In 2002 over 6% of all deaths in the European region were attributable to alcohol.1 Its contribution to mortality in Eastern Europe was particularly high with over 18% of male and 5% of female deaths resulting from the use of alcohol.2 Within this region much attention has been given on the role that alcohol has played in the alarming rise of mortality that has occurred in Russia during the transition period,3 when it has been linked to increases of both cardiovascular and external causes of death.4-6 There is also evidence however, that alcohol may have played an equally important role in mortality in other countries in this region in recent years.7 In the first half of the 1990s, the Baltic countries experienced a sharp increase in mortality that coincided with a great rise of alcohol-related causes of death.8 Similarly, the unprecedented rise in mortality that has occurred in several of the Slavic countries during the post-Soviet period has also been accompanied by a sharp rise of certain forms of alcohol-related mortality, such as alcohol poisoning.9
Although alcohol seems to be an important contributor to the burden of disease in the countries of Eastern Europe, as yet, with the obvious exception of Russia, little systematic research has been undertaken on its impact on mortality in specific countries in the region. This is especially true regarding Belarus where, since 1994, falling birth rates and high mortality rates have resulted in depopulation.10 The extent of the demographic problems affecting the country can be gauged by the reduction that has occurred in life expectancy which fell from 71.3 years at birth in 1990 to just 68.0 in 2002.11 Despite these changes, until now, only one (non-Western) study has attempted to systematically examine the effect of alcohol on increasing mortality in Belarus. It suggested that alcohol may be having a similar impact to that observed in Russia during recent years.12
The aim of the present study is to address this particular deficit concerning alcohol-related mortality in Belarus by using regional-level data drawn from death certificates. By doing this it will be possible to assess not only the level of alcohol-related mortality but also whether alcohol has a differential impact on death rates in Belarus in terms of such variables as age, sex and place of residence.
MATERIALS AND METHODS
The data used in this study come from Grodno province in Western Belarus which contains approximately 12% of the country's population with a majority of the province's residents (64%) living in urban areas. For the year 2002 all the death certificates from the province were obtained from hospital archives and examined to see if there was any mention of ' alcohol intoxication' as either a direct or underlying cause of death. From the 18,200 certificates that were retrospectively examined, 758 met the inclusion criteria for this study. Information regarding the sex, age, place of residence, place and cause of death (using the WHO ICD-10 nomenclature) as well who registered the death were extracted from each certificate. When the certificates were examined it was found that in only 3.8-5.6% cases of violent death (injuries, accidents suicides and homicides) was the presence of alcohol recorded. As the previous research has indicated that alcohol may be present in up to 50% of violent deaths in Belarus12 this suggests that in the current case, there was a situation of under-diagnosis of alcohol intoxication in cases of violent death. Death certificates relating to forms of violent death were thus omitted from the present study. Verification of the presence of alcohol in the remaining cases is likely to have been sufficiently rigorous in that, in 89.2% of cases the data were drawn from autopsies (in 70.5% of cases by forensic-medical experts and in 18.7% of cases by pathologists). In 5.3% of cases the death certificate was completed by the attending physician while in the remaining 5. …