Materials and methodsForty-five unpremedicated ASA class I or II patients scheduled to undergo either orthopedic or ear, nose, and throat surgery under general anesthesia with an anticipated duration of approximately 1.5-3 h were enrolled in this study after approval of the hospital ethics committee and written informed consent from the patients had been obtained. All patients were free from neuromuscular, endocrine, liver, or renal diseases and were not receiving drugs known to interact with neuromuscular blocking agents. They required muscle relaxation only for endotracheal intubation.An intravenous infusion of 0.9% sodium chloride solution was given initially via the basilic vein on one of the arms, while the other arm was kept for monitoring neuromuscular block. The heart rate (HR), mean arterial pressure (MAP), peripheral oxygen saturation (SpO 2 ) (Odam Physiogard SM 785, Wissenbourg, France), and end-tidal concentrations of CO 2 and volatile anesthetics (Artema MM 256, Sundbyberg, Sweden) were monitored. Each patient was allocated to one of three groups: enflurane (group E, n ϭ 15), sevoflurane (group S, n ϭ 15), and neuroleptanesthesia (group NA, n ϭ 15). The induction of anesthesia was performed with i.v. fentanyl 2 µg · kg Ϫ1 and thiopentone 5-7 mg·kg Ϫ1 , followed by a volatile anesthetic, either enflurane or sevoflurane, with assisted ventilation by mask in groups E and S, respectively. The volatile anesthetics (enflurane or sevoflurane) were administered in 66%/33% : nitrous oxide/oxygen at the endtidal concentration corresponding to 1 minimum alveolar concentration (MAC) in the present study. One MAC of enflurane and sevoflurane was assumed to be 0.57% and 0.66% in approximately 66% nitrous oxide, respectively [13]. In group NA, anesthesia was induced with i.v. droperidol 0.2 mg · kg Ϫ1 , fentanyl 5 µg · kg Ϫ1 , and thiopentone 1-2 mg · kg Ϫ1 during inhalation of 66%/33% : N 2 O/O 2 . Stable end-tidal anesthetic concen-