A statistical approach for reconstructing continuous series of mortality by cause of death

Carlo G. Camarda, Max Planck Institute for Demographic Research

Data on cause-specific mortality have been collected over several decades for different countries. Nevertheless this valuable source of information has not been sufficiently explored. This is mainly due to frequent revisions of the International Classification of Diseases which lead to discontinuities and disruptions in cause-specific mortality trends. This paper proposes a new methodology for reconstructing continuous series of mortality by cause of death. The model assumes death counts as realizations from a Poisson distribution. The expected values of this distribution is a composed mean which embodies the continuous series and the redistribution pattern among causes of death. An algorithm allows the simultaneous estimation of these two components. The known associations among causes is used as input and the only assumption that is made about the underlying continuous series is smoothness. We present a simulation study and an application for demonstrating the performance of our model and its practical characteristics.