Approximations for trimmed Fisher procedures in research synthesis

When combining the results of independent studies it often happens that som e studies are potentially aberrant either in quality or in actual values. B ecause aberrant studies are often at the extremes, we may wish to trim some of the largest and smallest effects. In such a case the use of p-values ma y well serve as a diag-nostic method. However, the use of ordered effects c hanges the distribution of the underlying statistics. We provide a discussi on of the exact distribution of the trimmed version of the Fisher procedure . Because of the complexity of the exact distribution, several approximatio ns are presented. These alternatives are applied to a metaanalysis on the e ffect of the dose of a drug on the risk of mortality.