NCJ Number
154647
Journal
Mathematical Computer Modelling Volume: 17 Issue: 2 Dated: (1993) Pages: 77-88
Date Published
1993
Length
12 pages
Annotation
This study proposes an analytic model based on individual decision theory and Bayesian acceptance sampling to develop cost- minimizing drug-testing strategies.
Abstract
A theoretical model of individual drug use is developed that implies a prior probability distribution over the number of drug users in a population. This prior distribution is then used to develop an optimal sampling strategy by minimizing an expected total cost function. The cost function is based on the cost of inspection (testing), the cost of treating or sanctioning drug users identified by the testing, the cost associated with additional testing and treatment associated with rejecting a lot based on the sample, and the cost incurred for undetected drug users that remain in an accepted lot. The impact of the drug- testing program on individuals in the program was modeled by using an expected utility function. Program effects impacted on the prior distribution through these utility functions. The single-period results showed the impact of relative costs on the optimal testing strategy. Findings show that under some circumstances, the best drug-testing strategy is not to test. Under other circumstances, screening the entire population was optimal. Under other cost scenarios, however, lower expected total costs could be achieved by using an acceptance sampling approach. An examination of the multi-period model suggests that consideration of the prior distribution and of changes in the prior distribution achieved by the drug-testing program could lead to cost savings. Alternative acceptance sampling plans could prove optimal over multiple periods. Directions for future research are suggested. 3 tables, 1 figure, and 24 references