Problem drinking in a general practice population: the construction of an interval scale for severity of problem drinking.

The main aim of this article is to describe the construction of an interval scale measuring the severity of problem drinking among patients of general practitioners. A questionnaire, including the existing screening instruments CAGE, SMAST, a shortened version of the SAAST and some additional items on alcohol-related problems, was given to three categories of patients: those who, according to the general practitioner, were not problem drinkers (Group 1), those who perhaps were (Group 2) and those who were known by the general practitioner to be problem drinkers (Group 3). According to the Classical Test Theory, the three existing screening instruments and the combination of all questions form a reliable scale. Advanced scaling techniques were applied to select the best items for the construction of an interval scale. These analyses showed that 18 of the 28 questions in the complete screening list fitted in a unidimensional Rasch scale. These questions can be used to estimate the severity of problem drinking. For research purposes an interval scale is attractive, since no information is lost by dichotomization. Such a scale enables the level of problem drinking to be specified at which correlations with other factors, whether as a cause or as an effect, can be found. The scale could also be a help to general practitioners for referral decisions. In more specialized settings it can be used to assign patients to various treatments.