UNIGE document Privat-docent Thesis
previous document  unige:43531  next document
add to browser collection

A review of methods for meta-analysis of aggregated survival data

Defense Thèse de privat-docent : Univ. Genève, 2014
Abstract In most of meta-analyses of aggregated survival data, a pooled measure of the intervention's effect is obtained by combining reported hazard ratios. Advanced statistical methods have been proposed for a better characterization of the effect. With these methods, published survival curves can be synthesized in a single summary survival curve and a variation over follow-up time of the effect can be tested. Besides the pooled intervention's effect, the synthesis of survival in each arm is helpful for the appraisal of the intervention's benefits. Moreover, some meta-analyses aim to evaluate survival in a single group. In meta-analysis of comparative studies, testing a variation of the effect over time can avoid simplistic conclusions. These methods have been published in journals of statistics and may not be accessible to a large audience. The purpose of this article is to review novel methods for meta-analysis of aggregated survival data. The principle of each method, and its advantages and limitations are explained.
Keywords Meta-analysisAggregated survival dataSummary survival curveHazard ratio
Full text
Thesis (719 Kb) - public document Free access
Research group Epidémiologie clinique (115)
(ISO format)
COMBESCURE, Christophe. A review of methods for meta-analysis of aggregated survival data. Université de Genève. Thèse de privat-docent, 2014. doi: 10.13097/archive-ouverte/unige:43531 https://archive-ouverte.unige.ch/unige:43531

905 hits



Deposited on : 2014-12-16

Export document
Format :
Citation style :