Sedation during flexible bronchoscopy (FB) should maintain an adequate respiratory drive, ensure maximum comfort for the patient, and warrant that the objectives of the procedure are achieved. Nevertheless, the optimal sedation method for FB has yet to be established. This study aimed to compare the standard recommended combination of midazolam-fentanyl (MF) with that of dexmedetomidine-ketamine (DK) for patient sedation during FB. Patients subjected to FB were randomly assigned to a DK (n=25) and an MF group (n=25). The primary outcome was the rate of critical desaturation events (arterial oxygen saturation <80% with nasal oxygen supply 2 L/min). Secondary outcomes included sedation depth, hemodynamic complications, adverse events, and patient and bronchoscopist satisfaction.

Study design and population This randomized controlled trial of adult patients (age >18 years) scheduled for FB was conducted between September 2019 and May 2020 at the Respiratory Medicine Department of Athens Naval Hospital, Greece.

Exclusion criteria included: known or suspected allergy to any of the study drugs, renal impairment (serum creatinine > 2 mg/dL), hepatic impairment (liver enzymes > 2 times the upper limit of normal), seizure disorders, history of psychosis or bipolar disorder, hemodynamic instability (heart rate - HR < 50 bpm or systolic blood pressure - SBP < 90 mmHg), and critically ill patients. The study protocol was approved by the Ethics Committee of Athens Naval Hospital (act number 296/13.08.2019), and written informed consent was obtained from all participants.

Protocol Demographic data and medical history were reviewed at the presentation, and the vital signs were assessed and recorded.

Eligible patients were then randomly assigned to two groups:

1. DK group: Dexmedetomidine solution of 1 μg/kg dissolved in 60 ml saline was administered over 15 minutes, followed by a maintenance dose of 0.5 μg/kg/h (continuous infusion). After the first 15 minutes, a bolus dose of 50 mg ketamine dissolved in 10 ml saline was given. An increase of 0.1 μg/kg in the infusion rate of dexmedetomidine was considered when necessary to optimize the level of sedation.
2. MF group: Bolus doses of midazolam (5 mg midazolam dissolved in 10 ml saline; 2 ml of the solution or 1 mg midazolam per bolus, premedication dose of 2mg, induction dose of 2 mg) and fentanyl (100 μg / 10 ml; 5 ml of the solution or 50 μg fentanyl per bolus, induction dose of 50μg) were administered at 20-minute intervals in between and titrated as needed to obtain the desired sedation depth. No more than 12 mg of midazolam and 150 μg of fentanyl were administered.

The depth of sedation was quantified at the onset of drug infusion using the Observer's Assessment of Alertness/Sedation scale; a maximum score of 3 was considered optimal for starting FB. Bolus doses of dexmedetomidine (DK group) and midazolam (MF group) were administered when necessary to maintain the same target score. An anesthesiologist was present throughout the procedure to oversee and monitor the sedation protocol. Only the anesthesiologist and the nurse of the bronchoscopy suite were aware of the sedation regimen. The attending pulmonologist remained blind regarding the sedation protocol until the completion of the bronchoscopist satisfaction questionnaire (see below).

Lidocaine gel was placed at the nostril of entry before the procedure, and lidocaine (4 ml of 2% solution) was sprayed under direct vision through the bronchoscope for vocal cord anesthesia before entering the larynx. All participants (in both the MF and the DK group) were initially premedicated with 2 mg of midazolam for anxiolysis and perioperative amnesia 10 minutes before commencing the procedure. They also received oxygen at 2 L/min via nasal cannula and 0.9 NaCl i.v. at 8 ml/h throughout the procedure. Continuous monitoring included electrocardiography, oxygen saturation (SpO2), and automated non-invasive blood pressure recordings.

Outcomes The depth of sedation was evaluated using the Ramsay sedation scale [28] and the Riker sedation-agitation scale. Both tools provide quantitative measurements in the maximum and minimum sedation-agitation range during the procedure. Higher Ramsay and lower Riker scores signify deeper sedation levels.

Bronchoscopist satisfaction was measured with a Likert-scale tool answering the question "How satisfied are you with both the ease and outcome of the procedure?" as follows: 1 = not satisfied at all; 2 = somewhat satisfied; 3 = more satisfied with the procedural outcome; 4 = mostly satisfied with the outcome and somewhat with procedural ease; 5 = exceptionally satisfied with both procedural outcome and ease.

Patient satisfaction was evaluated by a short, custom-made questionnaire that included the following questions: 1) I was satisfied with the sedation administered for the procedure, 2) I felt pain and/or discomfort beyond my tolerance during the procedure, 3) I believe my needs were met during the procedure 4) I felt pain and discomfort after the procedure, 5) I would be willing to undergo a second procedure if the first did not yield adequate results. The responses were measured in a Likert scale format with values 1 = strongly disagree; 2 = disagree; 3 = neither agree nor disagree; 4 = agree; 5 =strongly agree. Questions 2 and 4 were scored reversely. The questionnaire was given in printed form at discharge, and the responses were collected by telephone call the day after the procedure.

Sample size estimation and statistical analyses The primary endpoint of the study was the occurrence of major desaturation events (i.e., SpO2 <80% with nasal oxygen supply 2 L/min). Assuming a desaturation rate of 20 ± 5%, we calculated a lower critical number of 25 patients per group to prove non-inferiority within the above limits of the primary endpoint, with a p-value of <0.05 and 85% power.

Inclusion Criteria:

• Patients scheduled for flexible bronchoscopy at the Naval Hospital of Athens
• Written informed consent

Exclusion Criteria:

• known or suspected allergy to any of the study drugs
• renal impairment (serum creatinine > 2 mg/dL)
• hepatic impairment (liver enzymes > 2 times the upper limit of normal)
• seizure disorders
• history of psychosis or bipolar disorder
• hemodynamic instability (heart rate - HR < 50 bpm or systolic blood pressure - SBP < 90 mmHg)
• critically ill patients.