66 Chapter 3 ROBIS tool for systematic reviews and meta-analyses 19,20. Bias was visualized with risk-of-bias visualization tool in RevMan Web. After summarizing and presenting all the evidence related to each question, evidence-based statements were formulated by each WG to address all ALrelated questions. The formulation of the statements was based on the level of evidence (LoE) according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) guidelines, rated as ‘high’, ‘moderate’, ‘low’, or ‘very low’ 21. Standardized wording used to phrase statements is shown in Table 1. When the LoE was very low, the research team did not formulate a statement but flagged the topic as lacking evidence. Table 1. Phrasing of the statements based on the Level of Evidence (LoE) LoE according to GRADE Wording statement High Does (not) Moderate Probably does (not) Low May do (not) Very low / expert opinion No statement formulated Consensus on drafted statements A two-phase modified Delphi method, consisting of an online survey and an in-person consensus meeting, was used to achieve consensus on all evidence-based statements. In the first phase, the statements from each WG with LoE, were presented to all 32 team experts who subsequently voted online on each of these statements using a 9-point Likert scale, with consensus defined as >70% agreement (a score of ≥7 on the Likert scale). In the second phase, statements that did not reach consensus were discussed on Day 1 of a 2-day in-person consensus meeting held in Boston in October 2023. Research collaborators presented the data analysis for all statements that did not reach consensus and facilitated discussions and rephrasing of the statements among experts. Another round of voting was carried out with consensus achieved with >70% agreement. Patient engagement Patients’ personal experience with AL were captured across all the phases of patient journey through semi-structured interviews and a qualitative synthesis 22. In-depth semi-structured patient interviews were conducted among 10 patients who experienced an AL. To obtain a wide range of perspectives, a maximum variation sampling strategy was employed, ensuring diversity in age, sex, and the severity of AL. The results of this qualitative analysis were presented during the in-person consensus meeting and incorporated into the reporting framework. Development of the reporting framework and consensus process The reporting framework encompassing reporting elements along all four phases of the AL episode, was constructed from the statements that achieved consensus, with input derived from qualitative analysis of patients’ interviews as well as from experts. The framework was
RkJQdWJsaXNoZXIy MTk4NDMw