Background: Few studies have compared self-reported criminal behaviour with high-quality databases of criminal offences and judicial sanctions. Self-reported problems from drug abusers are generally believed to be valid. We assessed the validity of self-reported theft, drug offences and prison sentences from a five-year follow-up of female substance abusers who were originally treated in a compulsory care unit in Lund, run by the Swedish Board of Institutional Care.
Methods: Data from a total of 106 of a consecutive sample of 132 women inter-viewed in a five-year follow-up. All were thoroughly assessed for somatic complaints, psychiatric and psychological problems, background factors with standardized instruments. Data over the five years were linked to official records of judicial sanctions, retrieved from The National Council for Crime Prevention, Stockholm, Sweden. Register data have a full cover for the whole cohort. The current data base contain full data back to 1975 up to 2004.
Results: Agreement was assessed for each year, as well as for the total period. Statistical control was performed for other types of crimes and prison. Although statistically significant, agreement was modest, and in contrast to previous studies, patients under-reported violence charges.
Conclusion: The findings suggest that self-reports of criminal behaviour from women can be used with some caution, and that the validity of self-report may vary between types of criminal justice system involvement.