressively. Two researchers established the initial coding into a predefined framework with reference to the Medical Research Council Guidance of Process Evaluation. Final codes were reached by discussion and consensus. NVivo v12 software was utilized to manage the qualitative data. Results: We developed an open-access software package for the ‘R’ statistical programming language called ‘novelqualcodes’ and a video demonstrating the application. Our package converts NVivo coding matrices into graphs of the trends of novel and duplicate codes. Our package also enables users to document interview refinements as fieldnotes, providing a visual guide to help researchers decide on a stopping point. In the case study, one researcher (KCW) conducted 48 one-to-one in-depth semi-structured interviews with participants from September 2021 to July 2022. Interview refinements were identified and reported to provide transparency to the readers. The novel code graphs and examination of the code sufficiency in the predefined framework demonstrated the stopping point at which we were satisfied with the richness and insights of the interviews in answering our research question. Conclusions: We illustrated a new paradigm of systematic and practical methods and shared an open-access software package and a video to describe, document and visualize the path to the stopping point. This approach can be applied in other qualitative research using different predefined frameworks.