BAYESIAN METAANALYSIS, WITH APPLICATION TO STUDIES OF ETS AND LUNG-CANCER

Meta-analysis enables researchers to combine the results of several st udies to assess the information they provide as a whole. It has been u sed to give a systematic overview of many areas in which data on a pos sible association between an exposure acid an outcome have been collec ted in a number of studies but where the overall picture remains obscu re, both as to the existence or size of the effect. This paper outline s some innovations in meta-analysis, based on using Markov chain Monte Carlo (MCMC) techniques for implementing Bayesian hierarchical models , and compares these with a more well-known random effects (RE) model. The new techniques allow different aspects of variation to be incorpo rated into descriptions of the association, and in particular enable r esearchers to better quantify differences between studies. Both the cl assical and Bayesian methods are applied, in this paper, to the curren t collection of studies of the association between incidence of lung c ancer in female never-smokers and exposure to environmental tobacco sm oke (ETS), both in the home through spousal smoking and in the workpla ce. In this paper it is demonstrated that compared with the RE model, the Bayesian methods: (a) allow more detailed modeling of study hetero geneity to be incorporated; (b) are relatively robust against a wide c hoice of specifications of such information on heterogeneity; (c) allo w for more detailed and satisfactory statements to be made, not only a bout the overall risk but about the individual studies, on the basis o f the combined information. For the workplace exposure data set, the B ayesian methods give a somewhat lower overall estimate of relative ris k of lung cancer associated with ETS, indicating the care that needs t o be taken in using point estimates based on any one method of analysi s. On the larger spousal data set the methods give similar answers. So me of the other concerns with meta-analysis are also considered. These include: consistency between different geographic areas (Asia and the United States), and our studies show that Bayesian methods permit an account of the overall picture to be taken, thus improving the ability to estimate accurately in the subgroups; and publication bias which, as shown with the spousal exposure data, may lead to an inflated exces s risk.