Diminishing trend in alcohol poisoning mortality in Estonia: reality or coding peculiarity?

AIMS To examine whether the changes in coding practice could reduce alcohol poisoning mortality rates in Estonia. METHODS Individual death records in 1983-2009 (age at death 25-64) were used to calculate the 3-year moving averages of age-standardized mortality rates. From 2000 onwards, there was a sharp increase in mortality from mental disorders due to alcohol, and at the same time a remarkable decrease in alcohol poisoning mortality. We calculated expected alcohol poisoning mortality rates for 2000-2009, assuming that mortality rate ratio of alcohol poisoning and mental disorders due to alcohol remained stable. RESULTS Alcohol poisoning mortality rates fluctuated considerably, being the lowest in 1988 and the highest in 1994-1995. A sharp decline started in 2000. Expected alcohol poisoning mortality rates continued their growth from 2000 onwards with a small decrease in 2006-2009. Mortality rates of mental disorders due to alcohol followed the same curve as alcohol poisoning rates up to 1999, being roughly 10 times lower than alcohol poisoning rates in both genders. From 2000 onwards, mortality from mental disorders due to alcohol increased rapidly, exceeding alcohol poisoning mortality in 2006. CONCLUSION This study demonstrates an obvious misclassification in coding of alcohol poisoning and mental disorders due to alcohol as underlying causes of death in the Estonian Causes of Death Registry.