Narrowing of the drinking repertoire criterion: should it have been dropped from ICD-10?

Objective: To present data on the reliability of the narrowing of the drinking repertoire criterion in the ICD-10 substance use disorder nomenclature and possible justification for its exclusion from ICD-10. This criterion has been one of the hallmarks of the Edwards and Gross dependence concept.
Method: The sample of 318 consisted of 175 substance users from St. Louis area substance abuse treatment programs and 143 substance users from the St. Louis Epidemiologic Catchment Area Project, a general population survey. Respondents were interviewed about their substance use with the CIDI-SAM and reinterviewed approximately 1 week later. "Narrowing of the drinking repertoire" was reported by 24%.
Results: Overall, the agreement from interview 1 to interview 2 was fair to good (kappa = 0.5851). Men and drinkers in treatment were found to be more reliable reporters than women and drinkers recruited from the general population.
Conclusions: The decision to drop the repertoire criterion seems justified, especially when non-clinical samples are interviewed. However, since our data found it highly associated with meeting criteria for a dependence disorder it is clinically valid. Additional research among varied populations is necessary to determine if the concept of narrowing is culture or gender specific and if its deletion affects rates of dependence.