Interpretation of items in the AUDIT questionnaire.

Aims: To test for the possibility that tertiary students misinterpret certain items on the Alcohol Use Disorders Identification Test (AUDIT).
Methods: Responses to alternative question wordings were compared with responses to standard items.
Results: Alterations to items 5 and 9, so that consequences of drinking epitomized in these items were more specifically defined, resulted in markedly different response distributions to the item, but the total AUDIT score was not changed.
Conclusions: Caution is necessary before using individual AUDIT items as measures of consequences in population surveys, and the possibility of false positives in total scores should be borne in mind.