INFECTION

CONTROL

IN

ENDOSCOPY

Acknowledgements

Prepared and edited by: Alistair Cowen, M.D.

With Trudy Rayner, Bronwyn King and Dianne Jones

Contributors:

Gastroenterological Society of AustraliaAlistair Cowen MD FRACP, Chairman, Endoscopy Committee,Gastroenterological Society of Australia

Antony Speer BE. MB. BS. FRACP Gastroenterologist, RoyalMelbourne Hospital

Gastroenterological Nurses Society of AustraliaDianne Jones RN, RM, BApp Sc, FRCNA, ACGEN, Honorary SecretaryGENSA.

Bronwyn King RN, ACGEN, MRCNA, President GENSATrudy Rayner RN, ACGEN, President Qld GENSA.

Thoracic Society of Australia and New Zealand

David Fielding, MBBS, FRACP, MD, Respiratory Physician, Princess AlexandraHospital, Brisbane

Infectious Diseases Physician and Microbiologist

Peter Collignon MB.BS, B.Sc.,MASM, FRACP, FRCPA, FASM,Infectious Diseases Physician and Microbiologist, Canberra Hospital,Canberra

Michael Whitby MB BS, DTM & H, MPH, FRACGP, FRACP, FRCPA, FRC Path,FACSHP,FAFPHM, Director of Infectious Diseases, Infection Control & SexualHealth, Princess Alexandra Hospital, Brisbane

Published by: Gastroenterological Society of Australia145 Macquarie Street SYDNEY 2000 (1999)

© Copyright. Gastroenterological Society of Australia andGastroenterological Nurses Society of Australia

This booklet is copyright and all rights reserved. It may not be reproduced in whole or in part without permissionfrom the Gastroenterological Society of Australia, Sydney, and the Gastroenterological Nurses Society ofAustralia.

1.

Table of Contents

Introduction 3

Sterilisation and Disinfection 71. Sterilisation 72. Disinfection 7

Mechanisms of infection and major risk factors 101. Mechanisms of infection 102. Risk Factors 10

The infecting organisms 111. Bacteria 112. Viruses 133. Other infections 15

The Procedures 171. Diagnostic Endoscopy 172. Oesophageal dilatation 173. Endoscopic sclerotherapy /banding 174. Colonoscopy 185. E.R.C.P 186. Percutaneous Endoscopic Gastrostomy 197. Endoscopic Ultrasound 20

The Patient with increased susceptibility to infection 211.Compromised immune status 212.Procedurally induced tissue trauma 213.Intrinsic sources of infection 224.Increased risk of bacterial lodgement following septicaemia 22

Endoscopy and Antibiotic Prophylaxis 231.Indications 232.Antibiotic regimens 233.Antibiotic prophylaxis for E.R.C.P. 244.Bronchoscopy and antibiotic prophylaxis 25

Principles of effective decontamination protocols 261. Introduction 262. Effectiveness of Recommended Protocols 273. Endoscope structure 274. Cleaning equipment 295. Cleaning fluids 296. Rinsing water 307. Rinsing 318. Disinfectant 319. General maintenance 3210. Lubrication 3211. Work areas 33

2.

Decontamination regimens 341. Manual cleaning 342. Manual disinfection 353. End of list processing 364. Endoscopic accessory equipment 365. Reprocessing Flexible Bronchoscopes and Accessories 386. Variation in cleaning and disinfection regimens depending 40 upon the supposed infective status of the patient7.Single Use Only 41

Endoscopic disinfecting machines 43Machine Design and Principles 44

Proof of Process 45

Investigation of possible viral transmission 51

Microbiological testing of endoscopes 531.Introduction 532.Testing Schedules 533. Recommendation 554. Protocols 555. Interpretation of Cultures 566. Microbiological Surveillance of Automatic Disinfectors 57

Occupational Health and Safety in the Endoscopy Workplace 581. Introduction 582. Chemical and Hazardous Substance Exposure 583. Infectious Diseases 604. Other Health and Safety Issues 61

References 63

3.

Introduction

Infection Control in Endoscopy is in effect the 4th Edition of Infection and Endoscopy. TheGastroenterological Society of Australia and the Gastroenterological Nurses Society of Australiabelieve that this new monograph has benefited in a major way from the contributions of MichaelWhitby and Peter Collignon. The Australian Society for Infectious Diseases, the AustralianSociety for Microbiology and the Royal Australasian College of Pathologists were consulted forinput into the monograph and nominated Drs Whitby and Collignon. David Fieldingrepresenting the Thoracic Society of Australia and New Zealand has advised in the areas ofmycobacterial infection and bronchoscope reprocessing.

We make no apology that the importance of cleaning will be emphasised again and againthrough this monograph. We believe that there is a very real risk that prolonged chemicalimmersion times may be counterproductive since it is less easy to quantitate cleaning, andprolonging immersion times in busy endoscopy units may result in inappropriate pressure tohasten cleaning procedures. There is universal agreement by all societies that cleaning is themost important aspect of any cleaning and disinfection regimen, and it is totally unacceptable toshortcut any of the cleaning steps described in this edition. It needs to be emphasised thatsimply increasing the immersion time of endoscopes in chemical disinfectants is NOT theanswer to the real and extremely complex problems of infections associated with endoscopicprocedures.

A wide variety of chemical disinfectants are regularly appearing on the market. Some, such asthe Steris system, have adequate research and documentation of efficacy, but many othershave little or no proven efficacy. It cannot be stressed too strongly that the ability to kill a varietyof microbiological agents under laboratory conditions is only one part of demonstrating theeffectiveness of a chemical disinfectant. The Committee strongly believes that proven efficacyunder operating conditions, compatibility with instruments, and safety for staff must also bedemonstrated.

The modern endoscope is the clinical microbiologist's nightmare. It is built of material unable towithstand heat and many chemicals. It is honeycombed with multiple small channels, some withblind endings, none of which can be adequately inspected following cleaning. It is physicallydelicate, difficult to dry and difficult to adequately sample microbiologically.

Microbiological monitoring of endoscopes is fraught with controversy. Nevertheless, we havecontinued to recommend such assessment, while recognising that our protocols are empiric innature. This, however, should be of little import as long as it is recognised that such proceduresare utilised as an indirect marker of both adequacy and reliability of the cleaning/disinfectionprocess and of structural integrity of the instrument. In most circumstances, it is an assessmentof the number of organisms present after routine cleaning that is important and not the specifictaxonomy of residual microbes.

Despite all of these difficulties, clinical infections due to endoscopic procedures have been veryuncommon. With careful manual cleaning and adequate disinfection, endoscopically inducedinfections can and should be extremely rare.

4.

The three most important rules of any effectivecleaning and disinfection schedule are:

CLEAN IT

CLEAN IT

CLEAN IT

5.

NON IMMERSIBLE ENDOSCOPESSHOULD BE PHASED OUT

Scrupulous manual cleaning has been shown toremove all H.I.V. infective material fromendoscopes used in A.I.D.S. patients or

artificially contaminated with aqueous H.I.V. viralsuspensions.

Patients have died from Serratia Marcescensinfection acquired from a bronchoscope

inadequately cleaned and then Ethylene oxide"sterilised".

In several endoscopy-associated serial clinicalinfections, increasing chemical immersion timehas NOT solved the problem. Infections haveonly ended when inadequate manual cleaning(e.g. failure to ultrasonically clean spiral-wound

wire forceps, failure to adequately cleanspring-loaded biopsy port valves) has been

recognised and corrected.

Aldehyde chemical disinfectants and alcohol"fix" proteinaceous material, making it harderto remove and preventing chemical contact

with the organism.

6.

Dried biological material is much harder toremove and chemical disinfectants, including

aldehydes, penetrate much more slowly.

Organic material may inactivate the majorityof disinfectants including aldehydes,

resulting in a rapid reduction of effectivechemical concentration.

Immersion of a properly cleaned endoscope in2% Glutaraldehyde for 10 minutes will render

it safe for immediate clinical use.

7.

Sterilisation and Disinfection

1. STERILISATION

Sterilisation is a term describing the use of a physical or chemical procedure to destroyall microbiological life including bacterial spores. Major sterilising processes include dryheat sterilisation, steam sterilisation under pressure, low temperature hydrogen peroxideplasma sterilisation, automated peracetic acid systems and ethylene oxide gas1,2 Anumber of chemical germicides are capable of achieving sterilisation if used forprolonged periods. To achieve sterilisation with aldehyde based products, a contact timeexceeding three hours is required. At present modern flexible endoscopes cannot beregularly sterilised, either because processes such as heat and steam are incompatiblewith the materials of which they are composed or because processes such as ethyleneoxide and extremely prolonged chemical immersion are impractical and unlikely toachieve full sterilisation for the reasons subsequently outlined.

2. DISINFECTION

Disinfection is different from sterilisation. Disinfection is a process that only removes orkills organisms that are regarded likely to cause disease. Many organisms are relativelyresistant to disinfection. In general they are regarded as low virulence organisms, e.g.bacterial spores. Other forms of microbial structures designed to allow survival in hostileenvironments, e.g. protozoal cysts, are also resistant.

Any item that comes into contact with sterile body sites needs to be sterile. Sterilisationis also preferable for instruments that come in contact with an intact mucous membrane,but unfortunately because of the structure of many instruments (including endoscopes),this is not achievable either because the instrument cannot withstand heat or theimpracticable logistics of using other sterilisation processes (e.g. gas sterilisation).

Disinfection can be achieved by a number of means that include heat and chemicals.The cleaning process itself is a very efficient means of achieving disinfection. Cleaningremoves or destroys more organisms than a chemical disinfectant is likely to do over asimilar period of time (e.g. a 5 minute contact time). Organic material binds andinactivates many chemical disinfectants. Some disinfectants such as glutaraldehyde andalcohol fix protein. Thus chemical disinfectants may create a physical barrier ofdenatured protein that can protect organisms coated by organic material. Obviously noagent can be effective against micro-organisms it cannot reach. An advantage of heat asa disinfecting agent is that it is conducted and is able to penetrate better than chemicals.The action of heat will also be compromised by inadequate cleaning, but to a lesserextent than with chemical disinfectants. With high levels of wet heat and pressure(autoclaving) sterilisation is achieved. When heat is used at lower temperatures, e.g.boiling water or pasteurisation (60 - 80°C), heat is a very effective disinfectant.

For instruments that come in contact with mucosal surfaces, a high level disinfectant isrequired. Disinfecting agents need to kill all forms of bacteria (gram positive, gramnegative and mycobacteria), viruses (both the more sensitive lipid coated viruses suchas HIV and relatively resistant viruses such as the polio virus), fungi (e.g. Candida) andprotozoa (e.g. Giardia). High level disinfectants are able to kill the more resistant formsof microbial life such as bacterial spores and cysts but only with prolonged contact times(usually over 3 hours).

No sterilising or disinfection agent works instantaneously. They all require sufficientcontact times. The ability to achieve complete killing of microorganisms is dependent ona number of factors.

8.

1. Initial number of organisms present.

This is a critical factor as there is a log kill with time. Therefore the higher the number oforganisms present, the longer it will take to achieve a complete kill. This is a furtherreason why cleaning is a critical step in any cleaning disinfection protocol. A log fivereduction or more in the number of organisms present can certainly be achieved byscrupulous cleaning.

2. Temperature

In general the higher the temperature, the quicker the disinfecting agent will destroyorganisms. Most of the invitro testing of chemical disinfectants has been done at 20°C.Therefore if room temperatures lower than 20°C are used, the activity of a disinfectantmay be reduced and a longer than currently recommended soaking time is required. It isattractive to suggest that increasing the temperature of a disinfectant will reduce thesoaking time. However, this is a complex issue. Some aldehyde products (alkalineglutaraldehyde) cannot be heated to any significant degree without destroying thecompound. Preliminary results suggest that buffered aldehyde solutions (e.g. AidalPlus) can be heated and may be capable of much more rapid inactivation of difficultorganisms such as Mycobacterium tuberculosis and atypical mycobacteria. A number orstudies are presently underway to develop glutaraldehyde based automatic disinfectorswhere the temperature of the glutaraldehyde is raised significantly. A major difficultywith raising the temperature of glutaraldehyde is the increased vapour production, whichnot only raises Occupational Health and Safety issues, but also may mean the life of theglutaraldehyde solution will be significantly shortened.

3. Concentration

Concentration of a chemical disinfectant is critical. In general the lower theconcentration of the agent, the longer it will take to kill the same number of organisms. Itis particularly important to ensure that disinfectants do not become diluted with excesswater remaining on endoscopes after rinsing. Concentration of an agent (e.g. 2%glutaraldehyde) may be more than halved with repeated use and the activity of thedisinfection process significantly compromised. Presently there is considerable difficultyin accurately measuring the concentration of glutaraldehyde solutions.

There is no specific soaking time that will guarantee that all agents present are killed bychemical disinfectants. It is dependent on the number of organisms present, thepresence of inactivating compounds (e.g. organic materials), the pH, the temperature,the concentration of a disinfectant and the relative resistance (and therefore kill rate) ofthe organism involved. Recommendations given (e.g. 10 minutes in 2% glutaraldehyde)are for an adequately cleaned endoscope. If cleaning is compromised, even prolongedcontact time (in excess of 60 minutes) is unlikely to kill pathogenic organisms present onor in the endoscope. It has been shown that ten separate full disinfection cycles failed tokill Mycobacterium tuberculosis present in an inadequately cleaned bronchoscope3.

Agents, that can achieve high level disinfection, include 2% glutaraldehyde, peraceticacid, high concentrations of hydrogen peroxide and some chlorine releasing agents. Ingeneral peracetic acid and high concentrations of hydrogen peroxide can only be used inautomated processors which prevent staff exposure. Glutaraldehyde can be used ineither manual processing or in automated processors. Ethylene oxide achievessterilisation with prolonged contact time. However, it must be recognised that gassterilisation with ethylene oxide is subject to the same limitations as liquid chemicaldisinfectants. Gas sterilisation cannot be achieved in inadequately cleaned instruments.

9.

Other chemicals such as quaternary ammonia compounds (e.g. Cetrimide) are only lowlevel disinfectants and are inactive against many bacteria (pseudomonas, mycobacteria).They have little or no activity against viruses. Alcohol and iodine, while more effectivethan quaternary ammonia compounds, do not kill some forms of micro-organisms andare therefore not regarded as high level disinfectants.

It is customary to state that endoscopes undergo high level disinfection4. In practicalterms, however, endoscopes cannot always be rendered free of all bacterialcontamination by standard cleaning and disinfection processes. Endoscopes subjectedto the full cleaning and disinfection protocols advocated in this monograph and thenhaving their channels filled with culture medium and stored in sterile bags, may still growbacteria after several days. This is particularly so in older instruments whereirregularities at junctions, minor cracking or splitting of the surface layers of the internalchannels may allow protection of organisms5,6,7. The realistic aim, therefore, of anyreprocessing protocol is to have an endoscope, which will not transmit pathogens fromone patient to the next, nor hospital environmental contaminants from the endoscope oraccessories to the patient. In addition, it is important to recognise there are a widevariety of other factors which influence whether or not significant clinical infection willoccur when endoscopic procedures are undertaken. It is critical to have an appreciationof all the factors involved.

10.

Mechanisms of Infection and Major RiskFactors

1. MECHANISMS OF INFECTION

1. Clinical infections associated with endoscopy may occur because infective agentsare transmitted from one patient to the next via the endoscope or its accessoryequipment.

2. Hospital environment pathogens may contaminate the endoscope or accessoryequipment and be introduced into the patient during subsequent examination.Contamination may be from the general hospital environment, the water supply ordisinfecting machines. Previously the risks of clinical infection from this mechanismrelated mainly to E.R.C.P. but with the increasing use of disinfecting machines it israpidly becoming a more general problem.

2. RISK FACTORS

The important risk factors are:

1. The number and particular type of bacteria, virus or other infecting agents present onor in the endoscope, its water-feed system, diagnostic or therapeutic accessories.

2. The particular type of endoscopic procedure to be undertaken and whether tissuepenetration, damage or ischaemia occurs as a result of the procedure.

3. Patient factors:

a) Immune status.

b) Endovascular surface integrity.

c) Indwelling foreign material, e.g. prosthesis within tissues.

d) The presence of intrinsic infective foci.

11.

The Infecting Organisms

1. BACTERIA

(a) Salmonella and Related Species

Historically, salmonella and related species have been the infections most commonlytransmitted by endoscopy8-13. Many of the older literature reports of such infectionsoccurred with cleaning and disinfection regimens, which would not be consideredacceptable by today's standards. The majority of outbreaks were only recognised whenbacteriological laboratories reported unexpectedly large clusters of unusual salmonellaspecies triggering epidemiological investigation. The fact that these outbreaks were notrecognised by the endoscopy unit concerned but only on epidemiological investigationsuggests infections due to more conventional salmonella organisms may have beensignificantly under-reported. Some reports of salmonella outbreaks14 have beenassociated with inadequate cleaning of accessories, particularly the failure toultrasonically clean spiral wire wound accessories. Increasing chemical immersion timewas ineffective in at least one of these outbreaks and the problem was only terminatedwhen proper cleaning procedures were employed.

(b) Mycobacteria

Mycobacteria are relatively resistant to most chemical agents including aldehydes15.Atypical mycobacteria are even more resistant and there are reports of atypicalmycobacteria totally resistant to glutaraldehyde 16,17.

There is no proven case of transmission of tuberculosis by gastrointestinal endoscopy.Numerous reports of mycobacterial transmission by flexible bronchoscopy, however,have been reported18-23. Mycobacterial infections during bronchoscopy have beenrelated to contaminated suction valves18, cracked biopsy channels19, contaminatedtopical anaesthetic solutions20 and contaminated disinfecting machines21,22. Epidemicsof pseudoinfection associated with contaminated disinfecting machines have also been acause of considerable confusion24. Hanson25 has shown in a study using bronchoscopesheavily contaminated with Mycobacterium tuberculosis that adequate cleaning reducedcontamination by a mean of 3.5 log10 colony forming units. All bronchoscopes were freeof detectable mycobacteria after ten minutes in 2% glutaraldehyde. Nonetheless, thesheer number of cases of flexible bronchoscopic transmission of tuberculosis indicatesthat this is a significant clinical hazard. The Centre for Disease Control andPrevention recommends that bronchoscopy should not be performed on patientswith active T.B. unless absolutely necessary. Bronchoscopy should not be regardedas a first line investigation in the diagnosis of TB and repeated sputum smears should benegative for acid fast bacilli before bronchoscopy is considered.

Nowhere, but nowhere, has the critical role of cleaning been better demonstrated thanwith Mycobacterium tuberculosis and fibreoptic bronchoscopes. It has been welldemonstrated that immersion for sixty minutes in 2% glutaraldehyde will not eliminateMycobacterium tuberculosis from an inadequately cleaned bronchoscope3. Nicholson3

showed that a bronchoscope, which had undergone ten separate complete disinfectioncycles with 2% glutaraldehyde but had been poorly cleaned, was still contaminated withMycobacterium tuberculosis.

Rinsing of bronchoscopes after disinfection should be with sterile filtered water. Atypicalmycobacteria are frequently present in tap water. Full air/alcohol drying at the end oflists is critical.

12.

A further disturbing development in the mycobacterial area is the development ofmultidrug-resistant tuberculosis (MDRTB)26. These strains have been reportedprincipally from the eastern U.S.A. and infection transmission has largely been byrespiratory aerosols. An outbreak in southern United States with fatalities wasconvincingly traced to inadequate reprocessing of flexible bronchoscopes. The difficultyof tracing the bronchoscopic source of infection is well indicated in the report by Micheleet al. In this study a patient developed tuberculosis six months after bronchoscopy. Itwas shown by DNA fingerprinting that infection was from a strain of tuberculosis from apatient bronchoscoped two days earlier27. The cleaning and disinfection schedule wasinadequate in another study leading to infection28.

METICULOUS DETAILED MECHANICAL CLEANING BY STAFF PROPERLY TRAINEDIN BRONCHOSCOPE REPROCESSING REMAINS THE BEST AND INDEEDPROBABLY THE ONLY DEFENCE AGAINST TRANSMISSION OF MYCOBACTERIALDISEASE BY FLEXIBLE BRONCHOSCOPY. IT HAS BEEN UNEQUIVOCALLYDEMONSTRATED THAT EVEN EXTREMELY PROLONGED BRONCHOSCOPEIMMERSION IN 2% GLUTARALDEHYDE WILL NOT PREVENT DISEASETRANSMISSION IN INADEQUATELY CLEANED INSTRUMENTS ANDACCESSORIES.

(c) Serratia marcescens

If more evidence is required of the pivotal role of adequate mechanical cleaning inendoscope reprocessing then it is provided by Serratia marcescens. Several outbreaksof Serratia marcescens infection have been tracked to bronchoscopic transmission29,30.In an outbreak involving three fatalities29, the instrument had been inadequately cleanedbut then subjected to a full ethylene oxide sterilizing process, underlining the fact thatany attempts at sterilization or disinfection are likely to be ineffective in the presence ofinadequate cleaning.

(d) Helicobacter pylori

There is clear historical evidence that Helicobacter pylori was transmitted by researchstudies involving gastric tubes, endoscopy and biopsy, long before the organism wasclinically recognised (epidemic achlorhydria)31-33. Retrospective examination of biopsiesdemonstrated the presence of Helicobacter pylori31. Helicobacter pylori transmission bycontaminated biopsy forceps was demonstrated using restriction enzyme analysis ofbacterial DNA34. It is probable that endoscopic transmission of helicobacter has beenmore frequent than has been recognised because of:

(i) the high background prevalence of symptoms similar toHelicobacter pylori infection in the population examined;

(ii) the high background prevalence of Helicobacter pylori infection;

(iii) the non-specific nature of symptoms associated with Helicobacter pylori-induced gastritis; and

(iv) the frequency of asymptomatic infection

It has been suggested that a significant proportion of “adult reinfection” in some researchstudies is due to reinfection by inadequately processed biopsy forceps.

13.

(e) Clostridium difficile

There are several reports of possible endoscopic transmission of Clostridium difficile butnone have been definite35. Clostridium difficile spores are less resistant to a variety ofchemical disinfectants than test spores used in standard analytical chemical sporicidaltests35. Exposure for 10 minutes to 2% glutaraldehyde has been shown to inactivateClostridium difficile spores36.

(f) Pseudomonas

Pseudomonas is a common hospital environmental pathogen. Endoscope andaccessory contamination has almost invariably been acquired from the hospitalenvironment rather than from previous patients. Historically, endoscopy-associatedpseudomonas infections have largely been confined to E.R.C.P. and this problem isconsidered in more detail under that section. Clinical infections with pseudomonas,however, may become significant in the severely immuno-compromised patient,particularly when procedures involving tissue disruption are undertaken. However, evensimple diagnostic procedures not usually associated with mucosal trauma such asdiagnostic upper gastrointestinal endoscopy have been associated with pseudomonassepticaemia in severely immuno-compromised patient with gross oropharyngealmucocitis37,38 (e.g. leukaemia, bone marrow transplantation). Colonisation of automaticdisinfectors has resulted in serious disease transmission to patients.

2. VIRUSES

(a) Human Immunodeficiency Virus (HIV)

Infective HIV particles are present in the blood, semen and other body fluids of infectedindividuals. Needle stick injury with H.I.V positive blood has resulted in sero conversionranging from 0-0.42% in various studies39-44. The concentration of HIV in serum varieswidely with the stage of the infection. High viral concentrations can be found associatedwith all stages of HIV/AIDS. HIV is sensitive to many chemical disinfectants includingaldehydes45-47. A variety of studies has shown that when the virus is protected within adried protein coagulum, some chemical disinfectants including 1% glutaraldehyde will failto inactivate the virus within 5 minutes45, emphasising the absolute necessity to ensurethat scrupulous manual cleaning removes all traces of blood and proteinaceous material.Such cleaning should be undertaken without delay. In a series of studies Hanson et a148-49 contaminated the surface and internal channels of endoscopes with high titre HIVserum. Simple manual cleaning removed HIV activity from all except a singleendoscope and the remaining viral activity was removed from this endoscope after 10minutes or less in 2% glutaraldehyde. Where endoscopes were sampled after removalfrom HIV positive patients, all HIV present on endoscopes was removed by manualcleaning alone.

To date there has been no unequivocal demonstration of transmission of humanimmunodeficiency virus by gastrointestinal endoscopy. It is difficult to interpret the rarereports suggesting that some human immunodeficiency viral material may remain onendoscopes after recommended reprocessing protocols. The PCR techniques used mayidentify remaining nucleic acids, which do not constitute infective viral particles. Deva etal50 has shown that in the Duck Hepatitis B model, positive PCR material remaining onscopes does not correlate with infective transmission.

However the extremely long incubation time for clinical AIDS would make the detectionof a very isolated instance of HIV transmission difficult to detect.

14.

(b) Hepatitis B

Hepatitis B is a highly infectious virus and high concentrations of viral particles are foundin the blood of symptomatic hepatitis B sufferers and asymptomatic hepatitis B carriers,particularly those who are HBeAg positive. Clinical hepatitis B may occur as frequentlyas 1 in 3 following needle stick51-55 (compared with 1 in 400 becoming HIV positive afterneedle stick injury with HIV infected blood). Despite the high infectivity of hepatitis B,there is only a single well documented case of transmission of hepatitis B byendoscopy56. Clinical studies following up patients who have been endoscoped on thesame endoscopy list as known hepatitis B positive patients have produced no evidenceof infection56-64. Hepatitis B virus is moderately sensitive to the majority ofchemicals65,66. However, chemical inactivation requires that the germicide comes incontact with the virus and failure to remove blood, mucus and protein coagulums willallow the virus to be protected from chemical inactivation.

(c) Hepatitis C

In the Third Edition of Infection and Endoscopy there was no information available on thepossible transmission of Hepatitis C by endoscopy. Human body fluids including saliva,ascites and urine, all contain significant concentrations of Hepatitis C virus in infectedpatients. The risk of infection following needle stick injury with HCV positive blood isaround 3%. Given the known physical characteristics of the virus, its infectivity undergeneral clinical conditions and its known sensitivity to disinfectants, it was assumed thatthe risks of endoscopic transmission would be similar to Hepatitis B.

Unfortunately this appears not to be the case. There is now convincing evidence oftransmission of Hepatitis C associated with endoscopic procedures. In many cases thisseems related to inadequate cleaning. Tennenbaum et al reported the transmission ofHepatitis C following endoscopic sphincterotomy in 199367. Andrieu et al68 found in agastroenterology hospitalised population in patients over the age of 45, endoscopicbiopsy appeared to be the second highest risk factor for Hepatitis C, an odds ratio of 2.7compared with an odds ratio of 1.8 for blood transfusion. A national blood transfusionsurvey in France reviewed over two and a half million blood donations and found 30 antiHCV positive blood donors who had made a previous donation but had screened antiHCV negative. Six of 26 donors had a history of endoscopy between negative andpositive donations in the absence of any other identifiable risk factor69.

Bronowicki et al70 also reported from France HCV transmission during colonoscopy froma known infective patient to the two subsequent patients on the list. It appears likely thatthe cause of endoscopic transmission was a totally inadequate cleaning protocolincluding failure to brush the biopsy channel. However the biopsy forceps andpolypectomy snare were also inadequately processed.

Transmission of Hepatitis C during gastroscopy has also been reported by Crenn et al71.Single strand conformational polymorphism analysis of the hypervariable region of HCVRNA confirmed the patient to patient transmission. It is claimed that adequatereprocessing protocols were followed for the endoscope. It is unclear in this patientwhether the anaesthetic procedure or the endoscope was the cause of the transmission.Goudin et al72 have shown that there is a much greater incidence of HCV RNA inendoscope channels immediately after use when biopsies are taken compared with nobiopsy. They again found that conventional reprocessing techniques removed all HCVinfected material. In contrast to some of the above findings, Goudin et al in a study inLyon, France tested all patients referred for endoscopy for HCV and could find nodefinite evidence of HCV transmission and only one possible case73.

15.

At present proven transmissions of Hepatitis C by endoscopy remain confined to France.It is unlikely that this geographical restriction will continue. There are very few studiesaround the world which have prospectively examined the possibility of endoscopy as arisk factor for Hepatitis C transmission. A study by Kim et al74 from Korea could notidentify endoscopy as a risk factor. In French epidemiological studies it is impossible toknow what may have been the cause of Hepatitis C transmission.

However, in all except one clinical report there have been clear and gross deficiencies inthe endoscope and accessory reprocessing. This is not altogether surprising sinceRaymond in 199075 found that 73% of all units surveyed in France had gross protocoldeficiencies. This, however, should not lead to any sense of complacency elsewhere.Reynold’s survey in the U.S.A.76 in 1992 showed that 40% of units surveyed hadinadequacy in some aspects of their protocols. There are no recent Australian surveysbut past surveys were little better and there is recent anecdotal evidence that the veryprotocol failures associated with transmission of Hepatitis C at colonoscopy had beenpresent until recently in a small number of Australian endoscopy units.

At present the overwhelming evidence is that cleaning and disinfection protocols whenproperly applied during endoscope and accessory reprocessing will render instrumentsand accessories free of the risk of transmission of Hepatitis C. Failure to preventendoscopic transmission of Hepatitis C has been due to wilful or inadvertent deficienciesin appropriate cleaning and disinfection protocols or (possibly) inadequate anaesthetictechniques.

(d) Enteroviruses

Polioviruses are more resistant to many chemical disinfectants than the viruses whichhave a high lipid content (e.g. HIV). Hanson et al77 studied the elimination of enteroviruses from endoscopes using polio virus as the pilot agent. Endoscopes wereartificially contaminated with high levels of polio virus and subjected to standard cleaningand disinfection protocols. In further studies the effectiveness of glutaraldehyde againstcell free and cell associated polio viruses dried to a surface in a protein coagulum wasalso studied. Cleaning and disinfection was totally effective against a heavy viralcontamination and glutaraldehyde rapidly inactivated polio virus even when dried to asurface in serum.

3. OTHER INFECTIONS

A wide variety of other bacteria, viruses, fungi and protozoa could potentially betransmitted by endoscopy. Relatively little investigation has been undertaken in this areaalthough candidal infection of immunocompromised patients has been reported78. Anepidemic of pseudoinfection with the yeast Rhodotorula rubra has been reported inbronchoscopy patients79.

The sensitivity of many unusual organisms to chemical disinfectants is largely unknown.However some agents such as the oocysts of cryptosporidia are highly resistant to avariety of chemical disinfectants including 2% glutaraldehyde80,81. It is unlikely that suchorganisms pose a significant threat to patients with normal immune systems. Howeverthey could be responsible for serious and even fatal infections in theimmunocompromised.

16.

Creutzfeldt Jakob Disease (CJD - Prion Disease)

Spongiform encephalopathies are a family of aggressive neurological disorders whosesymptoms include dementia, ataxia, myoclonus, pyramidal and extrapyramidaldamage82-99. These diseases are characterized by the accumulation of a modified prionprotein PrPres which is an isoform of a normal protein PrPc. The abnormal form has anidentical amino acid sequence but the two isoforms differ in their three dimensionalconformation and glycosylation patterns. The protein appears to be encoded by PRNPGon chromosome 20. Mutations of Codon 200 result in familial forms of CJD andpolymorphism at Codon 199 may influence susceptibility to the infection.

The disease occurs in sporadic, familial, iatrogenic, occupational and new variant forms.Sporadic disease accounts for 90% of known cases and the mode of acquisition and/ortransmission is not known. Less than 10% of cases are familial. The “new variant”disease suggests that the prion source may be BSE contaminated beef.

There are approximately 160 cases worldwide of iatrogenic or occupationally acquiredCJD82,83,84,85,91,96,97. The vast majority of these relate to dura mater transplant (66) or theuse of human cadaveric growth hormone (76). Other sources include contaminatedneurosurgical instruments (4), human cadaveric pituitary gonadotrophin (4), brainelectrodes (2) and corneal transplants (3). Occupationally acquired CJD has beendocumented in only three workers, all of whom had percutaneous exposure to high risktissues in laboratory settings.

The risk of CJD transmission is a function of the relative concentration of the abnormalprion protein in a particular tissue84,90,96. Brain, dura mater and cornea are high risktissues. Cerebrospinal fluid, kidney, liver, lymph nodes and spleen contain a mediumconcentration of CJD agent. Low levels to no infectious agent are present in blood,urine, faeces, heart, bone marrow, nasal mucous, peripheral nerves, saliva, sputum ortears. It should be noted that there is no increase in risk of CJD in haemophiliacs whohave a massive life time exposure to transfusion and blood derived products.

Prion proteins are highly resistant to a variety of chemical and physical processes whichnormally inactivate microbiological agents. In addition, the catastrophic nature of thedisease has engendered such an emotional response that many recommended riskcontainment strategies have been grossly excessive. It again needs to be stressed thatno iatrogenic or occupationally acquired CJD has occurred from exposure to low to norisk tissues.

A conservative endoscopic approach, therefore, is:

§ Seek alternative diagnostic studies or therapeutic approaches in patients with knownor suspected spongiform encephalopathies.

§ Where such procedures are totally unavoidable, refer such patients to a large centrewhere specific endoscopes are reserved for patients with CJD.

§ Discard all endoscopic accessories used in patients with known or suspected CJD.

§ No change can be recommended to the indications for endoscopy or current cleaningand disinfection protocols in patients who have no increased risk of CJD.

17.

The Procedures

1. DIAGNOSTIC ENDOSCOPY

Bacteraemia occurs in any situation where there is mucosal trauma100-114 (even vigorousteeth cleaning). The presence of bacteraemia after an endoscopic procedure does notnecessarily indicate that a risk of serious clinical infection is present, but it does providean index of the degree of mucosal trauma. Significant rates of bacteraemia have beenreported where older style large diameter endoscopes have been studied103-109. Theorganisms found on blood culture were largely oropharyngeal commensals with a lowlevel of pathogenicity and the quantitative number of organisms recovered wassmall104,106. ANTIBIOTIC PROPHYLAXIS IS THEREFORE NOT INDICATED FORROUTINE UPPER ENDOSCOPY WHERE STANDARD PANENDOSCOPES AREUSED IN HEALTHY PATIENTS, THE IMMUNOCOMPROMISED, THOSE WITHCARDIAC VALVE PROBLEMS, ARTIFICIAL INDWELLING VASCULAR DEVICES,JOINT PROSTHESES, SILICONE TISSUE IMPLANTS OR PACEMAKERS. THESINGLE EXCEPTION TO THIS MAY BE TRANSPLANT OR CHEMOTHERAPYPATIENTS WITH SEVERE MUCOSITIS. Pseudomonas septicaemia has been reportedin leukaemia patients undergoing routine endoscopy37 and one study has reported a highrate of bacteraemia following endoscopy in bone marrow transplant patients38, although asubsequent study was unable to confirm this115. It is likely that the risk in this situationdepends upon the degree of mucositis present.

2. OESOPHAGEAL DILATATION

Disruption of the oesophageal mucosa invariably occurs during dilatation; indeed,endoscopic inspection following dilatation can show quite frightening tissue trauma. It isnot surprising that significant levels of bacteraemia have been recorded in associationwith oesophageal dilatation116,117. Organisms recovered at blood culture havesometimes been oropharyngeal commensals but more pathogenic organismscontaminating the endoscope or accessory equipment have also been recovered.ANTIBIOTIC PROPHYLAXIS IS RECOMMENDED FOR OESOPHAGEAL DILATION INIMMUNOCOMPROMISED PATIENTS OR THOSE WITH SIGNIGICANT CARDIAC ORVASCULAR ABNORMALITIES.

3. ENDOSCOPIC SCLEROTHERAPY/ BANDING

The degree of tissue damage occurring at the time of endoscopic sclerotherapywill depend upon the volume of sclerosant injected and whether this is intra orextra-variceal118. Very significant tissue destruction can occur with extra-varicealinjection. The majority of patients in whom endoscopic sclerotherapy is undertaken havea compromised immune system and it is not surprising that SERIOUS CLINICALINFECTIONS AND FATALITIES HAVE BEEN RECORDED IN ASSOCIATION WITHINJECTION SCLEROTHERAPY118-125. Some of these complications have beenlocal, e.g. mediastinal abscess, others have been more generalised includingsepticaemia or distant abscess, e.g. brain abscess. New needle catheters withcovered tips for endoscopic injection therapy have been reported126. The distal tip of thecatheter is covered with rubber and the needle only punctures the rubber when insertionof a needle into the varix or other tissue being injected (e.g. bleeding ulcers) is imminent.Investigations show this covered needle catheter system reduces the number ofcontaminating bacteria and may be useful in preventing bacteremia. Bacterialendocarditis has been reported following variceal sclerotherapy124,127.

18.

Again, the organisms recovered have varied widely but have included organismscontaminating endoscopes and accessory equipment. It is recommended thatANTIBIOTIC PROPHYLAXIS BE GIVEN IN SEVERELY IMMUNOCOMPROMISEDPATIENTS WHERE SIGNIFICANT ASCITES IS PRESENT OR SIGNIFICANT EXTRA-VARICEAL INJECTION HAS OCCURRED. Oesophageal banding is associated withsignificantly less tissue trauma and antibacterial prophylaxis is not usuallyindicated128,129.

Where latex allergy is known or suspected, latex free bands should be used (see sectionon Latex Allergy).

4. COLONOSCOPY

Low levels of bacteraemia have been reported in association with diagnosticcolonoscopy130-140 but the organisms recovered have been more pathogenic than thoseassociated with upper gastrointestinal endoscopy. There appears to have been anassociation with Streptococcus bovis endocarditis in some patients with villousadenoma141. It is important to recognise that manipulation of the sigmoid in patientswith acute peridiverticular inflammation or abscess formation is likely to result ingross bacteraemia. WHERE COLONOSCOPY IS UNDERTAKEN IN THE PRESENCEOF ACUTE PERIDIVERTICULAR ABSCESS, ANTIBIOTIC THERAPY IS INDICATEDEVEN IN PATIENTS WITH NORMAL IMMUNE COMPETENCE. ANTIBIOTICPROPHYLAXIS IS INDICATED IN COLONOSCOPY FOR THOSE WITH CARDIAC ORVASCULAR ABNORMALITIES, PERITONEAL DIALYSIS OR SEVERELYCOMPROMISED IMMUNE STATUS.

5. E.R.C.P

ENDOSCOPIC RETROGRADE CHOLANGIOPANCREATOGRAPHY IS THE ONLYENDOSCOPIC PROCEDURE WHICH HAS BEEN ASSOCIATED WITH ASIGNIFICANT RATE OF PROCEDURE INDUCED INFECTION142-154. Infections haveoccurred both sporadically and in mini epidemics. It is an unfortunate fact that most ofthe mini epidemics have only been detected by hospital infection control processes,mainly as a result of phage typing of pseudomonas found in bile recovered at operation.

Endoscopic retrogradecholangiopancreatography is the only

endoscopic procedure, which has beenassociated with a significant rate of

procedure-induced infection.

There are a number of reasons why E.R.C.P. is associated with this increased level ofprocedure-induced infections:

(i) Contamination of instrument and accessory equipment

INFECTIONS INDUCED AT E.R.C.P. HAVE ALMOST INVARIABLY BEEN WITHPSEUDOMONAS OR SIMILAR ORGANISMS (including Proteus spp.) These areubiquitous commensal organisms that colonise almost any damp surface. The usualsource of pseudomonas has been the channels within the endoscope itself althoughoccasionally contamination of accessory equipment has been responsible. The majorcauses of infection that have been traced as a result of single clinical cases of infectionor mini epidemics have included:

19.

(a) Inadequate disinfection of the endoscope with particular faults being related toinadequate cleaning and disinfection of the forceps raising channel155,156.

(b) Failure to rinse the channels at the end of the post-session cleaning anddisinfection process with alcohol and to subsequently dry the channels withforced air155,156.

Because of the risk of residual dampnessallowing proliferation of remaining

organisms, duodenoscopes must alwaysundergo full disinfection prior to the

commencement of a list.

(c) Contamination of the water feed system and water157,158. The water bottle andconnecting tube must be sterilised before the commencement of each E.R.C.Psession. Sterile water must be used in the water bottle. Use autoclavable waterbottle and tubing to eliminate the possibility that glutaraldehyde residue maycontribute to pancreatitis.

(d) Contamination of disinfecting machines by Pseudomonas (see DisinfectingMachines section).

Routine bacteriological surveillance of duodenoscopes and accessories should beperformed monthly (see Microbiological testing of endoscopes).

(ii) Patient related factors

Obstruction of the bile or pancreatic duct will greatly increase the risk of infectionfollowing E.R.C.P. and if duct obstruction is demonstrated at the time of examination,then every attempt should be made to ensure adequate drainage by sphincterotomy andstone extraction, sphincterotomy and stenting, or naso-biliary drainage. If drainagecannot be achieved by these methods then consideration should be given to earlysurgical or percutaneous transhepatic intervention. Significant tissue damage is likely tooccur if the pancreatic duct is over-filled or during manipulation to place stents. Theclinical risks of infection will be compounded if the patient has a compromised immunestatus. WHEREVER INTERVENTIONAL E.R.C.P PROCEDURES ARE LIKELY TO BEUNDERTAKEN OR WHERE THERE IS A SIGNIFICANT LIKELIHOOD OF DUCTOBSTRUCTION, THEN CONSIDERATION SHOULD BE GIVEN TO PROPHYLACTICANTIBIOTICS. It is important to obtain significant tissue levels of antibiotics at the startof the procedure. Antibiotics are therefore generally administered intravenouslycommencing one hour before the start of the procedure.

6. PERCUTANEOUS ENDOSCOPIC GASTROSTOMY

Bacteraemia may occur during percutaneous endoscopic gastrostomy insertion from thenecessary endoscopic manoeuvres. The passage of the inevitably contaminated end ofthe gastrostomy appliance through the stomach and abdominal wall makes the risk oflocal infection significant and antibiotic prophylaxis is recommended for all patients159,160.

20.

7. ENDOSCOPIC ULTRASOUND

There are conflicting reports of the rate of bacteremia during endoscopic ultrasound, onestudy161 reporting no evidence of bacteremia, another claiming significant bacteremia in6.3% of patients162. The longer distal segment without bending may result in someultrasonic endoscopes causing more tissue trauma during manipulation. Until moreevidence is available it may be prudent to offer antibiotic prophylaxis to patientsundergoing endoscopic ultrasound who have HIGH RISK cardiac lesions.

21.

The Patient with Increased Susceptibility toInfection

A variety of clinical circumstances may increase the danger of infection associated withendoscopy. These will include:

1. Compromised immune status.2. Procedurally induced tissue damage.3. Intrinsic sources of bacteraemia.4. Increased susceptibility to bacterial lodgement associated with septicaemia.

1. COMPROMISED IMMUNE STATUS

Impaired immune status is a major risk factor for significant clinical infection associatedwith endoscopic procedures37,115,120,121. The most important clinical conditions associatedwith impaired immune status include:

(a) Infections, particularly Human Immunodeficiency Virus infection.

(b) Neoplastic disease (especially malignancy associated with the lymphoreticularsystem (lymphomas and leukaemias).

(c) Cancer therapy (radiotherapy, chemotherapy).

(d) Transplant patients (particularly bone marrow transplantation).

(e) Advanced systemic disease including advanced liver and renal disease.

(f) Specific disorders of immune response (e.g. hereditary hypogammaglobulinaemia).

Patients with compromised immune status are more susceptible to infection with ordinarypathogens but are also at significant risk from organisms not ordinarily consideredpathogenic. In addition they may, themselves, harbour unusual organisms which may bedifficult to detect and resistant to chemical disinfectants (e.g. Cryptosporidia). Whilesuch atypical organisms pose a relatively minor risk to patients with normal immunesystems they may constitute a serious threat to other immunocompromised patients.Hospital water supplies when contaminated with Pseudomonas or atypical mycobacteria,even at low levels, may be sufficient to pose a significant threat to theimmunocompromised.

2. PROCEDURALLY-INDUCED TISSUE DAMAGE

Endoscopic procedures that result in tissue disruption or damage due to mechanical,chemical or inflammatory processes render the patient at even greater risk of infection.Significant tissue disruption occurs during oesophageal dilatation, removal of sessilepolyps and in difficult bile duct stone extraction associated with sphincterotomy. Majortissue damage due to chemical factors occurs during injection sclerotherapy if significantparavariceal injection has taken place. It may also occur where absolute alcohol orsclerosing agents are injected into tumours or in attempts to arrest bleeding from chroniculcers or Dieulefoy's abnormalities. Inflammatory processes such as post E.R.C.P. acutepancreatitis also increases the risk of infection. CONSIDERATION SHOULD BE GIVENTO PROPHYLACTIC ANTIBIOTIC ADMINISTRATION IN PATIENTS WITH IMPAIREDIMMUNE STATUS AND EVEN THOSE WITH NORMAL IMMUNE STATUS WHEREMAJOR TISSUE DAMAGE HAS OCCURRED.

22.

3. INTRINSIC SOURCES OF INFECTION

Intrinsic sources of infection which may be activated by endoscopic procedures willinclude acute peridiverticular abscess, E.R.C.P in the presence of cholangitis or infectedpseudocysts or any other situation where infected lesions are present in or adjacent tothe organ being examined. Consideration must be given to prophylactic antibioticadministration.

4. INCREASED RISK OF BACTERIAL LODGEMENT FOLLOWINGSEPTICAEMIA

(a) Endovascular integrity

Any abnormality of the endovascular surface, particularly if a high flow rate is associatedwith turbulence, will render the patient more susceptible to bacterial lodgement. Thehighest risk will be associated with prosthetic or incompetent valves. Other high riskfactors will include stenotic valves, arteriovenous shunts and previous endocarditis.

(b) Foreign material

Indwelling intravascular devices such as long-term venous access systems (e.g.portacaths, Hickmann catheters) will constitute a high-risk group. Foreign material withinthe body, but not in the intravascular space, carries a very substantially lower risk.Septic arthritis of artificial joints has been reported only very rarely in association withendoscopic procedures163,164 and does not warrant antibiotic prophylaxis.

23.

Endoscopy and Antibiotic Prophylaxis

1. INDICATIONS

There is no evidence to suggest that patients undergoing routine upper gastrointestinalendoscopy require antibiotic prophylaxis. Patients undergoing procedures, which have ahigher incidence of bacteraemia, e.g. those involving the biliary tract, sclerotherapy oroesophageal dilatation may benefit, although this remains unproven. Prophylacticantibiotics are of proven value in E.R.C.P.

For this reason recommendations from both cardiac and microbiological societies varyquite widely165-176. As a generalisation, antibiotic prophylaxis can be recommended for:

(a) All procedures in patients with previous endocarditis.

(b) Patients with prosthetic heart valves, major valvular damage, arteriovenous shuntsand indwelling vascular devices in all endoscopic procedures except simplediagnostic upper gastrointestinal endoscopy using a slim panendoscope.

(c) Patients with seriously impaired immune status having colonoscopy or anytherapeutic procedure where tissue disruption is likely. If severe mucositis is presentantibiotic prophylaxis may be appropriate for even simple diagnostic endoscopy.

(d) E.R.C.P. and associated pancreatobiliary procedures where there is duct obstruction,tissue disruption or impaired immune status.

(e) Patients having endoscopic procedures adjacent to intrinsic infective foci (e.g.colonoscopy and peridiverticular abscess, E.R.C.P with cholangitis, or infectedpseudo cyst).

(f) Percutaneous endoscopic gastrostomy.

2. ANTIBIOTIC REGIMENS

Protocols for antibiotic prophylaxis in endoscopy have been recommended by a numberof bodies165-176 including the Committee on Rheumatic Fever and Endocarditis and theCouncil on Cardiovascular in the Young, Endocarditis Working Party of the BritishSociety for Antimicrobial Chemotherapy and Australian Therapeutic Guidelines onAntibiotics (available from Therapeutic Guidelines, North Melbourne, Victoria).

All three protocols are based on similar microbiological principles and the followingrecommended regimen is an amalgam of these protocols:

1. Standard Regimen

Ampicillin or amoxycillin 2 grams (for children 50mg per kilogram up to adult dose)intravenously just before the procedure commences, or intramuscularly 30 minutesbefore the procedure commences. In exceptional circumstances this may befollowed by 500mg (25mg per kilogram up to adult dose) intravenously or orally sixhours later but a second dose is not recommended routinely.

Plus

Gentamicin 2.5mg per kilogram up to a maximum of 80mg intravenously just beforethe procedure commences or intramuscularly 30 minutes before the procedurecommences.

24.

2. For patients allergic to Penicillin or on long term Penicillin therapy or havingtaken penicillin or a related Beta-lactam antibiotic more than once in theprevious month:

Vancomycin 1.0g (20mg per kilogram up to an adult dose) intravenously slowly overone hour, the infusion ending just before the procedure commences.

OR

Teicoplainin 400mg (6mg per kilogram up to an adult dose) intravenously orintramuscularly just before the procedure commences;

TOGETHER WITH

Gentamicin 2.5mg per kilogram (up to a maximum of 80mg) intravenously justbefore the procedure commences, or intramuscularly 30 minutes before theprocedure commences.

3. ANTIBIOTIC PROPHYLAXIS FOR E.R.C.P.

Cholangitis can be a serious complication following E.R.C.P. The incidence variesbetween 1-10% depending on the patient population. Mortality can be as high as 10%.The risk of cholangitis is increased if there is biliary obstruction, particularly hilarstrictures and when attempts at drainage are unsuccessful.

The role of prophylactic antibiotics in E.R.C.P. has been assessed recently in placebocontrolled randomised trials177-181. However, no E.R.C.P. studies have been performedto identify the dose, duration of treatment, timing of administration or the choice ofantibiotics. The role of prophylactic antibiotics in surgery has been studied for over 30years and the following general principles have been established182,183.

Optimal protection is obtained when the antibiotic is present in therapeutic levels in thetissues at risk at the time of surgery. The best results are obtained if the antibiotic isgiven intravenously two hours before the procedure. Animal studies have shown that theprotection provided decreases as the time interval between the dose and proceduredecreases and disappears if the antibiotic is given after the procedure. In general, asingle dose is as good as a prolonged course. The antibiotic used must be active againstthe expected pathogens and achieve therapeutic levels in the tissues at risk of infection.

Bacteraemia following E.R.C.P. is common, asymptomatic and usually self limited in theabsence of established infection184. The role of prophylactic antibiotics in E.R.C.P. is toprevent infection becoming established in at-risk areas such as undrained bile ducts, gallbladders with stones and pancreatic pseudocysts. Antibiotics must achieve therapeuticlevels in tissues around these areas, i.e. in liver, gall bladder wall and pancreas.

The expected pathogens at E.R.C.P. are E. coli, Klebsiella and Strep faecalis and thereis a possibility of Pseudomonas if instruments are not meticulously cleaned. Of theantibiotics addressed in the randomised trial, cephalosporins have a variable penetrationinto bile and do not cover Strep faecalis or pseudomonas. Timentin has an appropriatespectrum of activity but penetrates poorly into the bile. Piperacillin is actively secretedinto the bile but since the early nineties E. coli have become increasingly resistant due tothe development of beta-lactamase. Worldwide only approximately 60% of E. coli arenow sensitive to piperacillin. Ciprofloxacin achieves good levels in the bile with a goodspectrum of activity apart from only modest activity against Strep faecalis. Bothciprofloxacin and piperacillin achieve good levels in pancreatic tissue.

25.

The power of the study to detect a benefit with prophylaxis depends on the number ofpatients enrolled and the risk of cholangitis in that population. Two trials have shown nobenefit with prophylactic antibiotics. However, these studies lack statistical power due tothe enrolment of modest numbers of low risk patients177,178. Other trials have hadproblems with choice of antibiotic or timing of administration. One study with 551patients enrolled showed a modest reduction of 27% in the relative risk of cholangitis butthis difference was not statistically significant179. Another study enrolling high riskpatients showed a significant reduction in cholangitis with prophylactic piperacillin180.Ciprofloxacin has a good combination of broad spectrum of activity and penetration intothe biliary tree and pancreas and has been studied in two trials177,181. Recently theproblem of developing resistance to piperacillin has been addressed by combining abeta-lactam inhibitor, tazobactam, with piperacillin (Tazocin). This combination isactively secreted into the biliary tree and has a broad spectrum of activity with betteractivity against Strep faecalis than ciprofloxacin.

Review of randomised trials show that prophylactic antibiotics provide a modest benefitin patients at risk for cholangitis. Animal studies and randomised trials in surgicalpatients suggest that an appropriate antibiotic should be given intravenously 1½-2 hoursbefore the procedure. The choice of antibiotic will depend on local resistance patternsand may change as bacterial resistance evolves. Tazocin or ciprofloxacin are currentlyappropriate for most Australian centres. Units should discuss local sensitivities with theirmicrobiologist. Patients at risk - those with biliary strictures, stones in the bile duct or gallbladder and pancreatic pseudo cysts – should be given an appropriate antibiotic twohours before E.R.C.P.

Not all Infectious Diseases Physicians would agree with the above choices for antibioticprophylaxis for E.R.C.P. There is concern that the use of broad spectrum agents suchas ciproxin may have long term detrimental effects from the point of view of antibioticresistance. An alternative view for prophylactic antibiotics would be cephalothin orcephazolin or gentamycin preceding E.R.C.P. with continuation of amoxil/clavulanateorally for three days.

4. BRONCHOSCOPY AND ANTIBIOTIC PROPHYLAXIS

Bacteraemia following fibreoptic bronchoscopy has a very low incidence. Blood culturestaken after bronchoscopy were negative in all 100 cases in one study and were positivein 1 out of 50 cases in a second study185,186. There is only one case report of infectiveendocarditis following fibreoptic bronchoscopy that occurred in a 24 year old HIV positivepatient187. The American Heart Association recommendations are that endocarditisprophylaxis is not indicated for flexible bronchoscopy except that it is optional forpatients with high risk cardiac lesions (prosthetic valves, cyanotic congenital heartdisease, surgically constructed systemic pulmonary shunts or conduits) and patients whohave previously had bacterial endocarditis188.