Is MOOSE a loose goose?

A proposal for reporting meta-analysis of observational studies in epidemiology ('MOOSE') has been published. Meta-analysis of observational studies is outwardly similar in methodology to that applied to randomised controlled trials (RCT), but it does differ inherently. It is less certain and less definitive. Single summary estimates of effect derived from it are suspect. RCT and observational studies are complementary in nature, however, and they provide assessments of efficacy and effectiveness respectively. Meta-analysis of both types of study is needed, and the MOOSE proposal is a useful one. [ABSTRACT FROM AUTHOR]