Study design
This study was approved by the Ethics Committee of the Eye, Ear, Nose, and Throat Hospital of Fudan University, Shanghai, China (2013005). Written informed consent was obtained from all subjects participating in the trial. This prospective, randomized, double-blind, single-center clinical trial was registered prior to patient enrollment at clinicaltrials.gov (NCT03918889, principal investigator: Rui Xu, date of registration: March 28, 2019) and performed at the Department of Anesthesiology in the Eye, Ear, Nose, and Throat Hospital of Fudan University. All procedures adhered to the applicable CONSORT guidelines.
Patients were randomly allocated to either the dexmedetomidine group (group D) (n = 43) or the midazolam group (group M) (n = 43). Randomized group allocation was performed using a computerized randomization table created by one staff member who was not involved in the patients’ anesthesia or recovery care. The randomization result was kept sealed in an envelope; only the nurse who prepared the anesthetics could open the envelope in order to prepare the allocated drug. A total of 83 medical records were analyzed, 43 from group D and 40 from group M. The patients, the nurse in the post-anesthesia care unit (PACU), and attending anesthesiologists were blinded to the medicine administration.
Inclusion criteria
We enrolled 86 adult male patients with American Society of Anesthesiologist physical status III or IV, aged 25–70 years, scheduled for partial or total laryngectomy.
Exclusion criteria
Patients with cardiac disease, neuropsychiatric diseases, pharyngeal paraganglioma, or uncontrolled hypertension (i.e., systolic blood pressure > 160 mmHg or diastolic blood pressure > 90 mmHg), taking β-adrenoreceptor blockers, with long-term (>6 months) abuse of alcohol, taking opioids or sedative-hypnotic drugs, with dexmedetomidine or midazolam allergies, undergoing awake fiberoptic intubation, with operation times shorter than 1 h or longer than 4 h, or with a tracheotomy history were excluded.
Anesthesia
Drugs, including sedative, analgesic, anti-emetic, and anti-itching drugs, were not given before an operation. After arrival at the operation room, the electrocardiogram, SpO2 levels, blood pressure, the bispectral index, end-tidal carbon dioxide levels, and temperature were continuously monitored and recorded. General anesthesia was induced with sufentanil (0.2 μg/kg) and propofol (2.5 mg/kg), and after confirmation of adequate muscle relaxation with the administration of cisatracurium (0.2 mg/kg) iv, an endotracheal tube with an internal diameter of 7 mm was inserted into the trachea. Endotracheal tube cuff pressure was maintained at 25 cmH2O measured using a calibrated handheld Portex Cuff Inflator Pressure Gauge (Portex Limited, Hythe, Kent, UK). Prior to the start of surgery, sufentanil (0.1 μg/kg) was given. Either dexmedetomidine (Precedex; Henrui Pharmaceutical, China) (group D, n = 43) infusion (0.5 µg/kg 10 min before tracheotomy, then adjusted to 0.3 µg/kg/h) or midazolam (Midazuolun injection; Enhua Pharmaceutical, China) (group M, n = 43) infusion (0.05 mg/kg 10 minutes before tracheotomy, then adjusted to 0.02 mg/kg/h) was administered in a blind mode. Anesthesia was maintained with a minimum alveolar end-tidal concentration of sevoflurane of 1–1.3 in 30% oxygen/air mixture to keep the bispectral index between 45 and 55. The maintenance infusion rate of cisatracurium was 1–1.5 µg/kg/min and the maintenance infusion rate of sufentanil was 0.002 μg/kg/min according to clinical needs. Granisetron (6 mg) was administered at the end of surgery to prevent post-operative nausea and vomiting (PONV). Endotracheal secretions were removed before tracheostomy tube insertion. Topical tetracaine hydrochloride gel was applied to the tracheostomy tube to enhance toleration.
After surgical procedures were finished, sevoflurane administration was discontinued, 100% oxygen was administered at 6 l/min, and patients were transferred to the PACU. Neostigmine (0.04 mg/kg) and atropine (0.02 mg/kg) were given to reverse residual neuromuscular block. After spontaneous ventilation returned, patients were considered to have fully recovered from muscle relaxation, and after patients opened their eyes, they were weaned from mechanical ventilation. Nurses who assessed subjects were blinded to the medicine intervention. If there was any adverse event, an attending anesthesiologist managed it. In the case of bradycardia (heart rate [HR] < 45 beats/min), 0.5 mg atropine was administered, and if systolic blood pressure (SBP) decreased to less than 90 mmHg, ephedrine (6 mg) was used.
Cough grading was based on a modified 4-point Minogue scale: grade 1, no cough; grade 2 (mild), coughing once or twice; grade 3 (moderate), fewer than 4 non-sustained coughs lasting 1–2 s each or overall coughing lasting less than 5 s; grade 4 (severe), at least 4 coughs lasting at least 2 s, or overall coughing duration more than 5 s8. Patients with grades 3 and 4 were categorized as “moderate to severe.” The patients’ levels of sedation were assessed by the Ramsay sedation scale (RSS): 1, the patient is anxious and restless or agitated, or both; 2, the patient is cooperative, tranquil, and oriented; 3, the patient responds to commands only; 4, the patient exhibits a brisk response to loud auditory stimuli or a light glabellar tap; 5, the patient exhibits a sluggish response to a loud auditory stimulus or a light glabellar tap; 6, the patient exhibits no response9. In addition, the nurse also assessed the post-operative pain score by a visual analog scale (VAS) (on a scale from 0 to 10, where 0 is no pain, and 10 is very much pain). If the pain score was above 5, sufentanil (0.1 μg/kg) was given to patients immediately as rescue analgesic; consumption of analgesics was recorded.
HR, SBP, diastolic blood pressure (DBP), and pulse oximetry results (SpO2) were recorded before induction (T1), after drug administration (T2), after intubation (T3), after medicine intervention (T4), at the moment of laryngectomy (T5), after completion of surgery (T6), at the point of awareness (T7), at departure from the PACU (T8), 2 h after surgery (T9), 24 h after surgery (T10), and 48 h after surgery (T11). Duration of surgery, respiratory recovery time, and duration of PACU stay were also recorded. The incidence of adverse events, including bradycardia, hypotension (<30% decrease from baseline), hypertension (>30% increase from baseline), vomiting, pale lips, delirium, subcutaneous emphysema, and hematoma, was noted by a nurse who was blinded to medicine intervention. The incidence of pneumonia 72 h after surgery was also recorded.
Statistical analysis
The primary endpoint was incidence and severity of cough. The secondary outcome measures were hemodynamic responses, post-operative pain scores, sedation scores, respiratory recovery time, duration of PACU stay, and incidence of adverse events.
PASS15 was used to calculate the sample size. On the basis of a preliminary study, the incidence of no cough in group M was about 65%, and in group D it was about 25% higher than in group M. The proportion in group D was assumed to be 0.65 under the null hypothesis and 0.90 under the alternative hypothesis. The proportion in group M was taken as 0.65. The test statistic used is the one-sided Z-test with unpooled variance. The significance level of the test is 0.025. Group sample sizes of 40 in group D and 40 in group M achieve 80% power to detect a difference between the group proportions of 0.25. Assuming a dropout rate of 8%, the final sample size was determined to be 43 patients per group, with a power of 80% and an alpha level of 0.05.
Student’s t test was used for between-group comparisons of HR, SBP, DBP, and SpO2. Repeated-measures ANOVA was used for within-group comparisons. The χ2 test or the Fisher exact test was used to analyze coughing severity, sedation, pain scores, and adverse events. A P-value of 0.05 or less was considered statistically significant.