Objectives. The aim of the study was to identify predictors of the problematic pattern of the cannabis users' behavior.
Samples and setting. The subjects consisted of a representative sample of 3928 Czech adolescents (boys = 1918; girls = 2010) from the ESPAD international survey.
The subjects (N = 1052; boys = 588; girls = 464) who reported cannabis use were included in the analysis. Based primarily on surmise rather than adequate evidence, 18 variables expected to be associated with a pattern of the behavior of use or problems experienced within the scope of cannabis use operationalized by screening instrument CAST were selected from the ESPAD database.
Results. Using a multiple regression analysis eight statistically significant predictors of the CAST variability were identified: Frequency of cannabis use during whole life, Frequency and extend of smoking, Frequency of contacts with friends using cannabis, Satisfactory relationships with friends, Parental emotional support, Number of friends using cannabis, Age at the first cannabis use, and Gender.
The model accounted for 43.7% of cannabis risk behavior. Comparison of the CAST 7 point value threshold as an indicator for professional help in the Czech sample to the French sample results did not indicate clinically important differences.
The instrument can be judiciously used with proper care in clinical screening praxis. Study limitation.
The study is not based on a priory formulated model and has only descriptive and at most tentatively predictive character.