gms | German Medical Science

63. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

02. - 06.09.2018, Osnabrück

Shrinkage estimation in meta-analyses of two studies

Meeting Abstract

Search Medline for

  • Christian Röver - Universitätsmedizin Göttingen, Göttingen, Deutschland
  • Tim Friede - Universitätsmedizin Göttingen, Göttingen, Deutschland

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 63. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Osnabrück, 02.-06.09.2018. Düsseldorf: German Medical Science GMS Publishing House; 2018. DocAbstr. 161

doi: 10.3205/18gmds046, urn:nbn:de:0183-18gmds0460

Published: August 27, 2018

© 2018 Röver et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at



Meta-analytic methods may be used to combine evidence from different sources of information. Quite commonly, the normal-normal hierarchical model (NNHM) is utilized for such analyses. This modeling framework may also be used to not only derive a joint outcome, but to also lend support a particular case within the set of combined estimates by deriving a shrinkage estimate [1], [2]. Especially in the context of rare diseases, supporting evidence for a small-scale randomized controlled trial may come from only a single external source, e.g. a non-randomized study or clinical registry [3]. We investigate to what extent performing a meta analysis still makes sense in this extreme two-study setup, as illustrated using a recent trial in Creutzfeld-Jakob disease (CJD). Depending on the exact circumstances, derivation of a shrinkage estimate within a Bayesian random-effects meta-analysis may substantially improve a given estimate even based on only a single additional estimate while providing robust inference. The use of shrinkage estimates from a 2-study meta analysis can be shown to be equivalent to a similar, "asymmetric" model formulation that may in many cases be easier to motivate.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.


Wandel S, Neuenschwander B, Röver C, Friede T. Using phase II data for the analysis of phase III studies: an application in rare diseases. Clinical Trials. 2017;14(3):277–85. DOI: 10.1177/1740774517699409 External link
Röver C. Bayesian random-effects meta-analysis using the bayesmeta R package. arXiv:1711.08683. 2017. Available from: External link
Friede T, Röver C, Wandel S, Neuenschwander B. Meta-analysis of two studies in the presence of heterogeneity with applications in rare diseases. Biometrical Journal. 2017;59(4):658–71. DOI: 10.1002/bimj.201500236 External link