An assessment of the use of Bayes' Theorem for forecasting in public health: the case of epidemic meningitis in China

A mathematical model based upon Bayes' Theorem (BT) was used to forecast the occurrence of epidemic cerebrospinal meningitis (ECM) in ten communities in North China. Reports of ECM from each ten-day period during the meningitis season and records of special population movement during 1960-82 were analysed to establish forecast models. Calibration, split-sample, random-sample selection, as well as actual forecast tests, were used to check the efficiency of the models. For all the tests, the theoretical occurrence of ECM forecast by the BT method was compared with the observational data. Since the BT method offers efficiency and convenience, it is recommended for use in planning for the prevention and control of ECM in China.