This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by Laopaiboon, M Search for Related Content PubMed PubMed Citation Articles by Laopaiboon, M Social Bookmarking                         What's this?

Meta-analyses involving cluster randomization trials: a review of published literature in health care

Department of Biostatistics and Demography, Khon Kaen University, Khon Kaen, Thailand, malinee{at}kku.ac.th

Throughout the 1980s and 1990s cluster randomization trials have been increasingly used to evaluate effectivenessof health care intervention. Such trials have raised several methodologic challenges in analysis. Meta-analyses involving cluster randomization trials are becoming common in the area of health care intervention. However, as yet there has been no empirical evidence of current practice in the meta-analyses. Thus a review was performed to identify and examine synthesis approaches of meta-analyses involving cluster randomization trials in the published literature. Electronic databases were searched for meta-analyses involving cluster randomization trials from the earliest date available to 2000. Once a meta-analysis was identified, papers on the relevant cluster randomization trials included were also requested. Each of the original papers of cluster randomization trials included was examined for its randomized design and unit, and adjustment for clustering effect in analysis. Each of the selected meta-analyses was then evaluated as to its synthesis concerning clustering effect. In total, 25 eligible meta-analyses were reviewed. Of these, 15 meta-analyses reported simple conventional methods of the fixed-effect model as method of analysis, while six did not incorporate the cluster randomization trial results in the synthesis methods but described the trial results individually. Three meta-analyses attempted to account for the clustering effect in the synthesis methods but approaches were in arbitrary. Fifteen meta-analyses included more than one cluster randomization trial, each of which included cluster randomization trials with a mixture of randomized designs and units, and units of analysis. These mixture situations might increase heterogeneity, but have not been considered in any meta-analysis. Some methods dealing with a binary outcome for some specific situations have been discussed. In conclusion, some difficulties in the quantitative synthesis procedures were found in the meta-analyses involving cluster randomization trials. Recommendations in the applications of approaches to some specific situations in a binary outcome variable have also been provided. There are still, however, several methodologic issues of the meta-analyses involving cluster randomization trials that need to be investigated further.

Statistical Methods in Medical Research, Vol. 12, No. 6, 515-530 (2003)
DOI: 10.1191/0962280203sm347oa


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				L. V. Hedges Effect Sizes in Cluster-Randomized Designs Journal of Educational and Behavioral Statistics, December 1, 2007; 32(4): 341 - 370. [Abstract] [Full Text] [PDF]

				G. A. Darlington and A. Donner Meta-analysis of community-based cluster randomization trials with binary outcomes Clinical Trials, October 1, 2007; 4(5): 491 - 498. [Abstract] [PDF]

				I. R White and J. Thomas Standardized mean differences in individually-randomized and cluster-randomized trials, with applications to meta-analysis Clinical Trials, April 1, 2005; 2(2): 141 - 151. [Abstract] [PDF]