EXAMINING ASSOCIATIONS BETWEEN OCCUPATION AND HEALTH BY USING ROUTINELY COLLECTED DATA

When examining a large number of associations simultaneously, as happe ns when routinely collected data are used to assess associations betwe en occupation and health, it is not obvious how best to identify assoc iations requiring further investigation since some risks may be high, or low, by chance alone. We have developed an approach to deal with th is problem which is relatively easy to apply and appropriate to applic ations where data are not too sparse. Observed to expected ratios are estimated using an empirical Bayes procedure. Anomalous associations c an be seen as outliers in a normal probability plot of the log-ratios. The method is illustrated in the analysis of 252000 cancers registere d in men in England during 1981-87.