The case-control method of determining human cancer causation has developed in the last 20 years as the major epidemiological approach to identify risk factors. Now in paperback, this volume brings together the methods available for analyzing case-control data, and makes them accessible to the practicing epidemiologist and statistician. It expounds both the statistical models and theory behind the methods, and the practical steps of applying them to specific sets of data. All the methods are illustrated with examples from real studies, with an emphasis on the added insight into the data that...

Methods for investigating generalized spatial clustering of disease in human populations have only recently become available. This volume presents the outcome of a unique practical test of these methods, in which authors of several newly-developed approaches conducted their own blind analyses of over 50 artificial datasets, some random, some generated by clustering processes. Results were then compared with the known spatial structure. An historical view of leukaemia clustering is also included. This book will be of particular interest to epidemiologists and public health specialists with...

This volume assesses formal methods for the quantitative estimation and prediction of human cancer risks. Quantitative estimates of cancer risk can be expressed in different ways. In some cases estimates of risk under conditions prevailing in the original data are of primary interest; in others predictions of risk under other conditions are required. Estimates of risk may be based on empirical models that provide a reasonable description of the available data or on models developed on the basis of plausible assumptions about the mechanisms of carcinogenesis. Throughout this volume established...