Differential haloperidol effect on two indices of fentanyl-saline discrimination.

Using a discrete-trial, two-lever, food-reward discrimination learning paradigm, we trained rats (n = 6) to discriminate 0.04 mg/kg fentanyl (s.c. t-30') from saline. Stimulus generalization experiments with an adequate dose range (0.01-0.04 mg/kg) of fentanyl revealed that the ED50 value for drug lever selection is 0.02 mg/kg, irrespective of whether the animals were pretreated (s.c., t-60') with either saline or 0.08 mg/kg haloperidol. With increasing doses of the haloperidol-fentanyl combination, the percentage of total responding on the selected lever progressively decreased, and reached the 50% level at the highest drug combination. It is concluded that this percentage is heavily contaminated by factors unrelated to the discrimination condition being studied; these factors seem to invalidate this percentage as a discrimination index under experimental conditions (e.g., behaviorally toxic doses of drugs) where they are likely to operate. The use of response selection as a discrimination index in drug discrimination research is further argued.