More than words: A targeted review of qualitative data visualization in PROM development studies

ABSTRACT

Aims: In qualitative research, descriptive text with or without excerpts or patient quotes is used to highlight key findings. Data visualization techniques can be used to efficiently communicate insights from large amounts of text data. However, their use in the context of qualitative research for patient-reported outcome measure (PROM) development is still scarce. The primary purpose of this study was to review the recent literature to get a sense of the data visualization techniques most used to illustrate PROM development studies. Method(s) A PubMed search (2010-2021) with the search terms ''patient-reported outcomes'' or ''self-report measures'' and ''development'' was conducted to identify the top 10 journals (by frequency of relevant publications) that were not specific to therapeutic areas). Studies published in the aforementioned journals that reported qualitative methods to develop a PROM were included in the review. Studies that used qualitative approach to develop conceptual frameworks for quality of life or well-being relative to a health condition or treatment or overall, with the objective of developing a PROM were also included. The title and , full-text screening, and data extraction were completed in duplicate by two independent reviewers using the software package Covidence. Information pertaining to form (e.g., word clouds, bar charts) and the effectiveness (e.g., simplicity of visuals, captioning, use of text) of the visualization approach was extracted. Result(s) The search yielded 3816 unique s. After the eligibility criteria were applied, 355 were screened for full text, out of which 323 were included in the review. A total of 142 studies included some form of visualization of the qualitative data. We found bar charts were most used to describe the frequency of a concept being elicited in an interview at the level of the participant sample and few studies used flow charts to illustrate the relationship between concepts. Most studies failed to meet the good data visualization principles for effective communication of text data. Computational approaches to categorizing and visualizing text data (e.g., sentiment analysis, topic modeling) were largely underused. Conclusion(s) This review highlights the urgent need for using effective data visualization techniques for qualitative research for PROM development.