Contingency management for accurate predictions of urinalysis test resultsand lack of correspondence with self-reported drug use among polydrug abusers

Contingency management procedures have proven effective in the treatment of drug-dependent patients. These procedures, however, often require frequent urine testing, which is too costly for community treatment programs. To ma ke urine-testing procedures more cost effective, the feasibility of reinfor cing accurate predictions of urine drug screen (UDS) results was evaluated. Participants made extremely accurate UDS predictions, particularly when th ey made drug-positive predictions, regardless of whether predictions were r einforced. However, self-reports of recent drug use had poor correspondence with predictions of UDS results. Results suggested that if programs only t ested samples predicted to be drug free, considerable cost savings could be incurred. Further research is needed to determine if validity would be enh anced by using a proportion of costs saved to provide nominal reinforcement when samples were verified to be drug free.