Required knowledge 143 4 Next step was to select abstracts from potentially relevant studies about knowledge needed to develop guidelines using Rayyan [15]. Then, full-text articles of selected abstracts were retrieved, and analysed and relevant data were extracted. Additionally, relevant items of the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy (draft version) and the GRADE (Grading of Recommendations Assessment, Development and Evaluation) Handbook were searched and adopted [16, 17]. Finally, ‘snowballing’ was applied to retrieve more relevant evidence. All steps of the literature review conducted by the first author (MKT) were discussed and approved by the other members of the project team (JSB, TvdW, MWL). Creation of a draft list of knowledge components Based on the results of the literature review, the first author (MKT) generated a draft list with specific knowledge components for the development of healthcare related testing recommendations in guidelines. General competencies for guideline development (such as the performance of a systematic review or group process techniques) and knowledge required to further develop guideline methods were not addressed. The draft list of knowledge components was discussed in detail within the project team for identifying overlapping items, deleting irrelevant items, and structuring the items into several groups. Semi-structured interviews The draft list of knowledge components was discussed in one-hour semi-structured online interviews. Relying on the Dutch and international network of the project team, we created a purposeful sample of nine internationally respected experts in the field of guideline development with specific knowledge and experience in: - Using tests in clinical practice - Test evaluation - Guideline development and GRADE for tests - Involvement of the public in guideline development, including testing recommendations - Training about guideline development concerning tests The interviewees received an information leaflet (Appendix 1) and the draft list (Appendix 2) beforehand. Two authors (MKT, MWL) undertook the interviews via Zoom, using an interview guide (Appendix 3) and tailoring the questions to the specific expertise of the interviewee. The interviews were videorecorded to facilitate data extraction and analysis.
RkJQdWJsaXNoZXIy MTk4NDMw