A DECISION-SUPPORT SYSTEM FOR PARTIAL DRUG-TESTING - DSS-DT

This paper presents a Decision Support System (DSS) for the applicatio n of partial drug testing to a population of individuals with a histor y of drug abuse. The need for such a system arose in response to a 40% reduction in drug testing funds allocated to probation offices in the State of Illinois' Intensive Drug Supervision Programs (IDSP) in 1995 . Recent work in adapting single-attribute Bayesian acceptance samplin g to the problem of drug testing in 'at risk' populations has shown th at the total cost of sampling can be reduced without adversely affecti ng the proportion of users in the population. The DSS for Drug Testing (DSS-DT) allows users the opportunity to: (1) readily access informat ion about the prior distribution of drug use by population and drug ty pe; (2) generate optimal sampling plans based on current population in puts; (3) generate near-optimal sampling plans using a heuristic; and (4) evaluate the sensitivity of the solution to changes in various inp ut parameters for the drug testing model. Use of DSS-DT expedites the dissemination of the partial drug testing results while offering infor mation and budget planning support to planners charged with implementi ng a random drug testing procedure. (C) 1998 Elsevier Science B.V. All rights reserved.