Group testing procedures with incomplete identification and unreliable testing results

Abstract

We consider multistage group testing with incomplete identification and unreliability features. The objective is to find a cost-efficient group testing policy to select a prespecified number of non-defective items from some population in the presence of false-positive and false-negative test results, subject to reliability and other constraints. To confirm the status of tested groups, various sequential retesting procedures are suggested. We also extend the model to include the possibility of inconclusive test results. We derive all relevant cost functionals in analytically closed form. Numerical examples are also given. Copyright (C) 2006 John Wiley & Sons, Ltd.