Mariska Tuut

106 Chapter 3 Identification of recommendations We analysed the content of the selected CPGs to identify relevant recommendations, including supporting evidence available online (e.g. tables with study characteristics, evidence documents, GRADE Evidence Profiles), as well as information about the methods of CPG development of the developing organisation (e.g. methodology manuals). Data extraction In the preparatory phase of this study, we piloted data extraction on two recommendations with four authors (MT, ML, JB, TvdW) to refine the data extraction form and define the variables for which we needed data extraction in duplicate. One author (MT) extracted the initial characteristics of each recommendation and CPG (CPG title (including English translation if relevant), initial developing organisation, country, publication year, recommendation text (including English translation if relevant)). Detailed information about each recommendation and CPG was extracted by one author (MT) and critically reviewed by another author (ML, JB or TvdW) using a predefined and piloted data extraction form (see Appendix 2 for the data extraction form and the categorisation of the variables). The form consisted of questions about scope and target audience of the CPG and composition of the CPG panel, involvement of methodologist(s), methodological quality of the CPG (using AGREE II, domain methodology, items 7-12) [17, 18], patient involvement (using AGREE II item 5) [17, 18], the types and extent of supporting evidence for the recommendation (consideration and inclusion of systematic evaluation with assessment of the certainty in the evidence about diagnostic accuracy, burden of the test, natural course, treatment effectiveness and link between test result and administration of treatment), grading of the recommendation and use of the GRADE approach, direction of the recommendation, and characteristics of the test and target condition. Disagreements between the reviewers were discussed until consensus was reached. Analysis We tabulated basic and detailed characteristics of the included recommendations and CPGs. Systematic evaluation (with a systematic review of the literature and assessment of the certainty in the evidence) of the impact of the full test-treatment pathway (diagnostic accuracy, burden of the test, natural course of target condition, treatment effectiveness, and link between test result and administration of treatment) on patient relevant outcomes was considered best practice for developing medical test recommendations.

RkJQdWJsaXNoZXIy MTk4NDMw