A0983
Title: Advancing statistical analyses for living systematic reviews
Authors: Lifeng Lin - University of Arizona (United States) [presenting]
Abstract: A living systematic review (LSR) is a progressive approach aimed at providing ongoing updates and instant synthesis of evidence. Trial sequential analysis (TSA) is an important tool for evaluating the sufficiency of evidence gathered in an LSR. It utilizes trial sequential monitoring boundaries to assess the effectiveness of an intervention and futility boundaries to determine if the intervention does not significantly differ from the control. While TSAs have been increasingly popular, their reproducibility is currently limited due to a lack of detailed information on their assumptions. Moreover, existing TSA methods face challenges due to their significant reliance on interim analyses of randomized controlled trials, which typically involve more homogeneous participant groups than those found in meta-analyses. The purpose is to introduce new methods aimed at preventing premature terminations of LSRs, thereby enabling more robust evidence syntheses. Numerical studies indicate that these proposed methods are more reliable than current methods.
