Alcohol attributable fractions and costs in Portugal

Aim: This study estimates alcohol attributable fractions which are used to obtain a measure of health expenditures related to alcohol misuse in Portugal. Methods and Sources: Studies estimating alcohol-attributable expenditures focus usually on specific diseases or conditions. This study takes a broader approach by examining all expenditures considered as being attributable to excessive drinking, i. e., those resulting from utilization directly and indirectly related to alcohol associated diseases. Unlike studies using the relative risk methodology, where population alcoholattributable fractions (or etiologic fractions) are based on epidemiological studies identifying potential causes of death and disease associated with alcohol excessive consumption, this study has adopted the microeconometric methodology which gets estimates of attributable fractions running econometric regressions of the annual costs of medical care (medical appointments or hospital admissions) for individuals. It has an advantage over the risk approach in that it allows to control for the effect of factors other than alcohol consumption on medical care expenditures. Actually, it includes in the analysis those factors that are linked to health care expenditures but are not linked to any particular alcohol related disease. To perform our analysis we have used two data sets provided by the Portuguese Ministry of Health: the 1995 National Health Survey. These surveys provide detailed information on individual demographic, economic, health utilization, health attributes and also alcohol and smoking use. Both surveys use a probabilistic representative sample of the non-institutionalized Portuguese population and a questionnaire of about 180 questions. It is a representative sample of the five main administrative regions of Portugal (North, Centre, Lisbon and Tejo valley, Alentejo and Algarve).