Bronchoscopy is performed on a daily basis for diagnostic and therapeutic purposes. This procedure may cause anxiety and discomfort requiring the use of sedative drugs.

In this study the investigators titrate the administration of sedatives using bispectral analysis of the EEG. The investigators compared 2 drugs, propofol versus midazolam regarding patient subjective tolerance, recovery of brain function, safety and operator's satisfaction.

Sedation using anxiolytic, hypnotic and /or analgesic drugs is currently used during flexible bronchoscopy (FB) to facilitate the diagnostic process and improve patient comfort.[The technique of sedation varies widely and depends primarily on the experience of the chest physician. FB can also be done without sedation. However, according to a recent survey, 80% of patients prefer to be sedated during FB and it is also well known that the operator has a tendency to overstate the tolerance of patients during the procedure.

The use of benzodiazepines for sedation undoubtedly enhances the perception and tolerance during FB and they are by far the most widely used drugs because of their ease of administration, their speed of action and the possibility to administer an antidote. The drawback of these sedative agents is related to a high variable individual sensitivity, which can lead to drug overdose and to severe respiratory depression requiring ventilatory assistance. In addition, their sedative and amnesic effects might persist several hours after the procedure.

Over the last decade, major advances in technology and pharmacology have resulted in better cerebral monitoring of anaesthesia with tools like electroencephalographic bispectral index and synthesis of sedative agents of short duration of action as propofol (P).

P is a lipidic emulsion used intravenously for induction and maintenance of anaesthesia. Its effect begins rapidly and ends faster than midazolam (M). The main advantage of P is therefore a faster recovery of mnesic, cognitive and motor functions, whereas M induces anterograde amnesia (1-2 h after bolus) and slower motor reaction . This gives to P a keen interest in ambulatory practice. While P is commonly and safely used in several gastroenterological endoscopic procedures, its use by pneumologist is currently hampered by a lack of expertise and could benefit from collaboration with the anaesthesia team.

This is a randomised trial comparing recovery of brain function and patient subjective tolerance following FB after administration either of P or M for sedation. Patients and operators are blinded to the sedative drug, which is administered and titrated by another chest physician.

For each procedure, the staff consists of a chest physician trained in FB (operator), a physician in charge of sedation and two nurses for technical assistance and proper data recording. Before the procedure, blood pressure (BP), heart rate (HR), oxygen saturation (SaO2) and weight are recorded. The patient is placed in supine position, fitted with a nasal oxygen cannula and a peripheral intravenous access (18-20 G) for the infusion of crystalloids. Oxygen is administered only if SaO2 was < 92%. Local anaesthesia is done using lidocaïne 10 % and 1% in the pharynx and central airways respectively. The sedative drugs syringes and lines are hidden from the operator by a sheet. Sedation is then started with P or M according to randomization for any FB as follows: We inject in both groups a 5 ml bolus of lidocaine 1% (50 mg) (prevention of pain associated with P injection), followed by either a 40 mg (4 ml) bolus of P or 2 mg of M. The re-injection of 20 mg (2ml) of P or 2 mg of M was allowed every 2 minutes to achieve and maintain BIS values between 70 and 85 and OAA/S of 3-4. Blood pressure, oxygen saturation, heart rate, BIS and OAA/S are monitored continuously and recorded every 3 minutes during the procedure and at 5, 15, 30, 45 and 60 minutes after its end. We record the time necessary to achieve the targeted BIS value after the injection of the sedative drug, the duration of FB, the electroencephalographic recovery time, defined as the time to reach a BIS value > 90 after the end of FB and total doses of M and P.

Sixty minutes and 24 hours after FB, operator and patient, both blinded to the allocation, evaluate the tolerance to the FB with visual analogic scales (VAS) (1 mm: excellent tolerance, 100 mm very low tolerance) to assess 1) the global tolerance of the procedure and 2) the intensity of 4 key symptoms during FB (pain, nausea, breathlessness and cough).

Before the procedure and 15 and 60 minutes after the procedure, patients are asked to complete CPT trials to evaluate cognitive recovery.](streamdown:incomplete-link)

Inclusion Criteria:

• patients scheduled for bronchoscopy
• ASA I to III
• age between 18 and 79 years
• capacity of discernment

Exclusion Criteria:

• visual and psychological problems that might interfere with psychometric testing
• chronic obstructive pulmonary disease with FEV1 < 50% predicted
• need for an airway intubation or laryngeal mask insertion
• fever or other signs of systemic infection
• hemodynamic instability defined as a resting heart rate (HR) < 60 or ≥ 120 and/or a systolic blood pressure (BP) < 100 or > 180 mmHg
• significant liver disease and documented soy allergy