Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Gibbs-Sampler approach for meta-analysis of multiple clinical trials using generalized linear model with random-effects
Ist Teil von
Chinese medical journal, 2000-06, Vol.113 (6), p.554-557
Ort / Verlag
China
Erscheinungsjahr
2000
Link zum Volltext
Quelle
MEDLINE
Beschreibungen/Notizen
To investigate the use of the Gibbs-Sampler method in evaluating the relationship between clinic events and health risks in a meta-analysis of multiple clinical trials.
By using a generalized linear model with random-effects, Gibbs-Sampler technique was used in a meta-analysis of multiple clinical trials of angiotensin converting enzyme (ACE) inhibitors in patients with myocardial infarction (MI).
When heterogeneity across different trials can not be ignored, compared with the classic method, the odds ratio of relative reinfarction risk estimated by the Gibbs-Sampler method would have less variation. The gain in the reduction of variation in estimate of the overall odds ratio was 9.52%.
Implementation of the Gibbs-Sampler technique in meta-analysis of multiple clinical trials has the potential of reducing the inaccuracy caused by heterogeneity across trials.