Approval was obtained from the Institutional Review Board of Liaocheng People's Hospital and the Fifth People's Hospital of Jinan for this retrospective clinical study, which was registered at chictr.org (ChiCTR-IPR-16008494). Patients provided informed consent. Patients who underwent neurosurgery from January 2015 to December 2015 were enrolled in this study if they met the following inclusion criteria: age between 35 and 65 years, American Society of Anesthesiologists (ASA) grade I–II. Exclusion criteria included a history of endocrinological disease, hypertension (diastolic blood pressure >160 mmHg), ischemic heart disease, second or third degree heart block, pregnancy, long-term abuse of alcohol (>6 months), opioids, or sedative—hypnotic drugs, DEX allergies, nausea, and vomiting after previous surgery, neuropsychiatric diseases, operation time shorter than 1 h or longer than 6 h, blood loss greater than 1200 mL, reoperation during the first 72 h after neurosurgery, or emergency operation.

Patients were divided into two groups: Group D and Group ND. A total of 153 medical records were reviewed retrospectively, 76 from Group D and 77 from Group ND. Electronic chart and DoCare clinic electronic anesthesia recording system data were utilized.

Electrocardiography, arterial blood pressure, pulse-oximetry, end-tidal CO₂, and temperature were continuously monitored using an automated system (Philips IntelliVue MP50) after patients arrived at the operating room, then two peripheral intravenous catheters were placed in all patients before induction. A forced-air warming device (EQUATOR Convective Warmer, EQ-5000) was used in both groups to maintain normothermia.

Patients in Group D received intravenous DEX (0.5 μg·kg⁻¹) over a period of 10 min before endotracheal intubation, which was then adjusted to 0.3 μg·kg⁻¹·h⁻¹ until incision suturing. Both groups were induced using the same agents: fentanyl (2 μg·kg⁻¹), lidocaine (1.5 mg·kg⁻¹), propofol (2 mg·kg⁻¹), and cisatracurium (0.2 mg·kg⁻¹). General anesthesia was maintained with a 1–1.2 minimum alveolar concentration of sevoflurane in an air/oxygen mixture, and a remifentanil (0.05–0.1 μg·kg⁻¹·min⁻¹) infusion and intermittent boluses of cisatracurium (0.05 mg·kg⁻¹) for muscle relaxation. Positive pressure ventilation and oxygenation were maintained with endotracheal intubation to achieve an arterial partial pressure of carbon-dioxide of 35–40 mmHg. The concentration of sevoflurane and rate of remifentanil were adjusted according to the hemodynamic limits and bispectral Idex. (40–60).

During the operation, bradycardia and tachycardia were defined as HR < 45 bpm or a >30% increase from baseline and were treated using atropine (0.2–0.5 mg) or esmolol (20–30 mg), respectively. Hypertension was defined as a >30% increase from baseline and was treated by increasing the inspired sevoflurane concentration by 0.2% or remifentanil by 0.02 μg·kg⁻¹·min⁻¹ in a stepwise titration; fentanyl (1 μg·kg⁻¹) was used if hypertension was persistent for 3 min, and if persistent after fentanyl treatment, urapidil (10–15 mg) was administered. Hypotension was defined as a >30% decrease from baseline and was treated by decreasing the inspired sevoflurane concentration by 0.2% or remifentanil by 0.02 μg·kg⁻¹·min⁻¹ in a stepwise titration, and if persistent, by administration of ephedrine ephedrine (6–12 mg) or phenylephrine (20–80 μg). DEX and sevoflurane were discontinued approximately 15 and 5 min before completion of the surgery, respectively. All patients received 5 mg of tropisetron and underwent routine reversal of neuromuscular blockade. After extubation in the operating room, all patients were transferred to the post anesthesia care unit (PACU).

Patients in Group ND received sufentanil at a continuous dose of 0.02 μg·kg⁻¹·h⁻¹, a bolus dose of 0.02 μg·kg⁻¹ with a 5-min lockout interval and a 1 h limit of 16 mL, while patients in Group D received both sufentanil and DEX under the same regimen as those in Group ND after arriving at the PACU.

Patients were encouraged to self-control the pump when they experienced pain during coughing at a severity of Visual Analogue Scale (VASc) score >4. If patients demonstrated a poor response to the patient controlled analgesia (PCA) (VASc scores of >6), or an obvious PCA-associated adverse effect took place, 30 mg of ketorolac was administered. If the rescue analgesia was ineffective 30 min after administration, 100 mg of tramadol was administered.

The intraoperative hemodynamic data (perioperative systolic, diastolic, mean arterial pressure, and heart rate) were obtained from the Phillips IntelVue monitor at the following points: arrival at the operating room (T0), just before intubation (T1), after intubation (T2), during head pinning (T3), during surgical incision (T4), extubation (T5), arrival at the PACU (T6), and 5 min (T7), 10 min (T8), 15 min (T9), and 20 min (T10) after arriving at the PACU. The requirement of narcotic and vasoactive drugs, fluid requirements, and estimated blood loss, incidence of complications during the operation and in the PACU were recorded. The length of the PACU stay was also recorded based on the Aldretes criteria (Aldrete, 1995).

The cumulative amount of self-administered sufentanil was recorded until 72 h after the surgery. Postoperative pain intensity (VAS, both at rest and with cough), the Ramsay Sedation Scale (RSS: recorded on a six-point scale: 1, subject anxious, agitated, or restless; 2, subject cooperative, oriented, and tranquil; 3, subject responds to commands; 4, subject asleep but with brisk response to light glabellar tap or loud auditory stimulus; 5, subject asleep, sluggish response to light glabellar tap or loud auditory stimulus; 6, subject asleep, no response), and the Bruggemann comfort scale (BCS: 0, persistent pain; 1, severe pain while deep breathing or coughing; 2, mild pain while deep breathing or coughing; 3, painless while deep breathing; 4, painless while coughing) were evaluated at 1, 4, 8, 12, 24, 48, and 72 h after surgery. The number of rescue analgesia and adverse effects (such as bradycardia, tachycardia, hypotension, hypertension, nausea and vomiting, agitation, and respiratory depression) was also recorded at the end of the study.

The data from a preliminary study showed that the requirement of sufentanil in patients not treated with DEX was approximately 165.34 ± 23.13 μg during the first 72 h after neurosurgery. A 25% reduction in sufentanil use was considered clinically feasible. For a study power of 80% (α = 0.05, β = 0.2), assuming a dropout rate of 15%, the required sample size per group was calculated to be 75, with a total of 150 patients allowed for adequate data collection (PASS 11.0, NCSS Statistical Software, Kaysville, Utah).

The Kolmogorov–Smirnov test was used to assess the distribution of variables. Homogeneity of variance was determined using Levene's tests. Quantitative data were expressed as mean and standard deviation or median and inter-quartile range. Inter-group comparisons were performed using repeated-measures analysis of variance. Bonferroni's correction was used for post-hoc multiple comparisons. The non-parametric Wilcoxon—Mann—Whitney test was used for variables that were not normally distributed. Categorical data were expressed as frequency and percentage and analyzed using chi-squared tests or Fisher's exact tests when appropriate. Probability (P) values of <0.05 were considered statistically significant. Statistical analysis was performed using SPSS for Windows Version 18.0 (SPSS Inc., Chicago, IL, USA).

The patient demographics are illustrated in a patient enrollment flow diagram (Figure 1). One thousand two hundred and eighty-three patients who underwent neurosurgery were screened from January 2015 to December 2015. 1130 patients were excluded due to not meeting the inclusion criteria: 514 patients required emergency surgery, 24 patients refused the surgery, the age of 143 patients did not fall in the specified range of 35–65 years, the ASA grade of 72 patients was more than II, 129 patients had cardiovascular and neuropsychiatric diseases, 35 patients had a long history of addiction to alcohol, opioids and sedative–hypnotic drugs, 78 patients had nausea and vomiting after previous surgery, the operation time of 67 patients was longer than 6 h, 35 patients demonstrated hemorrhage of more than 1200 mL, 23 patients required reoperation during the first 72 h after neurosurgery, three patients refused use the PCA after surgery, and seven patients were excluded after surgery due to incomplete clinical data. Finally, 153 patients were included in the primary analysis and divided into two groups (76 patients in Group D and 77 patients in Group ND).

The two groups were comparable regarding age, sex, BMI, ASA grade, history of comorbidity and tumor location (Table 1).

Baseline HR and MAP were not statistically different between the two groups (P > 0.05; Figure 2). Compared with Group ND, HR in Group D was significantly decreased from T1 to T10, while MAP in Group D was significantly decreased from T1 to T7 (P < 0.05; Figure 2). The lowest levels of HR and MAP of the two groups were observed at T1.

Compared with Group ND, the consumption of sevoflurane [1.38 ± 0.42 vs. 1.04 ± 0.35 minimum alveolar concentration (MAC), P < 0.001], remifentail (0.13 ± 0.03 vs. 0.19 ± 0.05 μg·kg⁻¹·min⁻¹, P < 0.01), fentanyl (3.24 ± 1.13 vs. 4.72 ± 1.41 μg·kg⁻¹ P < 0.01) and estimated blood loss (486.57 ± 33.38 vs. 505.04 ± 32.52 ml, P < 0.01) were significantly decreased in Group D (P < 0.05, Table 2). More patients in Group ND than in Group D required an intraoperative vasoactive agent: urapidil (32 vs. 21, respectively, P = 0.01), phenylephrine (17 vs. 7, respectively, P = 0.03) and esmolol (37 vs. 25, respectively, P = 0.01), while the requirements for ephedrine (6 vs. 5, respectively, P = 1.00) and atropine (8 vs. 9, respectively, P = 0.06) were similar (Tables 3, 4). There were no statistically significant differences between the two groups in terms of duration of surgery and anesthesia, infusion, and urine output, with the exception of recovery time at the PACU (27.72 ± 3.78 vs. 45.14 ± 4.29 min; P > 0.05; Table 2).

The total dosage and dosage per body weight of sufentanil were significantly lower in Group D than in Group ND at 4, 8, 12, 24, 48, and 72 h after surgery (P < 0.05, Figure 3). Compared with Group ND, the VAS scores at rest at 1, 4, and 8 h after surgery and with cough at 12, 24, 48, and 72 h after surgery were significantly lower in Group D (P < 0.05, Figure 4). In Group D, 11 patients (14.47%) required postoperative rescue analgesia, while 21 patients (29.87%) required rescue analgesia in Group ND (P = 0.02, Table 5). There were no significant differences between the two groups in terms of RSS and BCS during the first 72 h after neurosurgery (P > 0.05, Figure 5).

The main adverse events are recorded in Table 5. Patients in Group D showed a lower incidence of tachycardia (6 vs. 15, P = 0.04), hypertension (12 vs. 28, P = 0.02), delirium (6 vs. 19, P = 0.01), nausea (12 vs. 27, P = 0.01) and vomiting (5 vs. 14, P = 0.01) than patients in Group ND. However, there are no statistically significant differences between the two groups in terms of the incidence of bradycardia (3 vs. 5, P = 0.28) and hypotension (7 vs. 11, P = 0.42). Although more patients in Group ND suffered seizures during the period of this study than those in Group D, the difference was not statistically significant (P = 0.49, Table 6).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.