Evidence-based mapping of design heterogeneity prior to meta-analysis: a systematic review and evidence synthesis

摘要: Assessment of design heterogeneity conducted prior to meta-analysis is infrequently reported; it often presented post hoc explain statistical heterogeneity. However, determines the mix included studies and how they are analyzed in a meta-analysis, which turn can importantly influence results. The goal this work introduce ways improve assessment reporting summarization epidemiologic studies. In paper, we use an sugar-sweetened beverages (SSB) type 2 diabetes (T2D) as example show technique called ‘evidence mapping’ be used organize evaluate meta-analysis.. Employing systematic reproducible approach, evaluated following elements across 11 selected cohort studies: variation definitions SSB, T2D, co-variables, features population characteristics associated with specific diversity modeling strategies. Evidence mapping strategies effectively organized complex data clearly depicted For example, SSB 7 measured diet only once (with 16 years disease follow-up), 5 primarily low consumers, 3 defined study variable consumption either sugar or artificially-sweetened beverages. This exercise also identified analysis strategies, such adjustment for 17 co-variables large degree fluctuation SSB-T2D risk estimates depending on variables multivariable models (2 95% change estimate from age-adjusted model). Meta-analysis seeks understand addition computing summary estimate. strategy documents heterogeneity, thus improving practice by aiding in: 1) protocol planning, 2) transparent differences designs, 3) interpretation pooled estimates. We recommend expanding include table that summarizes would provide readers more evidence interpret