The validity of self-reports in a follow-up study with drug addicts