Required knowledge 151 4 Second, the literature review, including data extraction and generation of a draft list of knowledge components, was conducted by a single author. This may have introduced random error and risk of bias. However, the project team closely monitored this process and discussed the draft list of knowledge components in detail. Furthermore, the interviewees were invited to supplement the draft knowledge components with their own knowledge, derived from published evidence as well as from their own expertise and experience. Consequently, it is unlikely that the review would contain significant gaps in the published knowledge base. Third, we included a purposeful sample of nine experts in the field. However, these experts are internationally respected opinion leaders. They enlightened the topic of the study from different perspectives (clinical, consumer, researcher, methodologist, and teaching). Besides that, the experts complemented each other, and data saturation was reached. This resulted in a sound set of knowledge components required to develop healthcare related testing recommendations in guidelines. Fourth, we used a modified version of the cognitive domain of Bloom’s taxonomy. This domain consists of six levels (remember, understand, apply, analyse, evaluate, create) of which we used the first three and added a level zero ‘not necessary to know’. We deemed this sufficient since the scope of this study is restricted to guideline development and does not include the guideline methodology development, which would have required the latter levels of knowledge. We considered using other methods of distinguishing knowledge, such as ‘entrustable professional activities (EPA)’ and CanMeds [41, 42]. However, these methods highlight knowledge and competencies from the healthcare professionals’ perspective, which might put healthcare consumers in second place. Implications for practice Guideline panel members are often trained before or during their participation in a guideline panel. Examples are the INGUIDE course for guideline panel members and Dutch training programs about evidence-based guideline development [43]. We suggest extending available training programs with a module about the development of healthcare related testing recommendations. It can be confirmed that this work will be used to develop dedicated modules for INGUIDE (inguide.org), a Guidelines International Network led guideline credentialing and certification program. Such modules can be developed based on our findings and the scientific principles of developing educational courses.
RkJQdWJsaXNoZXIy MTk4NDMw