Printed In Great Britain A Note Concerning the Analysis of an Epidemic of Q Fever

Techniques of statistical analysis developed for the analysis of experiments, for which the usual precautions of randomization ensure independence, are not necessarily appropriate to the analysis of epidemics. An outbreak of Q Fever in a Swiss Alpine valley previously reported in the International Journal of Epidemiology \s analysed to make this point In an energetically prosecuted epidemiological investi-gation Dupuis et al { have traced the outbreak of an epidemic of Q Fever in the Val de Bagnes in Switzer-land in the autumn of 1983 to the movement of 12 flocks of infected sheep through the valley. Of three main reasons for identifying the sheep as the source of the epidemic the second is extremely interesting from the statistical point of view. The authors write, 'from a geographical point of view we observed that all of the villages in the lower part of the valley along the road followed by the sheep wer