XVI.Add.II.2.1. Data sources for qualitative research

Common data sources for qualitative research in healthcare are interviews, focus groups and different existing types of documentations (e.g. media reports or clinical guidelines). These data sources may contain information about cognitive processes and experiences of patients and healthcare professionals that provide in-depth understanding of the reasons for RMM achieving (or not) the intended outcomes.

The type of documentation to use as data source for understanding perception and information needs in certain patient or healthcare professional populations may be determined by their media preferences. Preferences for e.g. news, social or scientific media can be identified through qualitative or quantitative media research.

The recruitment of participants in focus groups or interviews, and the selection of documentation is aimed at saturation of data, i.e. all viewpoints, experiences, and suggestions prevalent in the concerned patient or healthcare professional population(s) are collected and the collated data provide for a robust understanding of the cognitive processes and experiences that are typical in the population of interest. This includes covering less common views or needs of sub-populations of patients and healthcare professionals. Therefore, diverse participants should be selected to provide in-depth insights. Appropriate sampling is a key requirement to obtain relevant information and to minimise bias. The sampling strategy’s target is relevance and representativeness of the information to be collected, and various strategies can be applied: representative sampling in relation to certain criteria describing the population of interest, oversampling of specific subpopulations or complete sampling to include all concerned people within a defined region or timeframe, or step-by-step sampling to identify all themes or investigate emerging themes more in depth (5–7). Alternatively, purposive sampling is a non-random method where researchers use their expertise to identify and select participants who provide informationrich cases, making best use of available resources (8). However, where the sampling strategy leads to non-representative sampling, the results need to be interpreted in a qualitative setting, i.e. they cannot be interpreted like the results of a representative study. Finally, the appropriate sampling strategy should be adapted to the diversity of the patient or healthcare professional population of interest and recruit also those who may be less proactive to participate in such research.

Data collection through interviews or focus groups should preferably use open questions and can be conducted with variable degrees of structure, depending on the study objective and the available evidence on the topic to be studied (9–11). Studies should be conducted to standards that avoid response bias, e.g. where questionnaires or semi-structured interview guides are used, these should be validated, and interviewers should be experienced.