Background and objectives Fentanyl is widely used as an intrathecal adjuvant to local anesthetics to enhance the duration of spinal anesthesia. Recent evidence suggests that intravenous dexmedetomidine prolongs the duration of spinal anesthesia. This noninferiority study evaluated whether bupivacaine alone could provide a noninferior duration of block compared with bupivacaine and fentanyl when intravenous dexmedetomidine was administered intraoperatively.

Methods Fifty-six patients scheduled for total knee arthroplasty under spinal anesthesia were randomly allocated to receive either bupivacaine 13 mg with intrathecal fentanyl 20 µg (Group BF) or bupivacaine 13 mg (Group B). Both groups underwent intravenous dexmedetomidine sedation throughout the surgery (1 µg kg^–1 for 10 min, followed by 0.5 µg kg^–1 h^–1). The primary outcome was the time to two-segment regression of the sensory block. The noninferiority margin for the mean difference was predefined as −10 min. Secondary outcomes included postoperative pain scores, analgesics consumptions, and the incidences of pruritus, nausea, and vomiting.

Results There was no significant difference in the two-segment regression time of sensory block (Group B 109.1±25.0 min vs Group BF 104.3±25.9 min; p=0.484). The mean difference in the two-segment regression time between the 2 groups was 4.8 min (95 % CI −8.9 to 18.6), demonstrating the noninferiority of bupivacaine alone. Secondary outcomes showed no significant differences between the two groups.

Conclusions The duration of spinal anesthesia with bupivacaine alone is noninferior to that of bupivacaine plus fentanyl in patients receiving intravenous dexmedetomidine intraoperatively. Our results suggest that intrathecal fentanyl may not be required when intravenous dexmedetomidine is administered.

Trial registration number NCT03105115.