OBJECTIVES: Concept elicitation (CE) interviews are commonly used in the development of clinical outcome assessments (COAs) to identify sign/symptom, impact, and treatment-related concepts that are relevant and important to assess from the patient perspective. The use of open-ended questions during CE often results in a large qualitative dataset representing a broad set of patient experiences. Our objective was to identify, assess, and recommend the most appropriate qualitative methods for analysis of CE data in COA development.

METHODS: Online hand-searches were conducted to identify published CE studies, and any existing guidance for analysing CE data, to determine the most frequently used and appropriate analysis approaches.

RESULTS: The review identified thematic analysis (TA) and qualitative content analysis (CA) as two predominant methods for CE data analysis. TA was not developed for CE analysis, so the techniques do not always align with the objectives of CE (e.g., TA discourages the use of saturation to justify sample sizes, and suggests limiting the number of themes which are constructed). Several elements conflict with FDA guidance for COA development, and restrict the ability to reflect the complete patient experience.

Qualitative CA, however, allows the researcher to adhere more closely to language used by participants, and report all relevant concepts. This is a critical step to support the content validity of fit-for-purpose COAs. As there is currently no specific method for analysis of CE data in COA development, Content Analysis for Concept Elicitation (CACE) was designed, consisting of the following steps: (1) immersion in the study, (2) coding, (3) iterative review of codes, (4) defining and refining concepts, and (5) reporting.

CONCLUSIONS: Although there are many published methods for analysis of large qualitative datasets, CACE is a patient-centred approach that can be used to analyse CE data in the context of COA development, while adhering to regulatory and wider industry guidance.