british thoracic society guidelines on diagnostic flexible ... toracica... · british thoracic...

of 21 /21
British Thoracic Society guidelines on diagnostic flexible bronchoscopy British Thoracic Society Bronchoscopy Guidelines Committee, a Subcommittee of the Standards of Care Committee of the British Thoracic Society Introduction These guidelines have been developed at the request of the Standards of Care Committee of the British Thoracic Society (BTS). Two particular problems have emerged since the pre- vious BTS guidelines were published. 12 Firstly, there have been many cases reported recently of atypical mycobacteria causing contamination of bronchoscopes leading to pseudoinfections. Sec- ondly, toxicity from glutaraldehyde has become a significant problem as it has been implicated in cases of occupational asthma among some nurs- ing and technical staV working in endoscopy units (the term “endoscopy unit” is used in these guidelines as most bronchoscopy services are now carried out within such units). The previous BTS bronchoscopy guidelines were brief and were not based on a formal search of published evidence. A Working Group was formed at the request of the BTS with instructions to develop formal evidence-based guidelines for flexible bron- choscopy. The Committee consisted of indi- viduals with a wide range of backgrounds including nurses, a microbiologist, an infection control expert, as well as respiratory physicians including one with a special interest in intensive care medicine. Full details are given in Appendix 1. The aim of the Committee was to produce evidence-based guidelines for subsequent use by medical, nursing, and technical staV. The areas to be covered were carefully defined and- were primarily to advise on bronchoscopy in adults, although the sections of the guidelines concerned with staV safety and instrument decontamination would also be relevant to paediatric flexible bronchoscopy. The areas covered by these guidelines are as follows: + complications, contraindications and pre- cautions; + sedation and anaesthesia/analgesia; + cleaning and disinfection including glutar- aldehyde usage; + staV safety; + bronchoscopy in the intensive care unit; + data collection and staV training; + patient satisfaction. Indications for bronchoscopy and therapeu- tic aspects are not covered by these guidelines. A comprehensive search of several databases was carried out in 1998 and updated in 1999 using subject terms or MeSH headings (medi- cal subject headings) and also controlled vocabulary search terms. The electronic data- bases used were Medline (searched from 1966 onwards), EmBase (from 1980), Biological Abstracts (from 1988), CINAHL (a nursing literature database searched from 1983), Psych- lit (from 1967), and the Cochrane Controlled Trial Register. The criteria for assessing the levels of evidence and grading of recommendations were based on those recommended in the Scottish Intercollegiate Guidelines Network, 3 as have also been used in some other BTS guidelines (table 1). The papers were rigorously assessed by members of the committee and decisions on levels of evidence for each paper were made by two or more members of the group. The draft guidelines were presented at the December 1998 meeting of the BTS and were subse- quently circulated to the members of the BTS, discussed in detail with the Standards of Care Committee and Training Committee, and sent for comments to the Department of Health (including the Medical Devices Agency and Microbiology Advisory Committee), Intensive Care Society, Medical Defence Union, Medi- cal Protection Society, Cancer BACUP Sup- port Service, and the patient representatives’ group of the British Lung Foundation. Liaison with the Royal College of Nursing was carried out by their nominee, Adrienne Brewin. These guidelines should be reviewed and updated in 2003. Table 1 Levels of evidence and grading of recommendations 3 Level Type of evidence Ia Evidence obtained from meta-analysis of randomised controlled trials Ib Evidence obtained from at least one randomised controlled trial IIa Evidence obtained from at least one well designed controlled study without randomisation IIb Evidence obtained from at least one other type of well designed quasi-experimental study III Evidence obtained from well designed non-experimental descriptive studies such as comparative studies, correlation studies, and case controlled studies IV Evidence obtained from expert committee reports of opinions and/or clinical experiences of respected authorities Grade Type of recommendations A (levels Ia, Ib) Requires at least one randomised controlled trial as part of a body of literature of overall good quality and consistency addressing the specific recommendation B (levels IIa, IIb, III) Requires availability of well conducted clinical studies but no randomised clinical trials on the topic of recommendation C (level IV) Requires evidence from expert committee reports or opinions and/or clinical experience of respected authorities. Indicates absence of directly applicable studies of good quality Thorax 2001;56:(suppl I) i1–i21 i1 Members of the British Thoracic Society Bronchoscopy Guidelines Committee: D Honeybourne (Chairman), J Babb, P Bowie, A Brewin, A Fraise, C Garrard, J Harvey, R Lewis, C Neumann, C G Wathen, T Williams (further details in Appendix 1) Correspondence to: Dr D Honeybourne, Department of Respiratory Medicine, Birmingham Heartlands Hospital, Birmingham B9 5SS, UK [email protected] .

Author: others

Post on 09-Feb-2020




0 download

Embed Size (px)


  • British Thoracic Society guidelines on diagnosticflexible bronchoscopy

    British Thoracic Society Bronchoscopy Guidelines Committee, a Subcommittee of theStandards of Care Committee of the British Thoracic Society

    IntroductionThese guidelines have been developed at therequest of the Standards of Care Committee ofthe British Thoracic Society (BTS). Twoparticular problems have emerged since the pre-vious BTS guidelines were published.1 2 Firstly,there have been many cases reported recently ofatypical mycobacteria causing contamination ofbronchoscopes leading to pseudoinfections. Sec-ondly, toxicity from glutaraldehyde has becomea significant problem as it has been implicated incases of occupational asthma among some nurs-ing and technical staV working in endoscopyunits (the term “endoscopy unit” is used in theseguidelines as most bronchoscopy services arenow carried out within such units). Theprevious BTS bronchoscopy guidelines werebrief and were not based on a formal search ofpublished evidence.

    A Working Group was formed at the requestof the BTS with instructions to develop formalevidence-based guidelines for flexible bron-choscopy. The Committee consisted of indi-viduals with a wide range of backgroundsincluding nurses, a microbiologist, an infectioncontrol expert, as well as respiratory physiciansincluding one with a special interest inintensive care medicine. Full details are givenin Appendix 1.

    The aim of the Committee was to produceevidence-based guidelines for subsequent useby medical, nursing, and technical staV. Theareas to be covered were carefully defined and-were primarily to advise on bronchoscopy in

    adults, although the sections of the guidelinesconcerned with staV safety and instrumentdecontamination would also be relevant topaediatric flexible bronchoscopy.

    The areas covered by these guidelines are asfollows:

    + complications, contraindications and pre-cautions;

    + sedation and anaesthesia/analgesia;+ cleaning and disinfection including glutar-

    aldehyde usage;+ staV safety;+ bronchoscopy in the intensive care unit;+ data collection and staV training;+ patient satisfaction.

    Indications for bronchoscopy and therapeu-tic aspects are not covered by these guidelines.

    A comprehensive search of several databaseswas carried out in 1998 and updated in 1999using subject terms or MeSH headings (medi-cal subject headings) and also controlledvocabulary search terms. The electronic data-bases used were Medline (searched from 1966onwards), EmBase (from 1980), BiologicalAbstracts (from 1988), CINAHL (a nursingliterature database searched from 1983),Psych-lit (from 1967), and the Cochrane ControlledTrial Register.

    The criteria for assessing the levels ofevidence and grading of recommendationswere based on those recommended in theScottish Intercollegiate Guidelines Network,3

    as have also been used in some other BTSguidelines (table 1).

    The papers were rigorously assessed bymembers of the committee and decisions onlevels of evidence for each paper were made bytwo or more members of the group. The draftguidelines were presented at the December1998 meeting of the BTS and were subse-quently circulated to the members of the BTS,discussed in detail with the Standards of CareCommittee and Training Committee, and sentfor comments to the Department of Health(including the Medical Devices Agency andMicrobiology Advisory Committee), IntensiveCare Society, Medical Defence Union, Medi-cal Protection Society, Cancer BACUP Sup-port Service, and the patient representatives’group of the British Lung Foundation. Liaisonwith the Royal College of Nursing was carriedout by their nominee, Adrienne Brewin.

    These guidelines should be reviewed andupdated in 2003.

    Table 1 Levels of evidence and grading of recommendations3

    Level Type of evidence

    Ia Evidence obtained from meta-analysis of randomised controlled trialsIb Evidence obtained from at least one randomised controlled trialIIa Evidence obtained from at least one well designed controlled study

    without randomisationIIb Evidence obtained from at least one other type of well designed

    quasi-experimental studyIII Evidence obtained from well designed non-experimental descriptive

    studies such as comparative studies, correlation studies, and casecontrolled studies

    IV Evidence obtained from expert committee reports of opinions and/orclinical experiences of respected authorities

    Grade Type of recommendationsA (levels Ia, Ib) Requires at least one randomised controlled trial as part of a body of

    literature of overall good quality and consistency addressing thespecific recommendation

    B (levels IIa, IIb, III) Requires availability of well conducted clinical studies but norandomised clinical trials on the topic of recommendation

    C (level IV) Requires evidence from expert committee reports or opinions and/orclinical experience of respected authorities. Indicates absence ofdirectly applicable studies of good quality

    Thorax 2001;56:(suppl I) i1–i21 i1

    Members of the BritishThoracic SocietyBronchoscopy GuidelinesCommittee: DHoneybourne(Chairman), J Babb,P Bowie, A Brewin,A Fraise, C Garrard,J Harvey, R Lewis,C Neumann,C G Wathen, T Williams(further details inAppendix 1)

    Correspondence to:Dr D Honeybourne,Department of RespiratoryMedicine, BirminghamHeartlands Hospital,Birmingham B9 5SS, [email protected] .

  • Recommendations


    + Verbal and written patient information improves tolerance of the procedure by the patient and should be provided. [B]+ Patients with suspected chronic obstructive pulmonary disease (COPD) should have spirometric parameters checked

    before bronchoscopy and, if the COPD is found to be severe (FEV1

  • + Longer immersion times of 60 minutes are recommended for known or suspected atypical mycobacterial infections andpatients known to be HIV positive with respiratory symptoms as they may be infected with Mycobacterium avium intracellu-lare or other atypical mycobacteria which are more resistant to glutaraldehyde. [B]

    + Patients with suspected tuberculosis should undergo bronchoscopy at the end of the list. [C]+ Automated washer disinfectors are recommended to minimise staV contact with disinfectants and their fumes. [B]+ Automated washer disinfectors must have facilities for disinfecting tanks, immersion trays, and all fluid pathways. [B]+ It is essential that sterile or bacteria free water is used for rinsing bronchoscopes; autoclaved or filtered water (using 0.2 µm

    filters) may be used. [B]+ All rinse water pathways (tanks, filters, and pipework) must be accessible for regular, preferably sessional, cleaning and dis-

    infection. [B]+ Some water-borne mycobacteria such as Mycobacterium chelonae are extremely resistant to glutaraldehyde and a chlorine

    releasing agent or peracetic acid may have to be used via the water filters. [B]+ A record should be kept of which bronchoscope and other re-usable equipment are used on an individual patient and also

    of the decontamination procedure. [C]+ The quality of the rinse water should be assured, but if this is in doubt then external surfaces of the bronchoscope should

    be wiped and the lumen flushed with 70% alcohol. This will destroy non-sporing bacteria including mycobacteria and willrapidly evaporate, leaving surfaces dry. This is also recommended at the end of a session and/or before storage. [B]

    + Glutaraldehyde, although widely used for endoscopes, is only slowly eVective against mycobacteria. Peracetic acid, chlorinedioxide, and superoxidised water are more rapidly eVective (within 5 minutes or less) but are more damaging to instrumentsand processing equipment, are less stable, and are more expensive. [B] They may, however, be less irritant than glutaraldehyde.

    StaV safety+ All staV should be vaccinated against hepatitis B and tuberculosis, and immunity and tuberculin status should be checked

    as appropriate. [B]+ During bronchoscopy staV should wear protective clothing (gowns or plastic aprons, masks/visors, and gloves). [C]+ High grade particulate masks should be worn when patients known to have multidrug resistant tuberculosis or those at high

    risk undergo bronchoscopy and the procedure should be carried out in a negative pressure facility. [C]+ Non-powdered latex or non-latex gloves should be worn instead of powdered latex gloves. [B]+ Injection needles should not be re-sheathed, and spiked biopsy forceps require careful cleaning. [C]+ Pre-employment health checks should be carried out on all staV working with aldehydes according to COSHH recommen-

    dations, and regular periodic screening with regard to lung function and occurrence of symptoms should be carried out bythe occupational health department. [C]

    + Bronchoscopes should be disinfected ideally in a dedicated room using well ventilated automated systems, preferably insidea fume cabinet, to prevent unnecessary exposure to disinfectants. [C]

    + During cleaning and disinfection staV need to wear protective clothing (nitrile gloves and plastic aprons with eye and res-piratory protection, where appropriate) to protect them from splashes, aerosols, and vapour. [C]

    + The use of disposable accessories, especially injection needles, may reduce the risk of infection which may occur during thecleaning of equipment. [C]

    + Wherever possible, autoclavable or disposable accessories should be used to prevent unnecessary exposure to disinfectants. [C]+ Bronchoscopy staV need to be trained in patient care, infection control, and instrument decontamination including the safe

    use of aldehydes and the potential health risks. [C]

    Bronchoscopy in the intensive care unit (ICU)+ The internal diameter of the endotracheal tube, through which the bronchoscope is inserted, must be taken into considera-

    tion before bronchoscopy.+ Intensive care units should have the facility to perform urgent and timely flexible bronchoscopy for a range of therapeutic

    and diagnostic indications. [C]+ Patients in ICU should be considered at high risk from complications when undergoing fibreoptic bronchoscopy. [B]+ Continuous multi-modal physiological monitoring must be continued during and after fibreoptic bronchoscopy. [B]+ Care must be exercised to ensure adequate ventilation and oxygenation is maintained during fibreoptic bronchoscopy via

    an endotracheal tube. [B]+ More profound levels of sedation/anaesthesia can be achieved in ventilated patients provided the clinician performing the

    procedure is acquainted with the use of sedative/anaesthetic agents. [C]

    Standards and performance of diagnostic techniques+ At least five bronchial biopsy specimens should be taken in cases of suspected bronchial malignancy. [B]+ Biopsies, brushings and washings should all be obtained in cases of suspected endobronchial malignancy. [B]+ A minimum diagnostic level of at least 80% should be obtained from a combination of biopsies, brushings, and washings in

    cases of endoscopically visible malignancy. [B]+ When taking transbronchial lung biopsy specimens in patients with diVuse lung disease, an attempt should be made to

    obtain 4–6 samples from one lung. [B]

    Patient care+ Verbal and written patient information improves tolerance of the procedure and should be provided. [B]+ It is suYcient for patients to have no food by mouth for 4 hours and to allow clear fluids by mouth up to 2 hours before

    bronchoscopy. [B]+ Patients who have been sedated should be advised not to drive, sign legally binding documents, or operate machinery for

    24 hours after the procedure. [C]

    British Thoracic Society guidelines on diagnostic flexible bronchoscopy i3

  • Complications, precautions andcontraindicationsFITNESS FOR THE PROCEDURE

    There are no controlled studies of the factorswhich may make a patient unfit for broncho-scopy so a decision to carry out the procedureis a balance between the likely benefit ofobtaining diagnostic material (including thera-peutic benefit) and an assessment of the likelyrisk in that individual patient.

    Flexible bronchoscopy is an extremely safeprocedure as long as some basic precautionsare taken.4 One study reported a mortality rateof 0.01% and a major complication rate of0.08% in a series of 24 521 procedures,5 andanother a 0.02% mortality and 0.3% majorcomplication rate in a series of around 48 000cases.6 A survey in the UK found a mortalityrate of 0.04% and a major complication rate of0.12% in about 40 000 procedures.7 Thesethree were retrospective studies. Smallerprospective studies have reported slightlyhigher rates.8 9 A more recent retrospectivestudy of over 4000 cases, including 2000lavages and 173 transbronchial biopsies,showed no deaths and overall major and minorcomplication rates of 0.5% and 0.8%, respec-tively.10 Flexible bronchoscopy with topicalanaesthesia has been shown to be safer thanrigid bronchoscopy.11

    Major life threatening complications includerespiratory depression, pneumonia, pneumo-thorax, airway obstruction, cardiorespiratoryarrest, arrhythmias, and pulmonary oedema.Minor non-life threatening complications in-clude, in order of frequency, vasovagal reac-tions, fever, cardiac arrhythmias, haemorrhage,airway obstruction, pneumothorax, nausea andvomiting.9

    Complication rates relating specifically tothe procedure of transbronchial biopsies arehigher—for example, pneumothorax occurredin 1–5% of cases12 and haemorrhage in 9%,13

    which was usually mild. Haemorrhage is morelikely in uraemic or immunosuppressed pa-tients.14 Overall mortality has been reported as0.1%.12 Major complications occurred in 6.8%in another series.10

    SIDE EFFECTS FROM LOCAL ANAESTHESIA (SEEALSO SECTION ON “SEDATION AND LOCALANAESTHESIA”)Lignocaine (lidocaine) is the topical agent usu-ally used. Toxic side eVects include seizuresand cardiac suppression. A threshold abovewhich these side eVects become more likely hasbeen put at a plasma level of 5 mg/l.15

    Lignocaine is mainly metabolised in the liverand this has led to worries about potential tox-icity in patients with metastatic malignancyinvolving the liver. A study of a small number ofsuch patients showed no increased risk in toxicblood levels, however caution should still bestressed because none of their patients hadhyperbilirubinaemia or clinical evidence ofhepatocellular failure.16 Lignocaine instilleddirectly into the endobronchial tree is rapidlyabsorbed and a correlation has been reportedbetween the dose given and the serum level.17 Avariable amount of lignocaine is, however,

    re-aspirated through the bronchoscope, indi-vidual hepatic metabolism varies, and otherdrugs such as cocaine may delay absorption.Clinical toxicity is seldom observed, even indoses above 8 mg/kg.17 However, particularcare should be taken in those patients withhepatic or cardiac insuYciency in whom ligno-caine metabolism may be impaired; a maxi-mum dose of 5 mg/kg has been suggested insuch cases.17 One study, which also took intoaccount levels of an active metabolite of ligno-caine, concluded that the total dose oflignocaine should not exceed 7 mg/kg.17 Amore recent study has recommended a maxi-mum dose of 8.2 mg/kg,18 which is equivalentto 29.3 ml 2% lignocaine for a 70 kg patient.

    Recommendations+ The total dose of lignocaine (lidocaine)

    should be limited to 8.2 mg/kg in adults(approximately 29 ml of a 2% solutionfor a 70 kg patient) with extra care inthe elderly or those with liver or cardiacimpairment. [B]

    + The minimum amount of lignocainenecessary should be used when instilledthrough the bronchoscope. [B]


    Hypoxaemia has been shown to occur duringflexible bronchoscopy in several studies. A fallin arterial oxygen tension (PaO2) of about2.5 kPa during the procedure is common.19 20

    This can be more pronounced if broncho-alveolar lavage is performed and is more com-mon with larger volumes of lavage fluid.21 Themechanisms causing hypoxia includeventilation-perfusion imbalance and hypoven-tilation secondary to sedation. Ventilation-perfusion mismatch may occur as a result ofpartial airway obstruction caused by the bron-choscope, suction, and due to anaesthetic solu-tions or lavage fluid in the alveoli. Correction ofanaemia should be considered before bron-choscopy. The hypoxaemia has been linked toan increased risk of arrhythmias; pulse oxi-metry increases the safety of bronchoscopyand has been recommended for routine use.22

    Stridor can occur during the procedure andmay lead to serious hypoxaemia. A safety oxy-gen saturation (SaO2) threshold of 90% (at restbreathing room air) has been proposed,normally equivalent to a PaO2 of 8 kPa.

    22 Ifsedation is given well before bronchoscopy (forexample, peroral diazepam), monitoring byoximetry will be required before bronchoscopyis started.

    Hypoxaemia may last for a considerable timeafter bronchoscopy has been completed. Oxy-gen supplementation after bronchoscopy hasbeen found to be beneficial, particularly insubjects with impaired lung function.23 Theduration of postoperative oxygen supplementa-tion partly depends on the type of sedationpreviously given23; a longer period of supple-mentation (1–2 hours) is required if the patienthas had oral diazepam rather than intravenousmidazolam, or if the patient has received anamnesic dose of midazolam.

    i4 BTS Bronchoscopy Guidelines Committee

  • Oxygen supplementation via nasal cannulaeshould be at a flow rate of at least 2 l/min.22

    Care should be taken, however, to be alert tosigns of respiratory failure in patients onoxygen supplementation who may have “safe”oximetry readings but who may be developingcarbon dioxide retention. Particular careshould be taken in those patients at higher riskof carbon dioxide retention and monitoring oftranscutaneous carbon dioxide and oxygenpressures may be useful.24 Preoperative arterialblood gas assessment is usually required insuch patients. In high risk hypoxaemic patientsrequiring bronchoscopy and lavage, non-invasive positive pressure ventilation via a facemask can be used.25

    Recommendations+ Oxygen supplementation should be used

    to achieve an oxygen saturation of atleast 90% to reduce the risk of significantarrhythmias during the procedure andalso in the postoperative recoveryperiod. [B]

    + Postoperative oxygen supplementationmay be required in some patients, par-ticularly those with impaired lung func-tion and those who have been sedated.[B]

    + Patients should be monitored by oxi-metry. [B]


    Arrhythmias and cardiac arrest have beendescribed during fibreoptic bronchoscopy.5 6

    Several studies have involved continuous ECGmonitoring during the procedure and somealso monitored pre and postoperatively. Amarked tachycardia during bronchoscopy wasfound in many patients in one study.26 Anotherfound that hypoxia at the end of the procedurecorrelated with occurrence of major arrhyth-mias,27 11% of patients developing significantbut self-limiting arrhythmias. The authors re-commended ECG monitoring during the pro-cedure in those patients with an abnormal12-lead ECG and preoperative hypoxaemia(PO2 50 ml) occurred in 1.6–4.4%of patients with diVuse lung disease fromwhom transbronchial biopsy specimens weretaken.33 34 The risk of bleeding during trans-bronchial biopsies seems to be unrelated to sizeof forceps35 but is slightly higher in those beingmechanically ventilated.36

    Some patients are known to be at increasedrisk of bleeding, including those who haveuraemia, immunosuppression, pulmonaryhypertension, liver disease, coagulation dis-orders, or thrombocytopenia.37 38 Concern isoften expressed about performing a broncho-scopic examination in patients with evidence ofsuperior vena caval obstruction, particularlybecause of worries about a possible increasedrisk of bleeding. However, no studies on thissubject were found in the literature. Routinepreoperative coagulation screening is unjusti-fied in those with no risk factors who undergoroutine bronchoscopy.39 Conversely, in patientswith a potential increased risk of bleeding—forexample, those with abnormal liver functiontests—platelet count, prothrombin time, andpartial thromboplastin time should be checkedbefore bronchoscopy. Information is unclearabout their value when transbronchial biopsyspecimens are taken,40 although many physi-cians routinely check the preoperative plateletcount and prothrombin and partial thrombo-plastin times. There is no information aboutwhat constitutes a “safe” level for clotting inthis context. If persistent bleeding occurs,turning the patient onto the side of thebleeding and topical instillation of cocaine orsmall amounts of 1:10 000 adrenaline solutionshould be considered.

    If a patient is taking oral anticoagulants,published guidelines for managing anticoagu-lation in the perioperative period are relevant.41

    These state that “the short termriskof thrombo-embolism in patients with mechanical heartvalves when not anticoagulated is very small”.

    Recommendations+ Routine preoperative checks of the

    platelet count and/or prothrombin timeare only required in those patients with

    British Thoracic Society guidelines on diagnostic flexible bronchoscopy i5

  • known risk factors having routine bron-choscopy without transbronchial bi-opsy. [B]

    + If it is anticipated that biopsy specimensmay be required at bronchoscopy, oralanticoagulants should be stopped atleast 3 days before bronchoscopy orthey should be reversed with low dosevitamin K. [B]

    + On the rare occasions when it is neces-sary to continue with anticoagulants,the international normalised ratio(INR) should be reduced to

  • reported in another recent retrospectivestudy.65 Others have emphasised extreme cau-tion66 and have stressed the increased risk in thepresence of poor left ventricular function andcongestive cardiac failure.67 The risks ofbronchoscopy are thought to be reduced 4–6weeks after myocardial infarction.68

    Recommendation+ Bronchoscopy should be avoided if

    possible within 6 weeks of a myocardialinfarction. [C]

    Patients with asthmaGenerally, bronchospasm complicating fibre-optic bronchoscopy is rare (0.02% in oneseries10). However, another study of 216asthmatic subjects reported that 8% developedlaryngospasm or bronchospasm during theprocedure.69 Lignocaine may producebronchoconstriction in patients with asthmaand this is attenuated by preoperative treat-ment with atropine.70

    A more pronounced postoperative fall inforced expiratory volume in one second (FEV1)compared with normal subjects was found inpatients with mild asthma during broncho-scopy, and this fall in FEV1 was inverselycorrelated with the preoperative concentrationof methacholine provoking a fall in FEV1 of20% or more (PC20).

    71 Sedation should there-fore be used with particular care in asthmaticpatients because of the risk that the broncho-scopic procedure may exacerbate bronchocon-striction.

    Another study showed no diVerence in thedegree of intraoperative fall in SaO2 betweennormal subjects and asthmatic patients (with aspectrum of severity) but the falls in both FEV1and forced vital capacity (FVC) were greater inthe asthmatic patients after a sequence ofbronchoalveolar lavage and bronchial biop-sies.72 The preoperative use of a bronchodilatorwas associated with no fall in the postoperativeFEV1 in a study of patients with mild asthma inwhom bronchoalveolar lavage was performed.73

    Recommendation+ Asthmatic subjects should be premedi-

    cated with a bronchodilator beforebronchoscopy. [B]

    Patients with chronic obstructive pulmonarydiseaseThe presence of chronic obstructive pulmo-nary disease (COPD) has been shown toincrease the complication rate of bronchoscopy(where FEV1/FVC

  • Although bronchoscopy can be carried outwithout sedation,84 85 most are performedunder sedation.7 86 87 However, the routine useof sedation is not a prerequisite based on theavailable evidence.88 Sedation should benefitthe patient who is particularly anxious or whoexpresses a strong preference for sedation.Conversely, sedation should be avoided or usedwith extreme caution in patients such as thosewith severe COPD who have an increased riskof responding adversely (see section on “Com-plications, precautions and contraindica-tions”). It is essential that all syringes are care-fully labelled.

    In a few of the randomised controlled studiesof sedation in bronchoscopy a significantproportion of patients receiving midazolamwere poorly sedated.88 89 If sedation is inade-quate, up to 60% of patients may find the pro-cedure unpleasant90 and up to 25% are unwill-ing to undergo repeat investigation.76 Whensedation is titrated to induce a light sleep,patient acceptance of bronchoscopy is high.91

    Arguments against sedation cite reducedpatient safety, although this can be largelyobviated by appropriate monitoring and theavailability of proper resuscitation equip-ment.91 92 Combinations of benzodiazepinesand narcotics, although widely used,61 areprone to induce hypoxia and CO2 retention.

    24 93

    Which sedative?Although premedication with oral lorazepamdoes have a significant amnesic eVect com-pared with placebo,94 most sedation regimensare based upon a single dose or incrementaldoses of an intravenous sedative agent admin-istered at the time of bronchoscopy.


    Midazolam is a water soluble benzodiazepinewith an elimination half life of about 2 hoursand is generally preferred to diazepam. Itsonset is rapid and duration of action brief inhealthy individuals. Less than 10% of thepopulation exhibit prolonged eVects due toimpaired metabolism.95 Many investigatorshave used a dose of 0.07 mg/kg midazolambefore bronchoscopy. However, sedation can-not be assured by single dose sedationregimens.96 A better approach is incrementaldosing which achieves improved tolerance ofbronchoscopy, induces amnesia, and thereforepatients are more willing to undergo repeatbronchoscopy if necessary.91 In this latter studythe dose of midazolam ranged from 0.07 to0.67 mg/kg. Particular care is required in thosewith severe COPD or those with neuromus-cular diseases. However, adequate sedation andamnesia can be achieved by smaller doses suchas an initial dose of 2 mg followed after 2 min-utes by increments of 1 mg/min if required.97

    Although amnesia may be important forpatient acceptance of bronchoscopy,91 94 thereare no studies that directly compare sedativedrugs in doses suYcient to produce completeamnesia with lower doses producing a lesserdegree of sedation.


    Flumazenil is a specific benzodiazepine antago-nist that can reverse oversedation and shouldbe available in the endoscopy unit. However, itsshort elimination time means that re-sedationmay occur unless repeated doses or an infusionare given. Flumazenil can induce withdrawaland seizures in patients on long term benzo-diazepine therapy. The usual initial dosage is250–500 µg. Specific antagonists are no substi-tute for careful observation in an appropriatelyequipped recovery area.


    Propofol (2,6 di-isopropylphenol) is an emul-sion formulation suitable for induction andmaintenance of anaesthesia. In three ran-domised controlled studies propofol has beenshown to produce adequate sedation, which isof rapid onset and resolution. It appears tooVer advantages over midazolam anddiazepam/alfentanil in combination.97–100 Theaverage sedative dose administered in onestudy was 155 mg, equivalent to an anaestheticinduction dose.97 In that study a microproces-sor controlled infusion device was used toachieve target blood levels. Propofol doesappear to oVer advantages over some othersedative agents but is expensive and requiresexpertise and experience in its administration.


    A combination of a benzodiazepine andnarcotic has been widely used.101 102 Combiningthe amnesic eVects of a benzodiazepine withthe analgesic and antitussive eVects of anarcotic is rational. Unfortunately, such acombination may be associated with morearterial desaturation and CO2 retention thanwhen using midazolam alone.24 93 Morphine iscommonly used but synthetic shorter actingnarcotics may be preferred. The combinationof nalbuphine (0.2 mg/kg) and midazolam(0.05 mg/kg) produced slightly higher CO2levels than midazolam alone.102 Alfentanil is asynthetic, potent and short lived narcoticwhich, when used alone in doses of 0.05–1.0 mg, achieves equivalent levels of sedationto midazolam with greater antitussive eVects.93

    Recommendations+ Sedation should be oVered to patients

    where there is no contraindication. [B]+ Sedatives should be used in incremental

    doses to achieve adequate sedation andamnesia. [B]

    LOCAL ANAESTHESIA (SEE ALSO SECTION ON“COMPLICATIONS”)Lignocaine (lidocaine)Lignocaine (lidocaine) is the most commonlocal anaesthetic for use in fibreoptic broncho-scopy and has a better safety profile thancocaine.103 Anaesthesia of the anterior narescan be achieved using 10% or 4% w/vlignocaine aerosol or a 2% w/v gel. The gelpreparation may be better accepted by patientsand yield lower blood levels.104–106 The orophar-ynx can be anaesthetised with 10% lignocainespray,107 or by sucking a 60 mg amethocaine

    i8 BTS Bronchoscopy Guidelines Committee

  • lozenge. Anaesthesia of the vocal cords can beachieved in several ways. Having introducedthe bronchoscope through the nose andnasopharynx, the visualised vocal cord can besprayed under direct vision with 2–4% ligno-caine (“spray as you go”). Thereafter, havingtraversed the cords, anaesthesia of the carinaand bronchi can be achieved using 1–2%boluses sprayed via the bronchoscope. Alterna-tively, a transcricoid injection of 5 ml 2%lignocaine will anaesthetise the vocal cords aseVectively as under direct vision with a lowerdose of anaesthetic.108 109 However, almost athird of the patients undergoing transcricoidinjection found the procedure unpleasant.108

    Inhalation of 4 ml 4% nebulised lignocaine viaa mouthpiece can also produce satisfactoryanaesthesia of the oropharynx and vocalcords.109

    Toxic blood levels (>5 mg/l) or signs of tox-icity are uncommon when using topicallignocaine during bronchoscopy.17 105 Themaximum dose of 8.2 mg/kg body weight sug-gested in a recent study19 is higher than thatrecommended when infiltrating lignocaine(

  • choscopy are Gram negative bacilli, mycobac-teria, and viruses.

    Gram negative bacilliGram negative bacilli, particularly Pseudomonasaeruginosa, are most frequently responsible forrespiratory infections and pseudoinfections inthose undergoing bronchoscopy.122–124 Theseorganisms flourish in warm moist sites such assinks, water bottles, and aspiration and clean-ing equipment. In addition to their presence inthe environment, P aeruginosa colonises severalbody sites and is isolated from throat swabs,sputum, and faeces. Other Gram negativebacilli associated with bronchoscope contami-nation and occasionally infection are Serratiamarcescens and Proteus spp. Inadequate clean-ing, the use of an inappropriate disinfectant,and contaminated rinse water are the mostlikely causes.125–127

    A pseudoepidemic of Legionella pneumophilaserogroup 6 contaminating bronchoalveolarlavage specimens has also been traced to tapwater used to rinse bronchoscopes afterdisinfection.128

    MycobacteriaThe emergence of multidrug resistant strains ofMycobacterium tuberculosis and the muchgreater incidence of infections with M aviumintracellulare among HIV infected patients, par-ticularly in the USA, has led to a greaterawareness of the risk of transmission of myco-bacteria during bronchoscopy, particularly inHIV infected or severely immunocompromisedpatients.129 Other species of mycobacteria arealso being incriminated in human infectionwith increasing frequency.

    Reports of the transmission of M tuberculosisand other patient associated mycobacteria arefew130–133 and have largely been attributed to afailure to clean the bronchoscope, particularlythe suction valve, or to introduce a suitablyeYcacious disinfectant into the lumen of theinstrument. Non-tuberculous mycobacteriasuch as M chelonae, M kansasii, and M fortuitumare most likely to be transmitted via contami-nated washer disinfectors and rinse water.134–137

    Although these organisms appear to be lesspathogenic than M tuberculosis, they occasion-ally cause infections, particularly in severelyimmunocompromised patients. Some isolatesof M chelonae have been shown to be extremelyresistant to glutaraldehyde, the disinfectantmost widely used for bronchoscopes.138

    A misdiagnosis of tuberculosis is made ifcontaminated water is used to remove disin-fectant residues as acid fast bacilli are depos-ited on surfaces and within the lumen of thebronchoscope. These may have been acquiredfrom the water supply or from the washer dis-infector. If such contamination is suspected,parallel samples should be sent for microscopicexamination and culture. Mycobacteria, par-ticularly M avium intracellulare,139 are moreresistant to disinfectants than other non-sporing bacteria and viruses and this must beconsidered when selecting disinfectants andformulating policy.140

    VirusesBlood-borne viruses such as hepatitis B(HBV), C (HBC), and HIV may be present inblood and most body fluids including salivaand alveolar fluid. In spite of the high degree ofinfectivity of HbeAg positive patients, docu-mented transmission of HBV during endo-scopy is rare with only one convincing casehaving so far been reported.141 In this case theendoscope was immersed in glutaraldehyde for21 hours but the air and water channels werenot disinfected by adequate irrigation with dis-infectant. The possible transmission of hepati-tis C virus has been reported after colonoscopybut guidance on instrument decontaminationwas not followed.142

    HIV has been isolated from endoscopes afteruse in patients with AIDS. In one study bron-choscopes were sampled after use for the pres-ence of HIV using polymerase chain reactiontechniques and culture infectivity assays.143 Allseven bronchoscopes sampled immediatelyafter use showed the presence of HIV but, aftercleaning with a detergent, only two of 88 sitessampled yielded HIV and numbers were small.No cases of HIV transmission have so far beenreported following endoscopy.

    Miscellaneous organismsBacillus spp and fungi have been recoveredfrom bronchial washings,144 145 none of whichwere associated with clinical infection. Theprobable sources were a contaminated suctionvalve and contaminated medicaments.

    Recommendations+ Decontamination and disinfection

    should be carried out at the beginningand end of a list and between patients.[B]

    + Cleaning and disinfection of broncho-scopes should be undertaken by trainedstaV in a dedicated room. [C]

    + Thorough cleaning with detergent re-mains the most important initial stageof the process. [B]

    + When 2% glutaraldehyde is used formanual and automated disinfection,immersion for 20 minutes is recom-mended for bronchoscopes at the be-ginning and end of a session andbetween patients. This will destroy mostvegetative bacteria including Mycobac-terium tuberculosis and viruses. [B]

    + Longer immersion times of 60 minutesare recommended for known or sus-pected atypical mycobacterial infec-tions and patients known to be HIVpositive with respiratory symptoms asthey may be infected with Mycobacte-rium avium intracellulare or otheratypical mycobacteria which are moreresistant to glutaraldehyde. [B]

    + Patients with suspected tuberculosisshould undergo bronchoscopy at theend of the list. [C]

    + Automated washer disinfectors are re-commended to minimise staV contactwith disinfectants and their fumes. [B]

    i10 BTS Bronchoscopy Guidelines Committee

  • + Automated washer disinfectors musthave facilities for disinfecting tanks,immersion trays, and all fluid path-ways. [B]

    + It is essential that sterile or bacteriafree water is used for rinsing broncho-scopes; autoclaved or filtered water(using 0.2 µm filters) may be used. [B]

    + All rinse water pathways (tanks, filters,and pipework) must be accessible forregular, preferably sessional, cleaningand disinfection. [B]

    + Some water-borne mycobacteria suchas Mycobacterium chelonae are ex-tremely resistant to glutaraldehyde anda chlorine releasing agent or peraceticacid may have to be used via the waterfilters. [B]

    + The quality of the rinse water should beassured, but if this is in doubt thenexternal surfaces of the bronchoscopeshould be wiped and the lumen flushedwith 70% alcohol. This will destroynon-sporing bacteria including myco-bacteria and will rapidly evaporate,leaving surfaces dry. This is also rec-ommended at the end of a sessionand/or before storage. [B]

    + Glutaraldehyde, although widely usedfor endoscopes, is only slowly eVectiveagainst mycobacteria. Peracetic acid,chlorine dioxide, and superoxidisedwater are more rapidly eVective (within5 minutess or less) but are moredamaging to instruments and process-ing equipment, are less stable, and aremore expensive. [B] They may, how-ever, be less irritant than glutaralde-hyde.

    + Compatibility of decontaminationmethods should be checked with themanufacturers of bronchoscopic in-struments and accessories. [C]

    + A record should be kept of which bron-choscope and other re-usable equip-ment are used on an individual patientand also of the decontamination proce-dure. [C]

    StaV safetyBronchoscopy poses a risk for staV both duringthe procedure and during disinfection of theinstruments.


    StaV are potentially at risk of infections duringbronchoscopy, particularly with regard tohepatitis, HIV, and M tuberculosis. Needle stickinjuries are potential sources of blood-borneinfections, while cough inducing proceduressuch as bronchoscopy pose a risk from airborneM tuberculosis. Accidental needle stick injuriesare the greatest risk to endoscopy staV146 andcan occur during re-sheathing of needles,cleaning of spiked biopsy forceps, and theremoval of biopsy specimens from the forceps.Biopsy material can be removed from thebiopsy forceps by carefully shaking in thebiopsy pot with formalin. The use of hypoder-mic needles should be discouraged and suit-

    able clean and sterile alternatives should besought—for example, sterile tooth picks. Singleuse accessories should be employed whereverpossible.121

    StaV involved with hazardous proceduresmust be provided with protective clothing/equipment and instructed in its correct use.147

    Cross infection can also occur from surfacesand from poor hand hygiene.148 149 Goodhygiene—such as appropriate hand washing,cleaning and disinfection of surfaces—will pro-tect staV and patients from potential crossinfection such as methicillin resistant Staphylo-coccus aureus (MRSA).

    Enhanced protection against tuberculosisshould be shown to be present by either a posi-tive tuberculin test or the presence of a definiteBCG scar.150 Immunity to hepatitis B needs tobe demonstrated by the presence of surfaceantibody.151

    Recommendations+ Injection needles should not be re-

    sheathed, and spiked biopsy forcepsrequire careful cleaning. [C]

    + The use of disposable accessories, espe-cially injection needles, may reduce therisk of infection which may occurduring the cleaning of equipment. [C]

    + All staV should be vaccinated againsthepatitis B and tuberculosis, and im-munity and tuberculin status should bechecked as appropriate. [B]


    Examination gloves protect staV from potentialcontamination but may not protect patientsand must be changed after each procedure.Appropriate selection of gloves should be madeavailable for those workers who have developedan allergy to latex.152 Individually sterilisedgloves are generally unnecessary unless dealingwith immunocompromised patients. Latexsensitisation is an occupational hazard inhealthcare workers.153 Powdered latex glovesare recognised as a major cause of latex in theair as latex binds to starch particles.154 The risksof latex sensitisation to staV and patientsshould be reduced by using non-powderedlatex gloves155 or non-latex gloves.

    Cough inducing procedures such as bron-choscopy may be associated with a high risk ofexposure to microorganisms including Mtuberculosis.156 Previous recommendations forrespiratory protection during bronchoscopytherefore have included the wearing of facemasks or high grade particulate face masks(“respirators”).156 There is a lack of dataregarding the eVectiveness of personal respira-tory protection in stopping transmission ofinfections in the health care setting. Visors mayprotect the face of staV from splashes coughedup during bronchoscopy, although they maynot adequately protect staV from inhalingaerosols of M tuberculosis. Similarly, there is alack of data in the literature about the relativemerits of eye protectors.

    High grade particulate face masks have beenrecommended where there is thought to be arisk of multidrug resistant tuberculosis.150 A

    British Thoracic Society guidelines on diagnostic flexible bronchoscopy i11

  • recent audit61 showed that face masks werepredominantly worn by operating theatre staV,but less often in other departments. However,medical prudence would suggest that masks/visors are a reasonable approach to providesome protection to healthcare workers.156 Ifbronchoscopy is required for a patient withknown or suspected multiresistant tuberculo-sis, this should be carried out in a room fittedwith negative pressure ventilation.157

    Gowns or waterproof aprons may protectstaV uniforms from contamination. Longsleeve gowns may provide further protectionalthough there are no studies to confirm this.

    Recommendations+ During bronchoscopy staV should wear

    protective clothing (gowns or plasticaprons, masks/visors, and gloves). [C]

    + Non-powdered latex or non-latex glovesshould be worn instead of powderedlatex gloves. [B]

    + High grade particulate masks should beworn when patients known to havemultidrug resistant tuberculosis orthose at high risk undergo broncho-scopy and the procedure should be car-ried out in a negative pressure facility.[C]


    Glutaraldehyde is a highly eVective, low costdisinfectant which is non-damaging to instru-ments and processing equipment. However, itdoes pose a potential risk to hospital staV.Aldehydes are irritants and sensitisers and thevapour can cause rhinitis, conjunctivitis, andasthma, while contact with the disinfectant cancause dermatitis.117 It has been stated that, ifthe smell of glutaraldehyde is detectable, thelimits are likely to be exceeded and action mayneed to be taken.117 A new legally enforceablemaximum exposure limit of 0.05 ppm has nowbeen published158 to take the place of the previ-ous occupational exposure standard (0.2 ppmexpressed as a 15 minute reference period159). Itshould be stressed that no safe level of exposurehas been identified.

    StaV and patients are exposed to aldehydesduring bronchoscopy if the instruments havebeen inadequately rinsed, and risks fromaccessory disinfection and exposure can bereduced if exhaust vented automated systemsare used and accessories are autoclaved or dis-posable accessories are used. The disinfectionprocess needs to be carried out in a dedicatedwell ventilated environment with local exhaustventilation,117 preferably in a fume cabinet in anarea separate from the main endoscopy room.Only in the absence of adequate control meas-ures should personal protective equipment beused.160 Personal protective equipment such asnitrile gloves, impermeable aprons, respiratoryprotective equipment, and eye protectionshould be worn when mixing aldehydes anddealing with spillages. All other processesincluding filling of machines and disposal ofthe chemical should be carried out with localexhaust ventilation,161 preferably directly fromthe automated system into the drains.

    All staV should be informed of the potentialhealth risks associated with glutaraldehydeuse160 and must receive training in the safe useof the disinfectant. Safe working procedureshave to be established in writing and given toall staV concerned. Only trained and compe-tent staV should be authorised to use alde-hydes.117

    Recommendations+ Pre-employment health checks should

    be carried out on all staV working withaldehydes according to COSHH recom-mendations, and regular periodicscreening with regard to lung functionand occurrence of symptoms should becarried out by the occupational healthdepartment. [C]

    + Bronchoscopes should be disinfectedideally in a dedicated room using wellventilated automated systems, prefer-ably inside a fume cabinet, to preventunnecessary exposure to disinfectants.[C]

    + During cleaning and disinfection staVneed to wear protective clothing (nitrilegloves and plastic aprons with eye andrespiratory protection, where appropri-ate) to protect them from splashes,aerosols, and vapour. [C]

    + Wherever possible, autoclavable or dis-posable accessories should be used toprevent unnecessary exposure to disin-fectants. [C]

    + Bronchoscopy staV need to be trainedin patient care, infection control, andinstrument decontamination includingthe safe use of aldehydes and the poten-tial health risks. [C]

    Flexible bronchoscopy in the intensivecare unitThe flexible bronchoscope facilitates the in-spection of the upper airways and bronchialtree in critically ill patients in an intensive careunit (ICU). In non-intubated patients the risksof bronchoscopy are those described in thesection on “Complications”. However, bron-choscopy in the ICU commonly involves intu-bated patients supported by mechanical venti-lation. The specifications of a bronchoscopesuitable for general ICU work requires a degreeof compromise. The internal diameter of theendotracheal tube may restrict the size ofbronchoscope, while eYcient suctioning re-quires a larger bronchoscope with a widesuction channel.

    Recommendation+ Intensive care units should have the

    facility to perform urgent and timelyflexible bronchoscopy for a range oftherapeutic and diagnostic indications.[C]


    Fibreoptic bronchoscopy may be performed inthe ICU for a wide range of diagnostic ortherapeutic indications.92 162 Often, broncho-scopy is used to investigate and rectify lobar

    i12 BTS Bronchoscopy Guidelines Committee

  • collapse that has failed to respond to measuressuch as physiotherapy.92 163–165 Retained bron-chial secretions may obstruct major airwaysand predispose towards infection. Local di-rected suctioning, particularly with a widechannel bronchoscope, combined with salineor acetylcysteine instillation is very eVective atremoving these secretions. Similarly, some for-eign bodies such as food material or tooth frag-ments can be removed with a wire basket orgrasped with bronchoscopy forceps.

    Minor endotracheal bleeding is a commonfinding during routine tracheal suctioning andmay result from tracheal epithelial abrasions. Ifbleeding becomes persistent or excessive,bronchoscopy may identify the source andextent of the haemorrhage and assist indeveloping a management plan.92 166 Massivehaemorrhage renders fibreoptic inspection dif-ficult and rigid bronchoscopy is then generallypreferred.

    Some ICUs use bronchoscopically directedlavage or brushing techniques to obtain micro-biological samples in patients with pneumo-nia.167 168 Whether directed techniques oVer sub-stantial advantages over non-directed methodshas not been conclusively demonstrated.168 169

    Bronchial lavage for microbiological specimensappears to be a relatively safe procedure withoutlasting or serious sequelae.170 One report hassuggested that the physical appearance of secre-tions welling up from distal airways aftersuctioning is characteristic of pneumonia, incontrast to bronchitis.169

    Transbronchial biopsy may be required forhistological diagnosis.171 In the ventilated pa-tient there is a significant risk of pneumothorax(approximately 10%) and haemorrhage (ap-proximately 5%), and histological diagnosismay only be achieved in about one third ofcases.36


    Risk assessmentCritically ill patients represent a high riskgroup for most invasive procedures. They willoften present with hypoxia, electrolyte distur-bances, clotting abnormalities and arrhyth-mias.170 It is therefore important that thepotential benefits of bronchoscopy outweighthe risks. Elevated prothrombin time, increasedactivated partial thromboplastin time (APTT),reduced fibrinogen titre, or thrombocytopeniaindicate clotting dysfunction making biopsyprocedures hazardous. Brushing or lavage forcytological and microbiological examinationsmay oVer a safer alternative.167 170 The samereservations apply to patients with renal failurein whom platelets may be dysfunctional. Criti-cally ill patients may be more susceptible to thetoxic eVects of local anaesthetics172; however, inthe ventilated patient intravenous sedation oranaesthesia is probably the most appropriatealternative.

    Recommendation+ Patients in ICU should be considered at

    high risk from complications whenundergoing fibreoptic bronchoscopy.[B]

    MonitoringA full range of physiological monitoring willgenerally be available in the ICU. This shouldinclude ECG (for heart rate and rhythm), con-tinuous intra-arterial blood pressure or inter-mittent cuV blood pressure measurement, andpulse oximetry (SpO2). Setting appropriatealarm limits for heart rate, blood pressure andSpO2 and requesting other attendant staV tomonitor physiological variables during thebronchoscopy improves safety. Adverse eventsrequire immediate withdrawal of the broncho-scope and resuscitation of the patient. The cli-nician must then weigh the benefits against therisks of proceeding further.

    Monitoring intracranial pressure (ICP) inhead injured patients is essential if sudden risesin ICP are to be avoided due to CO2 retentionor other causes. Monitoring endotracheal CO2in such patients may also help to detect falls inminute ventilation caused by the presence ofthe bronchoscope within the endotrachealtube.77 Profound anaesthesia, including eVec-tive neuromuscular blockade, is required inpatients with head injury while undergoingbronchoscopy.

    Recommendation+ Continuous multi-modal physiological

    monitoring must be continued duringand after fibreoptic bronchoscopy. [B]

    Ventilator settingsPre-oxygenation should be achieved by in-creasing the inspired oxygen concentration to100%. 100% oxygen should be given duringbronchoscopy and in the immediate recoveryperiod. The ventilator should be adjusted to amandatory setting. Triggered modes such aspressure support or assist control will not reli-ably maintain ventilation during fibreopticbronchoscopy. The ventilator pressure limitshould be increased to ensure that adequatetidal volumes are delivered during each respira-tory cycle and the ventilator rate increased ifnecessary. Most modern microprocessor con-trolled ventilators will monitor tidal volumeand minute ventilation.

    A special swivel connector (Portex, Hythe,UK) with a perforated diaphragm, throughwhich the bronchoscope can be inserted,allows continued ventilation and maintenanceof PEEP/CPAP. This is particularly importantwhen performing a bronchoscopy in hypoxicpatients with adult respiratory distress syn-drome (ARDS).173

    Recommendation+ Care must be exercised to ensure

    adequate ventilation and oxygenation ismaintained during fibreoptic broncho-scopy via an endotracheal tube. [B]

    Endotracheal tube sizeThe internal diameter of the tracheal tube rela-tive to the external diameter of the broncho-scope is an important consideration. Broncho-scopes in the non-intubated patient occupyonly 10–15% of the cross sectional area of thetrachea. In contrast, a 5.7 mm bronchoscope

    British Thoracic Society guidelines on diagnostic flexible bronchoscopy i13

  • occupies 40% of a 9 mm endotracheal tubeand 66% of a 7 mm tracheal tube. Failure torecognise this may lead to inadequate ventila-tion of the patient and impaction of or damageto the bronchoscope. Tracheostomy tubes arealso prone to damage the bronchoscope,particularly during withdrawal when the rigidedge of the end of the tracheostomy tube canabrade the covering of the bronchoscope.Lubrication is essential to facilitate passage ofthe bronchoscope.

    Recommendation+ The internal diameter of the endotra-

    cheal tube, through which the broncho-scope is inserted, must be taken intoconsideration before bronchoscopy.

    Sedation and analgesiaThe clinical status of the patient will oftendetermine the type and level of sedation.Unstable hypoxic patients with ARDS mayrequire deep sedation, analgesia, or even mus-cle relaxation to maintain oxygenation andprevent the patient “fighting” the ventilator.162

    Synthetic narcotics such as alfentanil orfentanyl will suppress cough and provideprofound analgesia. Sedation can be inducedusing incremental doses of a benzodiazepine orpropofol. Some patients may only require lightsedation for comfort during mechanical venti-lation. In such cases bronchoscopy can beundertaken with supplemental topical anaes-thesia with lignocaine injected through thebronchoscope. Usually the dose of lignocaine issmall compared with that used for awake, non-intubated patients. However, care must beexercised in patients with combinations ofrenal failure, liver dysfunction, and congestiveheart failure when accumulation of lignocaineand seizures has been reported.172

    Recommendation+ More profound levels of sedation/

    anaesthesia can be achieved in venti-lated patients provided the clinicianperforming the procedure is acquaintedwith the use of sedative/anaestheticagents. [C]

    Cleaning/disinfectionTo ensure that all staV comply with appropriatestandards of cleaning and disinfection, writtenpolicies and procedures should be developed(see Appendix 2).

    Data collectionWhile it has always been valuable to recordflexible bronchoscopic procedures for laterdata retrieval and audit, following the Calman-Hine report174 it is becoming more important tobe able to audit lung cancer management indetail. A core data set has been recommendedfor audit of lung cancer services.175 There istherefore a need for careful recording andfeedback of information relating to flexiblebronchoscopy. A computer database providesthe ideal tool for data storage and retrieval.

    Currently only one database is commerciallyavailable in the UK, although a number of oth-ers have been developed and used in varioushospitals.176

    The use of a computerised database in bron-choscopic reporting provides the followingadvantages:(1) It is a uniform reporting tool.(2) It avoids the need for a separate letter to

    the referring doctor.(3) It may be used for internal or wider audit.(4) It may provide minimal data sets required

    for the Cancer Intelligence Unit and infor-mation for audit of waiting times.

    (5) It may provide completed report and labo-ratory request forms.

    (6) It may be used to help in the assessment ofthe competency of the bronchoscopist.

    (7) It may provide more complete data on lungcancer management including staging,type of treatment, and outcomes.

    (8) It provides a record for the trainee.It is important for any reporting system to

    have a minimum data set. Such a set should besufficiently simple to be collected with theresources available in most NHS hospitals.

    TrainingFlexible bronchoscopy is a complex andpotentially hazardous procedure requiringtrained personnel (medical, nursing, and para-medical) to minimise the risk to both patientand staV.


    The optimal number of procedures whichshould be undertaken under direct supervision(trainer in bronchoscopy unit) and indirectsupervision (trainer able to assist if called)before undertaking bronchoscopy alone willvary, depending on the competency of thetrainee and the complexity of the procedurebeing undertaken. The importance of experi-ence has been shown, for instance, to influencethe yield of positive biopsy material from visibletumours.177

    It would seem reasonable to undertake aminimum of 50 procedures under directsupervision and a further 50 under indirectsupervision, although the trainer or othercompetent bronchoscopist should be availableto give advice if needed for any traineebronchoscopist.178

    The trainee bronchoscopist should be ableto:+ Describe bronchoscopy procedures to the

    patient in appropriate terms and obtaininformed consent

    + Administer appropriate sedation to thepatient

    + Maintain an adequate airway and adequateoxygenation at all times

    + Introduce a flexible bronchoscope via thenose, mouth, tracheostomy, and endotra-cheal tube

    + Complete a flexible bronchoscopic exam-ination of the entire bronchial tree, namingall the main segmental bronchi, and direct-ing the tip of the bronchoscope into anygiven segment of the bronchial tree

    i14 BTS Bronchoscopy Guidelines Committee

  • + Carry out the following diagnostic proce-dures: bronchial brushing, bronchial lavage,endobronchial biopsy and transbronchialbiopsy

    + Manage the complications of haemorrhage,tenacious secretions, hypoxaemia, andpneumothorax if necessary.It is desirable that the trainee bronchoscopist

    should have the opportunity to see the follow-ing procedures: rigid bronchoscopy, endobron-chial tumour ablation with laser or diathermy,and stent insertion.

    These recommendations are partly based onthe training guidelines developed by theAmerican College of Chest Physicians.179


    Training on external courses has been shown toimprove practice.61 Training in the safe use ofglutaraldehyde is a legal requirement underCOSHH.

    Training should cover the following areas:+ Preoperative assessment of patients+ Communication with patients and relatives

    about the procedure and aftercare+ Appropriate use of protective clothing/

    equipment (in accordance with COSHH)+ Adequate hand washing148 149+ Appropriate patient monitoring180+ Dealing with patients from the high risk

    group+ Possible complications and how to deal with

    them (for example, bronchospasm, vaso-vagal attacks, haemorrhage, respiratory orcardiac arrests)

    + Collection and processing of specimens+ Adequate cleaning of both endoscopes and

    accessories (contaminated equipment notonly poses an infection risk but may alsolead to a false positive diagnosis133 135)

    + Appropriate rinsing of bronchoscopes afterdisinfection to remove traces of glutaralde-hyde which otherwise may cause irritation tothe respiratory tract181

    + Post-procedure observations and manage-ment of potential complications.

    Patient satisfactionPatient satisfaction appears to be influenced bymany factors—including patient characteris-tics, their previous health care experience, theirexpectations, and how well they areprepared88 182 183—and to address this satis-factorily bronchoscopy should be viewed in thecontext of its surrounding care.

    Some patients find bronchoscopy unpleas-ant, whether sedation is given or not.76 84 88 184

    There are several factors which improveacceptability by reducing the patient’sanxiety—for example, careful explanation andrelaxing tranquil music has been found to calmthe nerves of patients and also of the staV.185 186

    Planning and careful attention to detail arerequired to follow national recommenda-tions,174 to streamline the patient’s journeythrough referral, bronchoscopy, diagnosis, badnews, onward referral for surgery, radiotherapyor chemotherapy, and palliative care.187 EY-ciency in running the service is vital, withappointments and instructions being given

    clearly in advance with as few changes as possi-ble, clerical help to ensure notes and radio-graphs are available for the procedure and col-lection of data for audit, and a good liaisonbetween all the healthcare professionals in thehospital and community.

    Many patients who undergo a bronchoscopyare fearful of a malignant diagnosis and so needindividualised care which should be providedin an empathic and caring way.183 188 They andtheir relatives may have many queries, bothabout the procedure and later management,and adequate time should be available for this.A lung cancer care nurse may be used to fulfilthis role.

    The need to provide relevant and under-standable patient information before diagnos-tic and therapeutic procedures is being increas-ingly recognised, and improves the patient’stolerance of the procedure.189 190 A tendency fordoctors to explain “why” rather than also“how” may contribute to patients’ fears.83 Inmost settings the information includes verbalexplanation before, during, and after theprocedure, backed up with written informationto include date, time, place with map if neces-sary, description of the procedure, the possibleafter eVects, and the need to be collected after-wards and not drive home, even if intravenoussedation is not given.191 These instructions canbe tailor made for the hospital so that all infor-mation is correct, with inpatients also beinggiven a full explanation.192

    Recommendation+ Verbal and written patient information

    improves tolerance of the procedureand should be provided. [B]

    Studies have shown that preoperative fastingtimes have been unnecessarily long.193 It is suf-ficient to starve for 4 hours before broncho-scopy, allowing clear fluids up to 2 hoursbeforehand, depending on the listorganisation.194–196

    Recommendation+ It is suYcient for patients to have no

    food by mouth for 4 hours and to allowclear fluids by mouth up to 2 hoursbefore bronchoscopy. [B]

    The recovery time for the return of an adequategag reflex will vary between patients. Return ofthe gag reflex and the ability to swallow clearfluids safely (normally 60–90 minutes after thebronchoscopy) should be established beforedischarge from hospital.

    After the procedure the patient should begiven some idea of what the next stage of man-agement will be, an appointment for follow up,and a brief written explanation that they havehad a bronchoscopy, may cough up smallamounts of blood, and where to contact if theydevelop a significant haemoptysis or pyrexia. Itis preferable that day case patients who havebeen sedated should be accompanied homeand that higher risk patients such as the elderlyand those who have had transbronchial biop-sies should have someone to stay with them at

    British Thoracic Society guidelines on diagnostic flexible bronchoscopy i15

  • home overnight.76 A post-bronchoscopy infor-mation leaflet may be helpful to patients andtheir attendants.180

    Recommendation+ Patients who have been sedated should

    be advised not to drive, sign legallybinding documents, or operate machin-ery for 24 hours after the procedure. [C]


    Patients have a right to information about theircondition and the treatment options open tothem.197 The amount of information to be pro-vided varies198 and, in the UK, it is for the doc-tor to determine, for the particular individualpatient, how much information is disclosed.However, doctors must answer truthfully anddirectly any specific questions put by apatient,198 but may be entitled to withholdinformation if he or she reasonably believesthat to provide it would be detrimental to thehealth of the patient. Further explanation maybe given by nursing staV.199

    To proceed without consent may lead tolegal proceedings for assault and battery. To failto provide adequate information to a patient,including information about risks and adverseoutcomes, may lead to a negligence actionagainst the doctor for failure to warn. Theamount of information to be provided in theUK is a matter for the doctor’s judgement.200

    He or she should follow a course whichcommands support from a responsible body ofprofessional opinion.201 There is no duty in thiscountry for the doctor to point out remoterisks.202

    It is the responsibility of the clinician whoperforms the procedure to ensure that validconsent has been obtained.197 For a patient togive valid consent he/she must have thecapacity—that is, be capable of giving consent,be in possession of suYcient information,203

    and act entirely voluntarily.204 205

    Standards and performance of diagnostictechniquesDiagnostic yield is known to be related to theappearances at bronchoscopy in patients withsuspected lung cancer. A recent study involvingfive bronchoscopy centres178 showed an 82%positive rate for malignancy with bronchialbiopsies in cases where malignancy was judgedto be present on appearances through thebronchoscope. This compares with 76%206 and77%207 in other studies. The combined use ofbronchial washings and brushings increasedthe diagnostic rates to 87%.178

    For the maximum diagnostic yield in bothbronchoscopically visible tumours and tu-mours with normal or non-specific appear-ances, a combination of biopsy material,brushings, and washings should be obtained,206

    although the number of tumours diagnosed bywashings alone is relatively small.206 Therelative yield from washings and brushingscompared with biopsies appears to be higher incases where the tumour is not definitely

    visible.178 The optimum sequence of obtainingthese three types of samples is uncertain andrequires further studies.

    The number of bronchial biopsy specimensrequired to obtain a high diagnostic rate hasbeen studied. In order to achieve at least a 90%probability of a positive malignant biopsy, atleast five samples were required in cases wherethere was a visible tumour.207 Another studyhas suggested that a minimum of four biopsyspecimens are required.208 It would seemreasonable to expect a bronchoscopist toachieve a diagnosis in at least 80% of caseswhere a tumour is visible at bronchoscopyusing the combined methods of biopsies,brushings, and washings.209 The diagnosticyield will also be aVected by the quality of thepathology services. More data are required todetermine minimum standards at variousstages of higher professional training.

    Transbronchial needle aspiration (TBNA)has recently been used more widely. For visibletumours the yield from TBNA and forcepsbiopsy is similar.210 However, TBNA may bemore sensitive than forceps biopsy in carcino-mas which are infiltrating the submucosa andperibronchial areas.211 TBNA can be used tosample the hilar-mediastinal lymph glands.Obviously the node from which the biopsymaterial is to be taken should be closelyadjacent to the airway. The sensitivity of TBNAfor node sampling varies widely in the literaturebut is higher (38% of cases) where there isradiological confirmation of lymph glandenlargement.212 One study has shown improvedsensitivity when using a larger (22 gauge)needle.213

    For peripheral pulmonary lesions transbron-chial biopsy has a sensitivity that varies accord-ing to the number of biopsy specimens takenand also the underlying disease. Fluoroscopyappears to be unnecessary in patients with dif-fuse lung disease46 but is advisable for localisedlesions. The type of forceps used does not seemto influence the diagnostic yield.214 About onethird of attempted transbronchial biopsy sam-ples consisted of bronchial rather than alveolartissue in one study.215 Diagnostic yields are highin patients with stages II or III sarcoidosis(about 75%) and in those with lymphangitiscarcinomatosis (66%), moderate in patientswith stage I sarcoidosis (58%), but lower inthose with interstitial pulmonary fibrosis(27%).215 Transbronchial biopsy is not usefulfor the diagnosis or staging of cryptogenicfibrosing alveolitis and the advantages of alter-native biopsy techniques, including open lungbiopsy or video-assisted thoracoscopic lungbiopsy, are detailed elsewhere.46 For diVuselung diseases 4–6 transbronchial biopsy sam-ples should be obtained from one lung.46 215 Incases of suspected sarcoidosis endobronchialbiopsy samples should also be taken.46 Forlocalised lung lesions 7–8 transbronchial bi-opsy samples have been proposed.215 In allpatients the clinician will need to assess thelikely balance between the risk of complica-tions and the diagnostic yield, and also thepossibility of alternative diagnostic approaches.For localised peripheral lesions a transthoracic

    i16 BTS Bronchoscopy Guidelines Committee

  • needle biopsy under imaging control may oVeradvantages over a transbronchial biopsy.

    Comprehensive guidelines are available inthe literature regarding the performance ofbronchoalveolar lavage and interpretation ofthe results.216 217

    Recommendations+ At least five bronchial biopsy specimens

    should be taken in cases of suspectedbronchial malignancy. [B]

    + Biopsies, brushings and washingsshould all be obtained in cases ofsuspected endobronchial malignancy.[B]

    + A minimum diagnostic level of at least80% should be obtained from a combi-nation of biopsies, brushings, and wash-ings in cases of endoscopically visiblemalignancy. [B]

    + When taking transbronchial lung bi-opsy specimens in patients with diVuselung disease, an attempt should bemade to obtain 4–6 samples from onelung. [B]

    Appendix 1: Details of the BTS FlexibleBronchoscopy Guidelines Working PartyJohn Babb, scientist with a special interest in disinfec-tion and decontamination, Hospital Infection ResearchLaboratory, City Hospital NHS Trust, Dudley Road,Birmingham. Member of the Microbiology AdvisoryCommittee of the Medical Devices Agency of theDepartment of Health and member of the EndoscopyGuideline Working Party of the British Society of Gas-troenterology.

    Pippa Bowie, research assistant in the ClinicalEVectiveness Unit at Kettering General Hospital,Northamptonshire, and MSc student who carried outthe literature search while a student in the Departmentof Evidence Based Medicine at the University ofOxford.

    Adrienne Brewin, respiratory nurse specialist at theKent and Sussex Hospital, Tunbridge Wells. Nomineeof the Royal College of Nursing.

    Adam Fraise, consultant in medical microbiology atthe City Hospital NHS Trust, Dudley Road, Birming-ham. Member of the Microbiology Advisory Committeeof the Medical Devices Agency of the Department ofHealth.

    Christopher Garrard, consultant physician anddirector of the Intensive Care Unit at the John RadcliVeHospital, Oxford.

    John Harvey, consultant physician at SouthmeadHospital, Bristol.

    David Honeybourne (Chairman), consultantphysician at Birmingham Heartlands Hospital.

    Richard Lewis, consultant physician at WorcesterRoyal Infirmary.

    Christiane Neumann, clinical nurse specialist inendoscopy at the City Hospital NHS Trust, DudleyRoad, Birmingham. General secretary of the EuropeanSociety of Gastroenterology and Endoscopy Nurses andAssociates.

    Christopher G Wathen, consultant physician atWycombe Hospital, High Wycombe, Buckinghamshire.

    Tim Williams, consultant physician at KetteringGeneral Hospital, Northamptonshire.

    Appendix 2: Cleaning and disinfectionprocedureBronchoscopes are a potential source of infection andmust be cleaned, disinfected, and rinsed before use,between patients, and at the end of a list. This requiresspecial training and knowledge. Always refer to themanufacturers’ instructions for specific instrumentadvice.

    For a detailed description of endoscope cleaning anddisinfection procedures see the Medical Devices Agencydevice bulletin 9607 “Decontamination of endoscopes”.117

    This section summarises the decontamination proce-dures for bronchoscopes. A record should be kept ofwhich bronchoscope is used on an individual patientand also of the decontamination procedure.121

    Immersible instruments should be leak tested toensure instruments are water tight before processing.


    Wear suitable protective clothing (gloves, plastic apronsand eye protection) if splashing is likely (see section onStaV safety).1. Wipe clean external surfaces with detergent.2. Remove and dismantle the suction valve for separate

    cleaning.3. Clean the channel and port with a suitable cleaning

    brush.4. Flush the channel through with detergent followed

    by air.


    1. Use automated systems whenever possible as theyprotect the user from hazardous processing chemi-cals. (NB: If an automated system is used, ensurethat machine disinfection is carried out at the start ofeach session/day. This includes filters and pipeworkfor rinse water.)

    2. Immerse the bronchoscope in glutaraldehyde (seetable 2 for contact times) or an alternative disinfect-ant (for example, peracetic acid, chlorine dioxide,superoxidised water, or ethylene oxide), ensuringthat the channel is irrigated and all air is expressed.

    3. Change the disinfectant when the manufacturer’srecommended use life is reached.

    4. Rinse the instrument with sterile or filtered tap water(0.2 µm).

    5. Change the detergent and rinse water regularly (ide-ally after each cycle) if reused.

    6. Dry the instrument with air and/or alcohol. A 70%alcohol rinse may be used for the channel if the qual-ity of the rinse water cannot be assured. Wipe dry theinsertion tube.

    7. Accessories should be cleaned, preferably usingultrasonics, and, wherever possible, sterilised byautoclaving.


    Alternatives to glutaraldehyde include peracetic acid,chlorine dioxide, superoxidised water, and ethyleneoxide.113 117

    For advice on the safe use of glutaraldehyde andother processing chemicals see section on “Cleaningand disinfection” in these guidelines.

    Table 2 Glutaraldehyde* disinfectant contact times

    Patient category

    When Low riskHigh risk (e.g. tuberculosis (TB), immunosuppressed,symptomatic HIV)

    Start of list 20 min 20 minBetween patients 20 min 1 hour after TB and before and after immunosuppressed

    patients†End of list 20 min 1 hour after TB and after immunosuppressed patients†

    *2% glutaraldehyde at room temperature.†Includes symptomatic HIV patients.Twenty minutes disinfection with 2% glutaraldehyde is suYcient to kill most non-sporing bacte-ria, including Mycobacterium tuberculosis, and blood-borne viruses. Longer contact times of 1 hourare recommended for some other mycobacteria such as Mycobacterium avium intracellulare,encysted parasites, and spores. Thorough cleaning will remove most problematic and pathogenicmicroorganisms.If an alternative disinfectant is used, advice should be sought from the disinfectant manufacturersand your Infection Control Team/Committee.

    British Thoracic Society guidelines on diagnostic flexible bronchoscopy i17


    Advice is available on the selection and use ofendoscope washer disinfectors.117 218 219 Automated sys-tems are preferred as they reduce the likelihood ofexposure to hazardous processing chemicals and aremore easily validated than manual processes. It is essen-tial that the machine selected eVectively cleans,disinfects, and rinses all channels and external surfaceswithout damaging the instrument. If glutaraldehyde orirritant processing chemicals are used, fumes must becontained or removed to protect processing staV. Sterileor bacteria free water (0.2 µm filter) must be used forrinsing bronchoscopes. Machines should be equippedwith a “self disinfect” facility which includes filters andall rinse water pathways. Water filters should be changedregularly in accordance with the manufacturers’ instruc-tions and more frequently if water quality is poor. Lowwater quality should be suspected if filters block, whichmay then lead to inadequate water pressure in themachine. Consideration should be given to flushing thedisinfected bronchoscope with sterile water for micro-biological analysis in order to monitor the adequacy ofthe disinfection process.

    1 Woodcock A, Campbell I, Collins JVC, et al. Bronchoscopyand infection control. Lancet 1989;ii:270–1.

    2 Harrison BDW. Guidelines for care during bronchoscopy:British Thoracic Society. Thorax 1993;48:584.

    3 Petrie J, Barnwell B, Grimshaw J, on behalf of the ScottishIntercollegiate Guidelines Network. Clinical guidelines: crite-ria for appraisal for national use. Edinburgh: Royal Collegeof Physicians, 1995.

    4 Zavala DC. Diagnostic fiberoptic bronchoscopy: techniquesand results of biopsy in 600 patients. Chest 1975;68:12–9.

    5 Credle W, Smiddy J, Elliott R. Complications of fiberopticbronchoscopy. Am Rev Respir Dis 1974;109:67–72.

    6 Suratt P, Smiddy J, Gruber B. Deaths and complicationsassociated with fiberoptic bronchoscopy. Chest 1976;69:747–51.

    7 Simpson FG, Arnold AG, Purvis A, et al. Postal survey ofbronchoscopic practice by physicians in the UnitedKingdom. Thorax 1986;41:311–7.

    8 Dreison RB, Albert RK, Talley PA, et al. Flexible fibreopticbronchoscopy in the teaching hospital. Chest 1978;74:144–9.

    9 Pereira W, Kovnat D, Snider G. A prospective cooperativestudy of complications following flexible fiberoptic bron-choscopy. Chest 1978;73:813–6.

    10 Pue C, Pacht E. Complications of fiberoptic bronchoscopyat a university hospital. Chest 1995;107:430–2.

    11 Lukomsky G, Ovchinnikov A, Bilal A. Complications ofbronchoscopy. Comparison of rigid bronchoscopy undergeneral anesthesia and flexible fiberoptic bronchoscopyunder topical anesthesia. Chest 1981;79:316–21.

    12 Zavala DC. Complications following fiberoptic broncho-scopy. Chest 1978;73:783–5.

    13 Hanson RR, Zavala DC, Rhodes ML, et al. Transbronchialbiopsy via flexible fibroptic bronchoscope: results in 164patients. Am Rev Respir Dis 1976;114:67–72.

    14 Zavala DC. Pulmonary haemorrhage in fibreoptic trans-bronchial biopsy. Chest 1976;70:584–8.

    15 Sutherland A, Santamaria J, Nana A. Patient comfort andplasma lignocaine concentrations during fibreoptic bron-choscopy. Anaesth Intensive Care 1985;13:370–4.

    16 Smith M, Dhillon D, Hayler A, et al. Is fibreoptic broncho-scopy in patients with lung cancer and hepatic metastasespotentially dangerous? Br J Dis Chest 1985;79:368–73.

    17 Milman N, Laub M, Munch E, et al. Serum concentrationsof lignocaine and its metabolite monoethylglycinexylidideduring fibreoptic bronchoscopy in local anaesthesia. RespirMed 1998;92:40–3.

    18 Langmarc EL, Martin RJ, Pak J, et al. Serum lidocaine con-centrations in asthmatics undergoing research broncho-scopy. Chest 2000;117:1055–60.

    19 Albertini R, Harrel JH, Moser KM. Hypoxemia duringfiberoptic bronchoscopy. Chest 1974;65:117–22.

    20 Matsushima Y, Jones RL, King EG, et al. Alterations in pul-monary mechanics and gas exchange during routinefiberoptic bronchoscopy. Chest 1984;86:184–8.

    21 Pirozynski M, Sliwinski P, Radwan L, et al. Bronchoalveolarlavage: comparison of three commonly used procedures.Respiration 1991;58:72–6.

    22 Milman N, Faurschou P, Grode G, et al. Pulse oximetryduring fibreoptic bronchoscopy in local anaesthesia:frequency of hypoxaemia and eVect of oxygen supplemen-tation. Respiration 1994;61:342–7.

    23 Kristensen M, Milman N, Jarnvig I. Pulse oximetry at fibre-optic bronchoscopy in local anaesthesia: indication forpostbronchoscopy oxygen supplementation? Respir Med1998;92:432–7.

    24 Evans E, Ganeshalingam K, Ebden P. Changes in oxygensaturation and transcutaneous carbon dioxide and oxygenlevels in patients undergoing fibreoptic bronchoscopy.Respir Med 1998;92:739–42.

    25 Antonelli M, Conti G, Riccioni L, et al. Noninvasivepositive-pressure ventilation via face mask during bron-

    choscopy with BAL in high-risk hypoxemic patients. Chest1996;110:724–8.

    26 Luck J, Messeder O, Rubenstein M, et al. Arrhythmias fromfiberoptic bronchoscopy. Chest 1978;74:139–43.

    27 Shrader D, Lakshminarayan S. The eVect of fiberopticbronchoscopy on cardiac rhythm. Chest 1978;73:821–4.

    28 Breuer H, Charchut S, Worth H. EVects of diagnostic pro-cedures during fiberoptic bronchoscopy on heart rate,blood pressure, and blood gases. Klin Wochenschr 1989;67:524–9.

    29 Matot I, Kramer M, Glantz L, et al. Myocardial ischemia insedated patients undergoing fiberoptic bronchoscopy. Chest1997;112:1454–8.

    30 Moisan T, Chandrasekhar A, Moran J. Arrhythmias inpatients with heart disease during fiberoptic bronchoscopy.Illinois Med J 1979;7:27–30.

    31 Katz A, Michelson E, Stawicki J, et al. Cardiac arrhythmias.Frequency during fiberoptic bronchoscopy and correlationwith hypoxemia. Arch Intern Med 1981;141:603–6.

    32 Cordasco EM, Mehta AC, Ahmad M. Bronchoscopyinduced bleeding: a summary of nine years ClevelandClinic experience and review of the literature. Chest 1991;100:1141–7.

    33 Mitchell D, Emerson C, Collins J, et al. Transbronchial lungbiopsy with the fibreoptic bronchoscope: analysis of resultsin 433 patients. Br J Dis Chest 1981;75:258–62.

    34 Hue S. Complications in transbronchial lung biopsy. KoreanJ Int Med 1987;2:209–13.

    35 Loube D, Johnson J, Wiener D, et al. The eVect of forcepssize on the adequacy of specimens obtained by transbron-chial biopsy. Respir Dis 1993;148:1411–3.

    36 O’Brien JD, Ettinger NA, Shevlin D, et al. Safety and yieldof transbronchial biopsy in mechanically ventilated pa-tients. Crit Care Med 1997;25:440–6.

    37 Borchers S, Beamis J. Flexible bronchoscopy. Chest SurgClin North Am 1996;6:169–92.

    38 Papin T, Lynch J, Weg J. Transbronchial biopsy in thethrombocytopenic patient. Chest 1985;88:549–52.

    39 Kozak E, Brath L. Do “screening” coagulation tests predictbleeding in patients undergoing fiberoptic bronchoscopywith biopsy? Chest 1994;106:703–5.

    40 Brickey DA, Lawlor DP. Transbronchial biopsy in the pres-ence of profound elevation of the international normalizedratio (INR). Chest 1999;115:1667–71.

    41 British Committee for Standards in Haematology. Guide-lines on oral anticoagulation: third edition. Br J Haematol1998;101:374–87.

    42 Milman N, Munch E, Faurschou P, et al. Fiberopticbronchoscopy in local anaesthesia. Indications, results andcomplications in 1323 examinations. Acta Endosc 1993;23:151–62.

    43 Joyner L, Scheinhorn D. Transbronchial forceps lung biopsythrough the fiberoptic bronchoscope. Diagnosis of diVusepulmonary disease. Chest 1975;67:532–5.

    44 De Fenoyl O, Capron F, Lebeau B, et al. Transbronchialbiopsy without fluoroscopy: a five year experience inoutpatients. Thorax 1989;44:956–9.

    45 Puar H, Young R, Armstrong E. Bronchial and transbron-chial lung biopsy without fluoroscopy in sarcoidosis. Chest1985;87:303–6.

    46 British Thoracic Society. The diagnosis, assessment andtreatment of diVuse parenchymal lung disease in adults.Thorax 1999;54(Suppl 1): S1–30.

    47 Ahmad M, Livingston D, Golish J, et al. The safety ofoutpatient transbronchial biospy. Chest 1986;90:403–5.

    48 Levy H, Kallenbach JM, Feldman C, et al. Delayedpneumothorax after transbronchial lung biopsy. Thorax1986;41:647–8.

    49 De Fijter J, van der Hoeven J, Eggelmeijer F, et al. Sepsissyndrome and death after bronchoalveolar lavage. Chest1993;104:1296–7.

    50 Rennard SI, Aalbers R, Bleeker E, et al. Bronchoalveolarlavage: performance, sampling procedure, processing andassessment. Eur Respir J 1998;26(Suppl):13–15S.

    51 Krause A, Hohberg B, Heine F, et al. Cytokines derivedfrom alveolar macrophages induce fever after broncho-scopy and bronchoalveolar lavage. Am J Respir Crit CareMed 1997;155:1793–7.

    52 Witte M, Opal S, Gilbert J, et al. Incidence of fever and bac-teremia following transbronchial needle aspiration. Chest1986;89:85–7.

    53 Vasanthakumar V, Bhan G, Perera B, et al. A study to assessthe eYcacy of chemoprophylaxis in the prevention ofendoscopy-related bacteraemia in patients aged 60 andover. Q J Med: New Series 1990;278:647–53.

    54 Haynes J, Greenstone M. Fibreoptic bronchoscopy and theuse of antibiotic prophylaxis. BMJ 1987;294:1199.

    55 Yigla M, Oren I, Bentur L, et al. Incidence of bacteraemiafollowing fibreoptic bronchoscopy. Eur Respir J 1999;14:789–91.

    56 Durack DT. Prevention of infective endocarditis. N Engl JMed 1995;332:38–44.

    57 Dajani A, Bisno A, Chung KD, et al. Prevention of bacterialendocarditis: recommendations by the American HeartAssociation. JAMA 1990;264:2919–22.

    58 Aelony Y, Finegold S. Serious infectious complications afterflexible fibreoptic bronchoscopy. West J Med 1979;131:327–33.

    59 British National Formulary. London: BMA, 1999.60 Simmonds NA. Recommendations for endocarditis prophy-

    laxis. The Endocarditis Working Party for AntimicrobialChemotherapy. J Antimicrob Chemother 1993;31:437–8.

    i18 BTS Bronchoscopy Guidelines Committee

  • 61 Honeybourne D, Neuman C. An audit of bronchoscopypractice in the United Kingdom: a survey of adherence tonational guidelines. Thorax 1997;52:709–13.

    62 British Society of Gastroenterology. Report of a workingparty on the staYng of endoscopy units. Gut 1987;28:1682–5.

    63 Davies L, Misre R, Spence PMA, et al. Cardiovascular con-sequences of fibreoptic bronchoscopy. Eur Respir J1997;10:695–8.

    64 Dweik R, Mehta A, Meeker D, et al. Analysis of the safety ofbronchoscopy after recent acute myocardial infarction.Chest 1996;110:825–8.

    65 Dunagan DP, Burke HL, Aquino SL, et al. Fibreoptic bron-choscopy in coronary care unit patients. Chest 1998;114:1660–7.

    66 Bein T, Pfeifer M. Fiberoptic bronchoscopy after recentacute myocardial infarction. Stress for the heart? Chest1997;112:295.

    67 Kvale P. Is it really safe to perform bronchoscopy after arecent acute myocardial infarct? Chest 1996;110:591–2.

    68 American College of Cardiology and American HeartAssociation Taskforce. Guidelines for peri-operativecardiovascular evaluation for non cardiac surgery. Circula-tion 1996;93:1278–317.

    69 Mavritsyn LE, Lifshits NA. Complications in the fiberbronchoscopy of bronchial asthma patients. KlinicheskaiaMeditsina 1980;58:37–40.

    70 Fish JE, Peterman VI. EVects of inhaled lidocaine on airwayfunction in asthmatic subjects. Respiration 1979;37:201–7.

    71 Djukanovic R, Wilson J, Lai C, et al. The safety aspects offiberoptic bronchoscopy, bronchoalveolar lavage, andendobronchial biopsy in asthma. Am Rev Respir Dis1991;143:772–7.

    72 Van Vyve T, Chanez P, Bousquet J, et al. Safety of broncho-alveolar lavage and bronchial biopsies in patients withasthma of variable severity. Am Rev Respir Dis 1992;146:116–21.

    73 Rankin J, Snyder P, Schachter E, et al. Bronchoalveolar lav-age. Its safety in subjects with mild asthma. Chest 1984;85:723–8.

    74 Peacock M, Johnson J, Blanton H. Complications of flexiblebronchoscopy in patients with severe obstructive pulmo-nary disease. J Bronchol 1994;1:181–6.

    75 Knox A, Mascie-Taylor B, Page R. Fibreoptic bronchoscopyin the elderly: 4 year’s experience. Br J Dis Chest 1988;82:290–5.

    76 Macfarlane JT, Storr A, Wart MJ, et al. Safety, usefulnessand acceptability of fibreoptic bronchoscopy in the elderly.Age Ageing 1981;10:127–31.

    77 Peerless J, Snow N, Likavec M, et al. The eVect of fiberopticbronchoscopy on cerebral hemodynamics in patients withsevere head injury. Chest 1995;108:962–5.

    78 Bajwa M, Henein S, Kamholz S. Fiberoptic bronchoscopyin the presence of space-occupying intracranial lesions.Chest 1993;104:101–3.

    79 Gong H, Salvatierra C. Clinical eYcacy of early and delayedfiberoptic bronchoscopy in patients with hemoptysis. AmRev Respir Dis 1981;124:221–5.

    80 Poe R, Israel R, Marin M, et al. Utility of fiberopticbronchoscopy in patients with hemoptysis and a nonlocal-izing chest roentgenogram. Chest 1988;92:70–5.

    81 Maguire GP, Rubinfield AR, Trembath PW, et al. Patientsprefer sedation for fibreoptic bronchoscopy. Respirology1998;3:81–5.

    82 Putinati S, Ballerin L, Corbetta L, et al. Patient satisfactionwith conscious sedation for bronchoscopy. Chest 1999;115:1437–40.

    83 Poi PJ, Chuah SY, Srinivas P, et al. Common fears ofpatients undergoing bronchoscopy. Eur Respir J 1998;11:1147–9.

    84 Pearce S. Fibreoptic bronchoscopy: is sed