Does sedation help in fibreoptic bronchoscopy?

BMJ 1994; 309 doi: (Published 05 November 1994) Cite this as: BMJ 1994;309:1206
  1. M Q F Hatton,
  2. M B Allen,
  3. A S Vathenen,
  4. E Mellor,
  5. N J Cooke
  1. Departments of Respiratory Medicine and Pharmacy, Leeds General Infirmary, Leeds LS1 3EX
  1. Correspondence to: Dr M B Allen, Chest Clinic, St Luke's Hospital, Bradford BD5 0NA.
  • Accepted 6 June 1994

Although sedation is associated with major complications1,2 sedative drugs are often given immediately before fibreoptic bronchoscopy in the belief that patients' comfort is improved. Uncontrolled studies have shown that fibreoptic bronchoscopy is well tolerated without sedation.3,4 Most comparative studies of premedication have looked at different drug regimens; we are aware of only one study that included an unsedated control arm.5

Opiates and benzodiazepines are frequently used for sedation during fibreoptic bronchoscopy; we compared two such regimens with placebo.

Patients, methods, and results

Of 184 patients undergoing routine diagnostic fibreoptic bronchoscopy without transbronchial biopsy, 182 consented to enter a consecutive double blind comparison of (a) intravenous phenoperidine and droperidol with saline placebo or (b) intravenous midazolam with saline placebo. Three doctors performed the bronchoscopy; all patients received supplemental oxygen, topical lignocaine, intravenous atropine 600 μg, and the trial drugs through individual syringes prepared by the pharmacy department. The dose given was varied to produce light sedation, with most patients receiving phenoperidine 1 mg and droperidol 5 μg or midazolam 70 μg/kg.

A visual analogue scale (100 mm) was used to score the answer to five questions about the ease and comfort of bronchoscopy, high scores indicating an unfavourable response. Results are expressed as medians with differences, 95% confidence intervals, and the significance of the differences (Mann-Whitney U test; Minitab). Immediately after bronchoscopy doctors stated whether they thought active drug or placebo had been used and rated the ease of performing the procedure. At the same time they and the attendant nurses rated how comfortable the patient had been during the procedure. Patients were assessed by a doctor and when fully recovered (usually around six hours) rated how comfortable they had been during the procedure and their willingness to undergo a repeat procedure if one was clinically indicated.

Fifty patients were randomly assigned phenoperidine and droperidol and 51 placebo. Midazolam and placebo were randomly assigned in the ratio of two to one to 51 and 30 patients respectively. For 180 patients the study bronchoscopy was their first such procedure, two patients having undergone one several years previously. Doctors correctly identified active treatment in 41 (82%) of those receiving phenoperidine and droperidol and in 27 (53%) of those given midazolam. The table shows median scores and 95% confidence intervals.

Age, sex ratio, and visual analogue scores of patients in placebo controlled comparison of sedation with phenoperidine and droperidol or with midazolam during fibreoptic bronchoscopy

View this table:


One of the main objects of sedation is to make fibreoptic bronchoscopy less unpleasant for patients. Several different regimens have been described; phenoperidine and droperidol and, more recently, midazolam are frequently used.2

In our study phenoperidine and droperidol produced sedation, with 82% of the injections being correctly identified. Doctors and nurses found that bronchoscopy was more easily performed and patients more comfortable with this regimen than with placebo. Patients, however, found no difference in comfort between the active and placebo regimens, and they were less willing to have the test repeated when they had received these drugs. Although this combination has the theoretical advantages of amnesia, analgesia, euphoria, and cough suppression, the dysphoric effects probably taint patients' perception of the procedure.

The benzodiazepine midazolam would be expected to be helpful by producing amnesia and anxiolysis. In the doses we used, however, there was little obvious effect, doctors correctly identifying its use on only 52% of occasions. Furthermore, although doctors found the procedure easier with midazolam, the drug was not significantly better than placebo in making patients comfortable or willing to have the test repeated.

These sedative regimens fail in their primary function of making the procedure more tolerable for patients. This may be partially explained by the amount of sedation given; we titrated the dose to produce light sedation to minimise the risks previously described.2 Given the patients' opinion, we suggest that phenoperidine and droperidol should no longer be recommended and that light sedation with midazolam has few benefits. Our results suggest that routine sedation has little part to play in patients undergoing a single diagnostic procedure.

We thank Mrs S Barrowclough and Miss K McConnely for their help with the bronchoscopies and completion of the visual analogue scales.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract