Artikel
Empirical distribution of between-study heterogeneity tau from published IQWiG reports to inform Bayesian meta-analyses
Suche in Medline nach
Autoren
Veröffentlicht: | 26. Februar 2021 |
---|
Gliederung
Text
Meta-analysis is the method of choice in systematic reviews to summarize data quantitatively. According to the methods paper of the Institute for Quality and Efficiency in Health Care (IQWiG), version 5.0 [1], the Knapp-Hartung method is the standard approach to conduct meta-analyses with random effects [2]. However, if few studies are to be pooled (i.e., less than five studies), heterogeneity cannot be reliably estimated leading to broad confidence intervals. Bayesian methods are an alternative to conduct meta-analyses in such situations. In the literature, several proposals exist for a prior distribution of the between-study heterogeneity tau [3], [4], [5].
In order to inform prior choice of future Bayesian meta-analyses in the case of very few studies, we explore the empirical distribution of tau from published meta-analyses in IQWiG reports from 2005 to 2019. We focus on the recalculation of between-study heterogeneity tau by random-effects meta-analyses and the Paule-Mandel method. Empirical distributions are summarized in different settings such as type of the report, type of comparison and effect measure.
The authors declare that they have no competing interests.
The authors declare that an ethics committee vote is not required.
References
- 1.
- Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen. General methods: version 5.0. 10.07.2017 [Retrieved 06.06.2018]. Available from:https://www.iqwig.de/download/General-Methods_Version-5-0.pdf
- 2.
- Veroniki AA, et al. Recommendations for quantifying the uncertainty in the summary intervention effect and estimating the between-study heterogeneity variance in random-effects meta-analysis. Cochrane Database of Systematic Reviews. 2015;(1): 1-72.
- 3.
- Rhodes KM, Turner RM, Savović J, Jones HE, Mawdsley D, Higgins JPT. Between-trial heterogeneity in meta-analyses may be partially explained by reported design characteristics. J Clin Epidemiol. 2018;95:45-54. DOI: 10.1016/j.jclinepi.2017.11.025
- 4.
- Turner NL, Dias S, Ades AE, Welton NJ. A Bayesian framework to account for uncertainty due to missing binary outcome data in pairwise meta-analysis. Stat Med. 2015;34(12):2062-2080. DOI: 10.1002/sim.6475
- 5.
- Friede T, Röver C, Wandel S, Neuenschwander B. Meta-analysis of few small studies in orphan diseases. Res Synth Methods. 2017;8(1):79-91. DOI: 10.1002/jrsm.1217