This double-blind randomized controlled clinical trial was performed at Anqing Medical Center of Anhui Medical University from October 1, 2020 to November 30, 2021. The Ethics Committee of Anqing Medical Center of Anhui Medical University approved the study protocol, which was prospectively registered at www.chictr.org.cn (No. ChiCTR2000038442, registration date September 22, 2020). Each participant signed an informed consent form before surgery. Our study followed the Consolidated Standards and Regulations.

A total of 120 patients aged 18–65 years with American Society of Anesthesiologists (ASA) physical status I–II scheduled for elective thyroid cancer surgery (conventional open total thyroidectomy with therapeutic central compartment neck dissection) were enrolled in the study. Patients were excluded if they had known lateral neck node metastases, distant metastasis, bilateral central neck node metastases, recurrent laryngeal nerve invasion, local anaesthetic allergy, depression, anxiety, posttraumatic stress disorder, illicit drug use, severe pain, general poor health, bradycardia (heart rate less than 50 beats/min), second- or third-degree heart block, severe hepatic, renal or respiratory dysfunction, or refusal to participate in the study.

Eligible participants were randomized at a 1:1:1 ratio to receive lidocaine, dexmedetomidine or normal saline placebo using a computer-generated random table. The randomization sequence was kept in opaque envelopes, and a research nurse who was responsible for preparing the anaesthetics opened the corresponding envelopes. Blinding of all eligible participants, data collectors, medical care personnel and researcher staff was maintained during the entire trial period. Details of scales of QoR-15, visual analogue scale (VAS) and postoperative nausea or vomiting (PONV) intensity were explained to each patient 1 day before surgery.

Standard monitoring, including heart rate, mean arterial pressure, peripheral pulse oximeter value and electrocardiogram, was performed throughout the procedure. According to the study design, patients in group L received a loading dose of lidocaine (1.5 mg/kg) over 10 min before induction of anaesthesia followed by continuous infusion at a rate of 1.5 mg/kg per hour until 30 min before the end of surgery. Patients in group D received a loading of dexmedetomidine (0.5 µg/kg) over 10 min before induction of anaesthesia followed by continuous infusion at a rate of 0.5 µg/kg per hour until 30 min before the end of surgery. To attain double-blinding, 50-ml syringes containing lidocaine (12 mg/ml), dexmedetomidine (4 µg/ml) or saline were prefilled. Before anaesthesia induction, each patient received an intravenous infusion at a rate of 0.75 ml/kg per hour over 10 min. This rate corresponded to 1.5 mg/kg lidocaine or 0.5 µg/kg dexmedetomidine. Each patient received continuous infusions at a rate of 0.125 ml/kg per hour until 30 min before the end of surgery, and this rate corresponded to 1.5 mg/kg per hour of lidocaine and 0.5 µg/kg per hour of dexmedetomidine.

Anaesthesia induction was performed using target-controlled infusions (TCIs) of propofol and remifentanil. The initial TCI levels of plasma propofol and remifentanil were set as 3.0 μg/ml and 5.0 ng/ml, respectively. Vecuronium (0.08–0.1 mg/kg) was administered when the bispectral index (BIS) value decreased to less than 60. A continuous intraoperative neuromonitoring of recurrent laryngeal nerves (internal diameter 7.5 mm or 7.0 mm) was inserted after adequate muscle relaxation using a portable video laryngoscope, as described previously. Volume-controlled ventilation was performed to maintain an end-tidal carbon dioxide concentration between 35 and 45 mmHg using an anaesthetic machine (Aespire 7100, Datex-Ohmeda, Madison, WI), and the inspired oxygen fraction (FiO2) was 0.5 (balanced with air) during the anaesthesia period. No neuromuscular blocking agent was administered after anaesthetic induction. The BIS value was maintained between 45 and 60 by adjusting the TCI of the plasma propofol concentration during surgery. As reported previously, remifentanil infusions were administered manually or switched to manual infusion to maintain the fluctuation of mean arterial pressure and heart rate within 20% of the preoperative baseline value during surgery. If hypotension was not effectively treated with fluid replacement therapy and adjustment of the TCI of plasma remifentanil concentration, ephedrine 6 mg was administered. Atropine 0.5 mg was injected intravenously when the heart rate was less than 50 beats/min during the perioperative period. To alleviate the intensity of postoperative pain, sufentanil (0.2 μg/kg) was given intravenously at 30 min before the end of surgery. The surgeon tried to avoid recurrent laryngeal nerve injury via proactive exposure and neuromonitoring during total thyroidectomy.

To prevent PONV, ondansetron (4 mg) was administered intravenously at 10 min prior to the end of surgery. Parecoxib (40 mg) was provided every 12 h for 24 h. If VAS pain score exceeded 3, tramadol (50 mg) was administered intravenously for breakthrough rescue pain not alleviated by parecoxib.

The QoR-15 score on day 1 after surgery (POD1) was the primary endpoint. The QoR-15 questionnaire was used to evaluate the quality of recovery in five dimensions, namely physical comfort (five items), emotional state (four items), pain (two items), psychological support (two items) and physical independence (two items). The total score of the QoR-15 ranged from 0 (extremely poor recovery) to 150 (excellent recovery) [11]. Time to awareness, length of PACU (post-anaesthesia care unit) stay, cumulative consumption of remifentanil, time to first rescue analgesia, postoperative tramadol consumption, VAS pain score and incidence of PONV were the secondary outcomes. The incidence of PONV was assessed by the PONV intensity scale. A 10-cm VAS scale (0 = no pain, 10 = the most pain imaginable) was used to rate the VAS pain score at 2, 4, 8, 12 and 24 h after surgery. Side effects, such as arrhythmia (bradycardia of 50 beats/min or slower, A-V block, bundle branch block), hypotension (systolic blood pressure less than 90 mmHg), need for vasopressors, and prolonged respiratory support were recorded.

On the basis of our pilot study, the QoR-15 score was 10.2 points lower in group C than groups L and D on POD1. The standard deviation was 8 points on the QoR-15 [11]. The allowable error was 0.05, and each group needed 36 patients (assuming a power of 80%). Forty patients were included per group, allowing for a 10% dropout rate.

SPSS 16.0 (SPSS Inc., Chicago, IL) was used for statistical analyses. The Kolmogorov–Smirnov test was used to determine whether continuous data obeyed a normal distribution. Normally distributed data are presented as means (standard deviation, SD). One-way analysis of variance (ANOVA) was used for continuous data analyses of three groups. Tukey’s post hoc test was performed for further analysis of significant group differences. Nonnormally distributed variables are presented as medians and interquartile range (IQR) and were compared using the Mann–Whitney U test. The qualitative data are expressed as numbers or percentages and were analysed using χ2 tests. The significance level was P less than 0.05.