Abstract
BACKGROUND: The transmission potential of primary pneumonic plague, caused by Yersinia pestis, is one of the key epidemiological determinants of a potential biological weapon, and requires clarification and time dependent interpretation. METHOD: This study estimated the reproduction number and its time dependent change through investigations of outbreaks in Mukden, China (1946), and Madagascar (1957). Reconstruction of an epidemic tree, which shows who infected whom, from the observed dates of onset was performed using the serial interval. Furthermore, a likelihood based approach was used for the time inhomogeneous evaluation of the outbreaks for which there was scarcity of cases. RESULTS: According to the estimates, the basic reproduction number, R(0), was on the order of 2.8 to 3.5, which is higher than previous estimates. The lower 95% confidence intervals of R(0) exceeded unity. The effective reproduction number declined below unity after control measures were introduced in Mukden, and before the official implementation in Madagascar. CONCLUSION: While the time course of the latter outbreak could be explained by intrinsic factors and stochasticity in this remote and scarcely populated area, the former in Mukden suggests the possible continued chains of transmission in highly populated areas. Using the proposed methods, the who infected whom information permitted the evaluation of the time inhomogeneous transmission potential in relation to public health measures. The study also tackles the problem of statistical estimation of R(0) based on similar information, which was previously performed simply by counting the number of secondary transmissions regardless of time.