Berlin JA; International Cochrane Colloquium (8th : 2000 : Cape Town, South Africa).
Evid Action Int Cochrane Colloq 8th 2000 Cape Town S Afr. 2000; 8: UNKNOWN.
University of Pennsylvania, Philadelphia, USA.
Heterogeneity of both study design and study results seems to be the rule in meta-analyses, rather than the exception. Heterogeneity of design, broadly defined, can come from inclusion of different study populations, use of different treatment protocols (e.g., different doses or duration of therapy), or different methods of treatment allocation, to name just a few. Such heterogeneity, often vaguely defined by critics of meta-analysis, is sometimes viewed as a weakness of the meta-analytic approach ("You are not just mixing apples and oranges, you are making fruit salad.")The theme of this talk will be that heterogeneity is, in fact, a strength of meta-analysis, and can be used to our advantage. Careful analyses of the relationships between aspects study design and the magnitude and direction of treatment effects, can potentially clarify the clinical implications of meta-analyses and/or help improve the design of future studies. This talk will use several examples from meta-analyses of clinical trials to illustrate the potential benefits of heterogeneity.
Publication Types:
Keywords:
- Biomedical Research
- Clinical Protocols
- Clinical Trials as Topic
- Meta-Analysis
- Research Design
- methods
- hsrmtgs
Other ID:
UI: 102273034
From Meeting Abstracts