Dosage adjustments in response to monitored plasma concentrations: can unblinded staff adhere to objective criteria?

A randomized, double-blind, placebo-controlled two-period crossover trial was conducted at two centers to evaluate felbamate for treatment of partial seizures in patients receiving concomitant treatment with two standard antiepileptic drugs, phenytoin and carbamazepine. During the pilot study, an interaction was seen; felbamate administration was associated with an approximate 20% increase in plasma phenytoin concentrations. Consequently, the controlled trial was designed to account for this interaction by reducing the phenytoin dosage by approximately 20% during the felbamate treatment period and using matching placebo capsules to maintain the blind. Unblinded staff were to make additional phenytoin dosage adjustments to keep the phenytoin concentrations within 20% of the patient's baseline mean. The data suggest that knowledge of treatment assignment may have influenced decisions regarding dosage adjustments even though there were objective rules for those adjustments. At one center, phenytoin dosages were adjusted at almost twice the rate during the active treatment period as during the placebo treatment period. To avoid potential bias, it is recommended that plasma concentrations be monitored by blinded staff.