Underreporting of alcohol-related mortality from cirrhosis is declining in Sweden and Denmark.

Death rates for alcoholic cirrhosis ran parallel to alcohol consumption in Sweden and Denmark from 1961 to 1986, according to official statistics. Reported death rates for alcoholic cirrhosis and for non-alcoholic cirrhosis increased linearly from 1965 to 1976 in Sweden and from 1967 to 1973 in Denmark. This suggests that constant proportions of deaths from real alcoholic cirrhosis are misclassified in official statistics as due to non-alcoholic cirrhosis during these time periods. Assuming that the mortality from real non-alcoholic cirrhosis is constant (c), a statistical model for estimating both c and the misclassification rates for real alcoholic cirrhosis (p) is proposed. In the Sweden population and in Danish males c was about 5 deaths per 100,000 adults annually. The model was not applicable to Danish females because their mortality from non-alcoholic cirrhosis decreased. The misclassification rate p was about 55% in both Swedish and Danish males and 70% in Swedish females during the above time periods. Moreover, during the past decade p has decreased to 30% in Swedish and Danish males.