English  Français

Comparability of sub-national cause-specific mortality data across national populations. Is there a reasonable way to redistribute ill-defined causes?

Pavel Grigoriev, Federal Institute for Population Research (BiB), Germany
Florian Bonnet, Institut National d'Études Démographiques (INED)
Michael Muehlichen, Federal Institute for Population Research (BiB), Germany
France Meslé, Institut National d'Études Démographiques (INED)

Ensuring the comparability of cause-of-death data represents a very serious challenge. In order to make cause-specific mortality comparable across time and space several methodological issues such as bridging different ICD classifications and redistributing ill-defined causes have to be resolved. The problem becomes even more serious when it comes to the comparability of sub-national cause-specific mortality data across national populations. The aim of this work is to test empirically alternative approaches of redistributing ill-defined causes and identify the one ensuring a reasonable comparability. We rely on the cause-specific mortality data by sex and 5-year age groups collected for the 96 German Raumordnungsregionen over the period 1992-2018 and 97 French departments (1979-2016). We perform the redistribution by each year, region, sex, and age group. First, we deal with the event of undetermined intent and redistribute it among other external causes of death. Then, we redistribute senility and ill-defined causes among other well-defined items. In both cases, we consider the application of two major approaches: 1) simple proportional redistribution 2) Ledermann’s method. We evaluate the effectiveness of the applied methods by comparing the plausibility of the obtained results with the initial (unadjusted) data.

Keywords: Cross-country comparative analyses, Mortality, Spatial analysis/regression, Methodology

See extended abstract.

  Presented in Session 183. Measurement of Mortality and Causes of Death