To the best of our knowledge, this is the first prospective randomized study of alfentanil in combination with continuous propofol infusion titrated to maintain a BIS level of 65–75 for FB sedation. This study revealed that BIS-guided propofol infusion is as safe as the current standard method of clinically-judged midazolam sedation, in terms of the proportion of patients experiencing hypoxemia and hypotension. The mean lowest blood pressure was lower in the study group compared to the control group, and all patients recovered spontaneously or after proper management. This study proves that BIS-guided propofol sedation provides better tolerance and faster recovery to patients undergoing FB, and provides faster induction and less procedural interference for the bronchoscopists. These findings are clinically valuable, especially for complicated and time-consuming interventional bronchoscopic procedures.
The BIS algorithm was developed from a large database of patients or volunteers receiving various anesthetic regimens 
. However, it has not been applied routinely during FB, and the optimal BIS level for FB sedation is not yet established. Grass et al. 
reported that 95% of healthy volunteers under propofol sedation become amnesic at a mean BIS level of 77 (95% CI 72–83), and 50% lose their response to verbal commands at a mean BIS level of 63 (95% CI 62–65). Bauer et al. 
, in another propofol sedation study of patients undergoing surgery, reported that patients could still move in response to pain stimulation when the mean BIS level was 69. In another study using BIS monitoring during gastrointestinal endoscopy sedation, Bower et al. 
reported that patients responded to mild prodding, which is considered the level of moderate sedation, at a median BIS level of 67.5 (range, 54–97). Thus, a BIS level of 65–75 was set for FB sedation, and a BIS level of 70 was set for induction in this protocol to achieve patients that are amnesic but still with reflex responsiveness to noxious stimulation at a relatively level of sedation. If cough-related BIS elevation occurred, coughing was first treated as mentioned in our method., instead of introducing propofol.
Recently, Clark et al. 
compared administration of single drug bolus of propofol to midazolam while maintaining the BIS level at 70–85 for a small number of patients undergoing simple FB with better ASA class. Patients receiving propofol showed better global tolerance, but no difference in the perception of coughing and there was no difference in the bronchoscopists' assessment, comparing to the control patients. Stolz et al. 
reported that the discomfort score and safety profiles of patients receiving sedation for FB with propofol bolus based on clinical judgment were similar to those with sedation induced using combined midazolam and hydrocodone, although coughing was more severe with propofol treatment. Compared to these studies, the FB procedures employed in the current study were more complicated and time-consuming, but the study revealed that BIS-guided propofol infusion with alfentanil administration provides additional benefits for the bronchoscopists (less procedural interference) and patients (less discomfort from scope insertion, dyspnea, and cough). Compared to intermittent administration of boluses, which may result in fluctuating plasma concentrations and risk of over- or under-sedation, continuous propofol infusion provides a more steady plasma concentration that can be titrated within the therapeutic window 
. Adding alfentanil can modify the pharmacokinetic property of propofol, reduce the required dose of propofol, and facilitate faster recovery with less cardiovascular depression 
. A BIS monitor provides a trend of EEG change during a short processing time (usually 15 s) and gives the feedback of conscious level changing fromthe patient responsiveness to the procedures and drugs, thereby helping sedative drug adjustment. Thus, a BIS-guided propofol infusion can provide a steadier drug concentration and effect, as well as allow instant and individualized titration. This may explain why BIS-guided sedation was better tolerated and had reduced irritated movements by patients, without increased adverse effects in this study. Despite higher doses of propofol, and more complicated and longer procedures, the incidence of hypotension was similar to that reported by Clark et al. 
(3.7–7.4% vs. 4.7%) and that of hypoxemia (39.9%) was consistent with previous reports (34.9% and 32%, respectively) 
Logistic regression revealed that electrocautery, ASA class 3, and male gender were associated with hypoxemia in the study group. Electrocautery often generates tissue debris and blooding that contribute to ventilation/perfusion mismatch in patients with endobronchial lesions during prolonged procedures. Patients with higher ASA physical status may have relatively inadequate cardiopulmonary function. In the current study, men received more alfentanil during induction than women (310.8 vs. 271.1 µg, p
<0.001) because of higher mean weight (64.4 vs. 56.0 kg, p
<0.001), and had a higher incidence of hypoxemia during induction (20% vs. 9.2%, p
0.023). The higher amount of alfentanil administered to men may be a contributing factor to the increase in the incidence of hypoxemia in men. Another explanation may be the gender-based difference in the pharmacokinetic properties of alfentanil or propofol 
. However, this study was not designed to investigate such parameters. For male patients with ASA physical status 3 or those with endobronchial obstructions where electrocautery is indicated, propofol infusion should be performed with caution.
Logistic regression also revealed that a lower induction dose of propofol was associated with hypotension. Patients who required less propofol for induction may be more sensitive to propofol. This may be explained by different interpatient susceptibilities to propofol due to the difference in cardiac output, hepatic perfusion, body fat, and haplotype differences in metabolic genes, and requires to be studied further
. This is a valuable hint for clinical practice. If induction is achieved rapidly, sedative procedures should be performed with caution. The BIS has been used to monitor propofol titration for sedation of various procedures 
; however, there are other aspects of FB sedation that require further study. Besides the cost-effectiveness of the additional equipment and personnel, the regimen of opioid administration for induction as well as the procedure, the optimal BIS level for FB sedation, and the propofol infusion profile for better pharmacokinetic control (e.g., targeted control infusion) require further investigation to improve FB sedation.
This study has some limitations that should be considered. First, the investigators and bronchoscopists were not blinded to the sedation procedures. Because the aim of this study was to compare the current standard practice of clinically-judged incremental midazolam sedation to BIS-guided propofol continuous infusion, it is difficult to use completely blinded conditions because of maior difference in the protocols. The bronchoscopist can distinguish the protocol by observing the actions of the investigators while patients without irritated motion but BIS level reaching to the criteria for drug titration. Nonetheless, the primary endpoints were hypoxemia and hypotension events, which were recorded objectively. Second, patients with a history of FB were not excluded. Many patients indications for interventional bronchoscopy often undergo FB first for lesion site evaluation or for pathologic confirmation. Such previous experience with FB may affect a patient's judgment regarding their tolerance of FB. However, the numbers of patients with previous FB experience in the two groups were similar (data not shown); therefore, this should not influence data interpretation.
In conclusion, the current study showed that BIS-guided propofol infusion combined with alfentanil is feasible and safe. It provides excellent tolerance and fast recovery for the patients undergoing FB. It also facilitates the performance of procedures and reduces patient interference. Further studies on induction, BIS level, and propofol infusion profile for FB sedation as well as cost-effectiveness of BIS guided propofol infusion are warranted to improve the safety and quality of FB sedation.