Trial Sequential Analysis: An approach to furthering the interpretation of ‘significant’ and ‘non-significant’ meta-analyses‘

Background: The problem of multiple testing is a well-known concept when analysing data from clinical trials and observational studies, and many statistical approaches have been developed to adjust for multiple comparisons. The issue of multiple testing is also relevant to meta-analysis, particularly when systematic reviews and their associated meta-analyses are updated. However, until recently the issue was often overlooked within meta-analyses. Trial Sequential Analysis (TSA) is one approach that can be used to adjust for multiple testing in meta-analyses. Objective: This presentation will describe the problem of multiple comparisons in meta-analyses, and use contemporary examples to demonstrate the usefulness of TSA in establishing firm conclusions from a meta-analysis, and how TSA can be used to determine the sample size of future study to enable the meta-analysis results to be conclusive. International Journal of Evidence-Based Healthcare (C) 2016 The Joanna Briggs Institute
Source: International Journal of Evidence-Based Healthcare - Category: Nursing Tags: Abstracts of Oral Presentations: Evidence Synthesis Source Type: research