DOP02 Developing explicit thresholds for outcomes to inform (GRADE) Evidence to Decision frameworks for Inflammatory Bowel Disease guidelines

Abstract

Background
Evidence to decision (EtD) frameworks are a key recent development in Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) methodology. They aim to ensure that decisions about healthcare interventions are based on a thorough, transparent and unbiased assessment of the available evidence. Core to this is to agree the importance of outcomes to be used, the specific measures for each outcome that have the most consensus and finally to develop explicit thresholds for interpretation of effect sizes of outcomes to inform (GRADE) evidence to decision frameworks. This ensures decisions are made in an unbiased fashion and allows stakeholders to focus on discussions based on their significant experience and expertise, rather than interpretation of evidence. Therefore, our study aims to develop these thresholds for the EtD framework to establish their use in several key IBD guidelines.

Methods
This online survey using Survey Monkey was conducted by inviting experts and stakeholders from the United Kingdom and British Society of Gastroenterology. Each expert was asked to select important clinically relevant outcomes and were asked about what magnitude of the effect they consider large, moderate, small, or trivial for each of the clinical, endoscopic, radiological, biochemical, histological outcomes and adverse events. Questions were framed as neutral statements without introducing a specific direction. Response options included sliding bars going from 1 to 100% for the desirable magnitude of effect for each outcome (Figure 1). The mean average of responses of experts were calculated.

Results
A total of 89 clinical experts/stakeholders participated in this online survey. Clinical remission, clinical response, endoscopic remission, withdrawal due to adverse events, and serious adverse events were considered critical outcomes. Trivial to small, small to moderate and moderate to large thresholds were as follows - clinical remission (10%, 20% and 31%), clinical response (13 %, 24% and 36%), endoscopic remission (9%, 17% and 28%), withdrawal due to AE (7 %, 14% and 23%), serious AE (6%, 11% and 17%) (Table 1).

Conclusion
This is the first study to develop thresholds for outcomes in IBD which can be used in EtD for the development of IBD guidelines. These thresholds have been used in the development of upcoming British Society of Gastroenterology guidelines for the management of IBD.