The optimal size of attributor for use with the University of Manchester Drug Misuse Database

Abstract

This paper describes the anonymous, attributable code which is applied to different individuals when they are reported to the regional drug misuse databases using the University of Manchester system. It then reports on the effect of removing selected data items included in the attributor on the system's ability to distinguish between different drug users reported to the database. All of the attributable codes that did not use the full set of data items were found to reduce the system's ability to distinguish between the reporting of new drug users and new episodes of drug use for users previously reported to the database.