| Popis: |
Background and Objectives: The average treatment effect estimate from a randomized controlled trial (RCT) may not be valid for deciding what treatment is best for individual patients with specific characteristics. Characteristics may include contextual factors related to the person (e.g. age), disease (e.g. disease duration), or environmental (e.g. healthcare system). However, individual RCTs rarely report results stratified by potentially important contextual factors in a consistent manner. In this study we will summarize existing guidance from key organizations and other research initiatives, and subsequently demonstrate how subgroup results from RCTs can be presented in a consistent manner that enables future evidence synthesis. Study Design and Setting: We will use a tutorial-type approach to explain and demonstrate the recommended presentation of subgroup results from RCTs. In part 1, we will systematically review papers and official documents. Relevant guidance will be identified via a broad search strategy, including searching MEDLINE (via PubMed) and Web of Science, reviewing official websites of FDA, EMA, ICH, Equator Network etc., ‘asking around’, and snowball methods. Recommendations will be extracted and summarized. In part 2, we will demonstrate the recommendations in practice. We will use data from four previously published RCTs of rheumatic and musculoskeletal conditions from the Parker Institute, Bispebjerg and Frederiksberg Hospital, Denmark. For each RCT, we will reanalyze the data by stratifying according to proposed potential candidates of effect modifying contextual factors and apply the reporting guidance described in part 1. Dissemination and perspectives: This study will be disseminated in a peer-reviewed publication and conference presentations. It is anticipated that this tutorial can guide researchers in reporting trial results stratified by effect modifying contextual factors in a consistent manner, allowing for investigation of subgroup effects in future meta-analyses, potentially leading to more patient-specific evidence-based management of patients with rheumatic conditions. |
