Reconstruction of coherent cause-specific mortality time series for Russia and its regions

Inna Danilova, Max Planck Institute for Demographic Research and Higher School of Economics (Moscow)
France Meslé, Institut National d'Études Démographiques (INED)
Jacques Vallin, Institut National d'Études Démographiques (INED)

The analyses of long-term trends in cause-specific mortality is complicated due to changes in international classifications of diseases. Such changes often produce discontinuities in trends in mortality rates by cause of death. The most proper and comprehensive approach for overcoming the problem is reclassification (reconstruction) of all death in time series according to one and the same classification system (ICD). Vallin and Meslé proposed the reconstruction method based on transition coefficients calculated for each change of classification for each cause of death. (Vallin and Meslé, 1988; Meslé and Vallin 1996). This method was used for recalculating Russian series of mortality by CoDs for years 1956-1998 (Meslé et al. 1996; Meslé et al. 2003). The year 1999 mark a new methodological challenge because of the introduction of the ICD10 in Russia. Therefore, in this paper new algorithm for transition from ICD9 to ICD10 for Russia is presented and discussed. In order to solve this methodological task, we analyze specific features if this transition with particular focus on unique experiences of different regions in implementing new classification. We found that the ICD10 in some regions was implemented in 1999, while in some other regions it was postponed for 1-3 years. As a result, Russian mortality statistics by causes of death show some inconsistences for the period 1999-2002. In order to account for such methodological challenges we propose several complimentary modifications of the original reconstruction method by F.Meslé and J.Vallin which allow to produce continuous series of causes of death for Russia.