WHO MONICA Project: geographic variation in mortality from cardiovascular diseases. Baseline data on selected population characteristics and cardiovascular mortality.

Death certificates are the most commonly used source of information for the estimation of the frequency of cardiovascular diseases (CVD) in populations. National mortality statistics have been the major tool used for studying geographic variations in CVD. Such comparisons have suggested that major differences in CVD mortality trends can be found between countries, and that even national trends can show diverging temporal variations (1-6). To draw conclusions on the incidence and prevalence of CVD from national mortality statistics only is very difficult. Many epidemiologists and statisticians have questioned the accuracy of disease-specific death rates which are obtained from the certification of an underlying cause of death based on the International Classification of Diseases (I CD) (7-10). The current state of medical knowledge, improper or varying coding practices and insufficient clinicalor laboratory data are the main causes of inaccuracies in death certification. The ways in which underlying causes of death are classified in vital statistics offices may also introduce a bias that can contribute to international differences in cause-specific mortality. Another problem, specific to geographic comparisons, is that national mortality statistics are produced by pooling data from large populations that may include several heterogeneous subpopulations presenting different mortality experiences.