Objectives:Systematic reviews (SR) of observational studies addressing patients' prognosis may provide robust estimates of the likelihood of adverse outcomes in broad populations or the role of patient’s characteristics in modifying their risk. As in any SR, consumers require estimates of the certainty of evidence (synonyms: confidence in or quality of evidence). Using GRADE methodology, we have presented structured approaches to rating certainty of evidence from prognostic studies, which have been adopted in several guideline development programmes. Work is in progress for studies of individual prognostic factors and prognostic models.
Description: In this workshop, we will summarise the approach to broad population prognosis and work through examples. We will begin with a description of how to apply five elements or rating certainty to prognostic studies: risk of bias, precision, consistency, directness and publication bias. We will explore the many aspects involved in incorporating evidence about risk factors in assessing effects in subgroups analysis. Finally, we will discuss how prognostic information can be incorporated in evidence profiles and evidence to decision tables, and ultimately applied to a patient management decision.
After successful completion of the workshop, participants will:
1. know the principles of the GRADE approach;
2. be able to prepare EPs and Summary of Findings tables for prognostic questions addressing broad populations.