Certain patterns of cannabis use have become a major public health concern in developed societies. In several European countries, current work is under way to develop and validate instruments adequate for monitoring intensive, dependent or problematic forms of cannabis use in population surveys. For such purposes, several easy to administrate questionnaires exist. However, there is yet no generally accepted, validated screening instrument. One of
the instruments used in Swiss population surveys is the Cannabis Use Disorders Identification Test (CUDIT). The CUDIT is a ten-items-questionnaire that was constructed by New Zealand clinicians who have adapted the well-established Alcohol Use Disorders Identification Test (AUDIT) to cannabis.
The aim of our study was to examine the CUDIT's dimensionality and to explore its psychometric properties by means of Item Response Theory. For that purpose we have created a test version of the CUDIT by adding four additional items to the original ten-items version. Our analyses are based on data from the second wave of the Swiss Cannabis Monitoring Study, which was conducted under the patronage of the Swiss Federal Office for
Public Health. In the representative population sample of 5714 Swiss adolescents and young adults (13 to 32 years old), 558 were actual cannabis users, to whom the CUDIT was administered. Our results show that it makes sense to treat the CUDIT as a unidimensional construct. At the same time, three of the CUDIT's original items were found to be of weak psychometric
performance, and we suggest replacement with new items. The aim of our presentation is not to introduce an accomplished new screening instrument
but to discuss the CUDIT's potential as an instrument to screen for problematic cannabis use in population surveys (or clinical settings), and to make some suggestions for its improvement and further development.