Synopsis  

The prevention of infections associated with endoscopy is complex and requires meticulous attention to detail. Practical limitations of applying recognized sterilizing and high-level disinfection processes to endoscopes must be understood, as must the mechanisms of infection and the organisms which provide the greatest clinical risks. Antibiotic prophylaxis will depend not only on the particular endoscopic procedure undertaken but also on a variety of host factors. Detailed endoscope reprocessing protocols are discussed with particular emphasis on the problems of biofilms and the difficulties of providing rinse water of adequate quality. Automatic flexible endoscope reprocessors are widely used but the numerous problems associated with these machines are often inadequately addressed. Part of any quality control program must be adequate microbiological surveillance of endoscopes.

 Sterilization and disinfection   

Sterilization   

Sterilization is defined as the complete elimination or destruction of all forms of microbial life [1]. Sterility is a state more easily conceived of than demonstrated in a practical way. How can the sterility of a batch of medical devices be proven? Microbiological examination of a few items of a batch does not exclude a 1/100 000 or a 1/1 000 000 possibility that the batch contains an article which is unsterile. In practice safety assurance levels (SAL) are used [2]. This technique studies a selected microorganism (usually a resistant bacterial spore) under fixed conditions in a sterilizing process and extrapolates the chance of remaining organisms from the inactivation curve. The usual convention is for a medical device labeled as 'sterile' to have an SAL of 10^-6, i.e. there is one in one million chance or less that the item is unsterile. This is an arbitrary definition which has no intrinsic scientific merit. For example, there is no evidence that items with an SAL of 10^-3 have more adverse patient outcomes than items with an SAL of 10^-6.

High-level disinfection   

High-level disinfection is defined as the elimination or destruction of all forms of microbial life except high concentrations of some spores [3].

What level of disinfection is required?   

Earle H. Spalding devised a practical approach to patient safety and medical device usage.

Critical items   

Critical items are those which enter sterile tissues, including body cavities and vascular spaces, and should be sterile.

Semi-critical items   

'Semi-critical items' are those which come into contact with mucous membranes or 'skin that is not intact', and should undergo at least high-level disinfection.

This has proved to be a workable, practical, and largely safe classification. In clinical practice the boundaries are not so clearly defined. Where does 'non-intact skin' end and 'sterile tissue' begin? Who can know with certainty that an endoscope will 'come into contact' with intestinal mucosa but not breach it?

It is generally accepted that endoscopes should undergo high-level disinfection and that accessories which breach mucosa, or may breach mucosa, should be sterile [4]. These will include biopsy forceps, snares, injecting needles, dilators, ERCP equipment, and implantable devices such as stents.

The practical problem   

There is a curious belief that if endoscopes are subjected to a 'sterilizing process' then many of the problems of endoscopy associated infections will disappear. Indeed, at least one state of the United States of America is considering passing a law to this effect. Unfortunately, this attitude reveals a basic ignorance of the fundamental problem faced in endoscope reprocessing. Endoscopes cannot be subjected to reliable, easily validated, high temperature disinfecting or sterilizing systems because of the heat sensitive nature of the materials from which they are constructed. In addition the complexity of long, fine, interconnecting channels which are difficult to clean, impossible to inspect, and which cannot be adequately assessed for channel surface irregularities and damage, make endoscope reprocessing a special challenge for any system of sterilization or disinfection [4].

These physical characteristics of endoscopes have dictated that time consuming manual cleaning by well trained staff is necessary. It has been demonstrated on numerous occasions that any attempt to achieve high-level disinfection or sterilization by any of the currently available techniques, including ethylene oxide exposure and all chemical sterilants, will be ineffective if appropriate meticulous manual cleaning has not been completed [5,6].

Biocides   

Chemical biocides which are used for endoscope reprocessing throughout the world include glutaraldehyde, peracetic acid, chlorine dioxide, hydrogen peroxide, and ortho-phthalaldehyde. A number of new processes including those based on acid electrolytic water are currently undergoing evaluation. None of these chemicals or processes can be considered ideal.

In most developed countries biocides used for high-level disinfection of endoscopes must receive appropriate government approval after submission of extensive validation data. It must be recognized that the effect of biocides on microorganisms is a complex process where variation in temperature, concentration, and remaining organic material on devices, markedly alter efficacy and it is therefore critical that the agent be used strictly as directed by the manufacturer.

The organisms   

The level of intrinsic microbial resistance to chemical sterilants varies widely [4]. Spore forming organisms are relatively resistant. Fortunately the common microorganisms of greatest concern to endoscopy, particularly HBV and HIV, are among the most sensitive of all microbiological agents to some chemical sterilants. HCV, being a lipid containing virus, is predicted to have a similar sensitivity to HBV. Intermediate levels of resistance are shown by common vegetative bacteria. Some increased resistance is present with mycobacteria and non-lipid containing viruses, e.g. polio virus, hepatitis A virus. These levels of resistance are far less than for bacterial spores.

The critical points in reprocessing   

The critical points of any endoscope reprocessing system:

• Adequate cleaning is by far the most important part of any endoscope reprocessing system.
  
• Numerous studies show that no disinfecting or 'sterilizing' process capable of use with current flexible endoscopes will achieve even high-level disinfection, let alone sterilization, if adequate cleaning has not been undertaken first.
  
• Currently there are no studies in peer reviewed journals that consistently show that automated machine cleaning of endoscopes can achieve satisfactory removal of biological material.
  

 Risks of infections associated with endoscopic procedures   

Mechanisms of infection   

• An infectious disease may be transmitted from a patient to subsequent patients because infected biological material is not totally removed from the endoscope or accessories during reprocessing.
  
• Endemic health care facility pathogens, including those present in the water supply, may contaminate and colonize endoscopes, automatic flexible endoscope reprocessors (AFERs), storage areas, or water feed systems.
  
• Endoscopic manipulation may result in bacteremia from the patient's endogenous gut flora.
  

Clinical infections   

The American Society for Gastrointestinal Endoscopy has estimated that the overall risk of patient to patient transmission of a serious infection at endoscopy is 1 in 1.8 million examinations. Since this estimate is based on retrospective rather than prospective studies, it is almost certainly a significant underestimation, but it does indicate the rarity of patient to patient transmission of serious infectious diseases. The risk of endoscopy associated infections due to the contamination of instrument or accessory items by health care facility environmental pathogens, or infection with the patient's own flora, is very significantly higher.

Whether or not a clinical infection occurs as a result of an endoscopic procedure will depend upon:

• infecting organisms
  
• the endoscopic procedures
  
• host factors.
  

Infecting organisms   

A vast array of microorganisms could be associated with endoscopy transmitted infections. In practice a relatively limited number of infective agents have been transmitted by endoscopy. The more important agents together with those that are of major theoretical concern are considered here.

Bacteria   

Vegetative bacteria   

Numerous reports in the older literature of transmission of salmonella and related species were associated with cleaning and disinfection protocols which would be totally inadequate today [7]. Transmission of such infections indicates serious deficiencies in cleaning and disinfection which would be cause for major concern and investigation.

Clostridium difficile   

Clostridium difficile is a spore forming organism and therefore might be expected to be highly resistant to chemical disinfectants. There is clear evidence of person to person transmission by environmental contamination at hospital ward level. Fortunately, Clostridium difficile spores are much less resistant to chemical disinfectants than most other spore forming organisms. To date there is no definite evidence of endoscopic transmission of Clostridium difficile.

Mycobacterium tuberculosis   

There is no proven case of transmission of Mycobacterium tuberculosis by gastrointestinal endoscopy. Numerous infections have been transmitted as a result of bronchoscopy [8]. Contaminated suction valves, damaged biopsy channels, contaminated topical anesthetic sprays, and colonized AFERs have all been implicated.

Hanson et al. [9] have shown that proper cleaning of bronchoscopes deliberately contaminated with Mycobacterium tuberculosis reduced the bacterial bioburden by at least 3.5 log₁₀ and subsequent immersion in 2% glutaraldehyde for 10 min rendered the instrument free of infective material. Nicholson et al. [6] dramatically illustrated the critical role of cleaning by showing that a deliberately contaminated flexible bronchoscope which was cleaned inadequately still had remaining viable Mycobacterium tuberculosis after 10 sequential full disinfection cycles.

An additional hazard is the development of multidrug resistant tuberculosis (MDRTB). Infection transmission has largely been by aerosol. An outbreak with deaths in the southern United States was convincingly traced to inadequate reprocessing of flexible bronchoscopes.

The Center for Disease Control and Prevention recommends bronchoscopy should not be performed on patients with active TB unless absolutely necessary. Bronchoscopy should not be regarded as the first line investigation in the diagnosis of TB. Patient to patient and patient to staff transmission of mycobacterium tuberculosis by aerosols associated with coughing during and after bronchoscopy is a significant hazard.

Scrupulous, mechanical cleaning by properly trained and certified endoscope reprocessors remains the best defense against the transmission of mycobacterial disease by flexible bronchoscopy. (See also section on pseudomonas).

Atypical mycobacteria   

A variety of atypical mycobacteria may be present in the hospital water supply. These organisms frequently colonize AFERs and are prone to develop serious chemical resistance. Some strains of atypical mycobacteria (particularly Mycobacterium chelonei) have developed almost total resistance to glutaraldehyde. Atypical mycobacteria contaminating flexible bronchoscopes have caused pseudo-epidemics. Positive specimen cultures taken at bronchoscopy have been wrongly assumed to be patient infections, leading to potentially seriously toxic and prolonged drug therapy [10].

Serratia marcescens   

There are several reports of Serratia marcescens being transmitted by flexible bronchoscopy. In an outbreak involving three fatalities the instrument had been inadequately cleaned but then subjected to a full ethylene oxide sterilizing cycle [5].

Helicobacter pylori   

Helicobacter pylori has been transmitted by endoscopy biopsy forceps which were inadequately cleaned [11]. Given the high background prevalence of Helicobacter pylori infection it is likely that disease transmission by endoscopy has been significantly under reported. There is conflicting evidence about whether endoscopists and endoscopy nurses have an increased risk of helicobacter infection.

Pseudomonas   

Pseudomonas aeruginosa and related species are common environmental contaminants which may be present in tap water. Clinical endoscopy associated infections due to Pseudomonas spp. have largely been confined to ERCP and related procedures [12,13]. There are several reports of transmission by ordinary endoscopy in severely immunocompromised patients. Pseudomonas spp. frequently colonize AFERs and may be difficult to eradicate. Kovacs et al. [14] have reported a strain of Pseudomonas aeruginosa which appeared to have developed chemical resistance. The organism was responsible for several ERCP-related chemical infections. A disturbing feature of ERCP-associated pseudomonas infections has been the frequent failure by endoscopy units to recognize the problem.

Several serious episodes of bronchoscope contamination with pseudomonas have been reported from the United States recently. Serious clinical infections and even fatalities have been linked to the contaminated bronchoscopes. The mechanisms of contamination remain hotly debated, but a poorly designed port valve, defective or wrong connectors between bronchoscopes, and AFERs during reprocessing appear to be major factors.

The valve port in question was designed as a non-removable part of the bronchoscope (this in itself would seem to be a seriously flawed concept). Apparently microbiological cultures of the bronchoscope channels taken using samples collected by perfusion of the channels from the control head to the distal tip were negative but samples collected with reverse flushing grew pseudomonas.

The episodes apparently associated with AFER connectors yet again emphasize the risk of AFERs which do not have flow alarms on individual channels and do not test rinse water quality [15–18].

Viruses   

Human immunodeficiency virus (HIV)   

The transmission of HIV to a number of patients undergoing minor surgical procedures has created major public concern that medical procedures including endoscopy may be a serious source of disease transmission. Inflammatory articles in popular magazines in the USA, e.g. 'Do scopes spread sickness?''Medicine's dirty little secret', and 'Blood money' have all resulted in increasing public scepticism concerning the self-regulation of the medical profession. There is increasing demand for greater official regulation and accountability. Despite these concerns there is no proven case of transmission of HIV by endoscopy. However, the extremely long latent period before the development of clinical AIDS increases the difficulty of detecting transmission. High concentrations of viral particles may be present in the blood during most stages of HIV infection. Fortunately the virus is very sensitive to most chemical disinfectants including glutaraldehyde [19].

Elegant 'in use' studies by Hanson et al. [20,21] have demonstrated that endoscopes used in patients with HIV infection have had all infectious material removed by standard cleaning and disinfection protocols. There have been some reports suggesting that artificially contaminated endoscopes had remaining viral material after standard reprocessing. However, in all of these reports detection has used PCR techniques which do not distinguish between live infective viral particles and non-infective degraded viral material. Deva et al. [22] have shown in the duck hepatitis model that PCR positive material remaining after reprocessing is not infective.

Hepatitis B   

Despite the highly infectious nature of this virus there is only a single proven case of hepatitis B transmission by endoscopy [23]. The reprocessing protocol used on this instrument was clearly inadequate. A number of studies have followed patients undergoing endoscopic procedures with instruments and accessories used on known hepatitis B positive patients and none have found evidence of disease transmission [24].

Hepatitis C (HCV)   

Hepatitis C has been transmitted to a patient undergoing ERCP and endoscopic sphincterotomy [25]. HCV transmission occurred during colonoscopy from a known infective patient to the two subsequent patients [26]. A French National blood transfusion study suggested that endoscopy was a significant risk factor for HCV acquisition. Another French study, however, could find no evidence of increased prevalence of hepatitis C amongst patients undergoing endoscopy. Seven cases of hepatitis C transmitted at a Brooklyn endoscopy clinic were unrelated to the endoscopy procedure [27,28]. The infections were due to reuse of syringes or needles used for the administration of sedation during the procedures. A large number of hepatitis C infections appear to have been transmitted in an American day surgery center by the same mechanism [29]. Transmission of hepatitis C during simple endoscopy has been proven by detailed viral analysis [30]. In this case it was claimed that the reprocessing protocol was adequate as distinct from the other cases reported above. In this report it remains unclear whether disease transmission was associated with the endoscopic procedure or anesthetic administration.

Prions   

Prion diseases include Creutzfeldt–Jakob disease (CJD), new variant CJD (vCJD), and kuru in human beings. Animal diseases include bovine spongiform encephalopathy (BSE) and scrapie. Prions are unique amongst infectious agents because nucleic acid (DNA or RNA) has not been detected. The disease is characterized by an abnormal isoform of a cellular protein called prion protein (PrP^c). Mutations in the PrP^c gene on chromosome 20 may result in transformation into the pathological isoform (PrP^sc).

CJD   

CJD occurs in sporadic, familial, iatrogenic, and occupational forms. Less than 200 cases worldwide of iatrogenic or occupational acquired CJD have been reported [31]. The majority of these relate to dura mater transplants or the use of human cadaveric growth hormone. Rarer sources include contaminated neurosurgical instruments, cadaveric pituitary gonadotrophin, brain electrodes, and corneal transplants.

The risk of infection transmission depends on the concentration of abnormal prion in the tissue examined. High concentrations exist in the brain (including the dura mater), spinal cord, and eye. Low levels are present in liver, lymphoid tissue, kidneys, lungs, and spleen. Prions are undetectable in intestine, bone marrow, whole blood, white cells, serum, nasal mucus, saliva, sputum, urine, feces, and vaginal secretions. Attempts to transmit CJD by blood and blood products from humans to primates have consistently failed [32].

Prions unfortunately display a markedly increased resistance to conventional methods of sterilization. Commonly used disinfectants including glutaraldehyde are ineffective. Steam sterilization (standard gravity displacement at 120°C) is only partly effective even after exposure times as long as 2 hours.

What to do in practice about CJD?   

The dilemma faced by endoscopists is whether extraordinary precautions in reprocessing endoscopes need to be taken because of the risk of CJD. The approach described by Rutala and Webber [33] is recommended—that endoscopes which are exposed to essentially 'no risk tissues' will not be exposed to detectable levels of prion and therefore should be reprocessed by conventional disinfection protocols. Clearly this does not apply to any endoscopic device used in neurosurgery. This appears to be sound scientific advice. However, given the widespread emotional and press reactions to this devastating disease, few will be able to resist taking additional precautions. Many will believe that patients with known Creutzfeldt–Jakob disease, Gerstmann–Straussler–Scheinker syndrome, and fatal familial insomnia, and patients with undiagnosed rapidly progressive dementia should not undergo endoscopic examination unless there is no acceptable diagnostic or therapeutic alternative. If endoscopy is necessary in proven cases then examination should be undertaken with an endoscope reserved for this purpose. For those with a high suspicion of disease many would prefer to use endoscopes which have reached the point of retirement from clinical service and not to reuse the scope. This may be an argument for retaining some older endoscopes, at least in large facilities.

New variant CJD (vCJD)   

Variant CJD in humans appears to be due to the transmission of the identical prion strain causing bovine spongiform encephalopathy in cattle [34]. The disease has a different clinical course with a much earlier age of onset, presentation with psychiatric symptoms, and a rapid downhill progression. In complete contra-distinction to other forms of CJD large quantities of the abnormal prion (PrP^sc) are found in lymphoid tissue [35]. In fact vCJD can reliably be diagnosed by tonsil biopsy. This obviously raises concerns that endoscopes exposed to lymphoid tissue in the alimentary tract may become contaminated with tissue in which significant concentrations of abnormal prions exist. Currently quantification of risk transmission is hampered by a lack of knowledge about relative prion titres in alimentary lymphoid tissues. Recent development of a highly sensitive immunoblotting assay for vCJD should shortly allow a clearer understanding of transmission risks associated with endoscopy [36]. Until such information is available it is difficult to recommend any effective changes to cleaning and disinfection protocols for endoscopes.

Other infections   

A wide variety of other bacteria, viruses, fungi, protozoa, and helminths could potentially be transmitted by endoscopy. Candida infections in immunocompromised patients have been related to endoscopy. Pseudo-infection with Rhodotorula rubra was associated with bronchoscopy. Strongyloides infection of the esophagus and cryptosporidial infection have been linked to upper endoscopy. Many of these infectious agents are likely to represent a greater hazard to immunocompromised patients.

The endoscopic procedures   

Bacteremia may be associated with simple everyday events such as teeth cleaning. The significance of the bacteremia will depend on the particular types of organisms present, the numbers of organisms, whether the tissues are inflamed, and the degree of mucosal trauma.

Upper gastrointestinal endoscopy   

Simple diagnostic endoscopy and biopsy are associated with low levels of bacteremia, usually with non-virulent organisms. Clinically significant bacteremia may, however, occur in patients with compromised immune status and severe mucositis (e.g. bone marrow transplantation, leukemia).

Disruption of the esophageal mucosa almost invariably occurs during esophageal dilatation. This procedure is associated with high levels of bacteremia [37] and a number of serious clinical infections have been reported.

Endoscopic sclerotherapy, particularly where there is a significant submucosal injection, is associated with high levels of bacteremia. In addition these patients are often severely immunocompromised [38].

Endoscopic banding is associated with significantly lower rates of bacteremia [39].

Lower gastrointestinal endoscopy   

Somewhat surprisingly, only low levels of bacteremia have been reported in association with diagnostic colonoscopy. However, manipulation of the sigmoid colon in patients with acute peridiverticular inflammation or abscess formation undergoing colonoscopy may result in gross bacteremia.

Endoscopic retrograde cholangiopancreatography   

ERCP is the one endoscopic procedure which has consistently been associated with significant clinical infection rates. Clinical infections may be due to the patient's endogenous flora, particularly if duct obstruction is not totally relieved at the time of the procedure. However, the vast majority of clinical infections have been with Pseudomonas aeruginosa or closely related species and have been due to contamination of the endoscope or accessory equipment [12,13,40,41]. AFERs, contaminated water feed systems, inadequate cleaning and disinfection of the forceps elevating channel, and failure to alcohol rinse and air dry ALL duodenoscope channels have been the most common underlying causes.

Percutaneous endoscopic gastrostomy   

Bacteremia during the procedure and wound infection have led to significant procedure related complications [42].

Endoscopic ultrasound   

There have been conflicting reports on the degree of bacteremia occurring during endoscopic ultrasound.

Mucosectomy   

Significant rates of bacteremia have been reported following endoscopic mucosectomy particularly where large quantities of tissue are removed.

Host factors   

Immune competence   

Patients with compromised immune status are significantly more susceptible to endoscopy associated infections. A wide variety of disorders can affect immune competence. These will include infections (e.g. HIV), neoplastic disorders (particularly hematological malignancies), chemotherapy, radiotherapy, bone marrow transplantation, advanced diseases of the liver and kidney, and specific disorders of immune response. These patients are not only more susceptible to infections with conventional organisms but may also be susceptible to a variety of other organisms not usually associated with human disease (e.g. atypical mycobacteria infection). In addition they may harbour unusual organisms. In most cases these organisms will not pose a clinical threat to other patients with normal immune status but may be a hazard to those with compromised immune systems. It is important to remember that hospital water contamination with atypical mycobacteria, pseudomonas, or even cryptosporidia may pose a particular hazard to the immunocompromised.

The degree of tissue damage   

The greater the procedural damage caused during an endoscopic procedure, the greater the risk of subsequent infection. Significant tissue disruption can occur during esophageal dilatation. Chemical and ischemic damage to tissues occurs with injection sclerotherapy. Endoscopic sphincterotomy and stone extraction, removal of foreign bodies, endoscopic placement of stents, mucosectomy, and removal of large sessile polyps are all associated with significant tissue damage.

Intrinsic sources of infection   

Intrinsic sources of infection within the patient may contribute to clinical infection. Poor oral hygiene with severe bacterial gingivitis will ensure greater contamination of instruments and accessories in the upper GI tract. Peri-diverticular abscess, infected pseudocysts, and intra-abdominal collections may be traumatized during endoscopic procedures.

Damaged valves and implants   

There is a significant risk that bacteria present during periods of bacteremia will lodge on damaged or foreign tissues. The most important factor here is endovascular integrity. Colonization can occur with indwelling vascular devices, vascular grafts, and coronary artery stents before complete re-epithelialization. Mechanical heart valves, valve abnormalities which create turbulent flow, and other irregular endovascular surfaces are prone to bacterial lodgement. Artificial joints and other indwelling foreign devices may be slightly more prone to infection following endoscopic procedures but the risk is very small and mainly within the first few months following insertion.

 Antibiotic prophylaxis for endoscopic procedures   

There are three situations where antibiotic prophylaxis may be considered for endoscopic procedures:

• Patients at increased risk of bacterial endocarditis.
  
• Patients at increased risk of general infections following particular procedures.
  
• ERCP.
  

Principles of prevention of bacterial endocarditis   

The American Heart Association stresses that 'there are currently no randomised and carefully controlled human trials in patients with underlying structural heart disease to definitively establish that antibiotic prophylaxis provides protection against development of endocarditis during bacteremia inducing procedures. Further, most cases of endocarditis are not attributable to an invasive procedure'[43]. There is no definite consensus on which patients should receive antibiotic prophylaxis for which endoscopic procedures and indeed there is no real consensus on the preferred antibiotic regimen [44]. Nonetheless, some general guidelines may be offered. There is a consensus that some comorbid cardiovascular conditions pose a greater risk than others.

High risk cardiovascular conditions [43]   

• Prosthetic heart valves.
  
• Previous history of endocarditis.
  
• Complex cyanotic congenital heart disease.
  
• Surgically constructed systemic pulmonary shunts or conduits.
  

Moderate risk cardiovascular conditions [43]   

• Uncorrected shunt defects.
  
• Bicuspid aortic valves.
  
• Coarctation of the aorta.
  
• Acquired valvular dysfunction.
  
• Hypertrophic cardiomyopathy.
  
• Mitral valve prolapse without regurgitation is considered a low risk condition but a moderate risk where regurgitation exists.
  

Recommendations for antibiotic prophylaxis   

Who should receive antibiotics?   

The perceived need for antibiotic prophylaxis varies also by the type of endoscopic procedure. General recommendations are given in Fig. 1.

Clinical problems where opinions diverge   

Indwelling vascular devices. Antibiotic prophylaxis may be of value for patients undergoing endoscopic procedures with a high rate of bacteremia, particularly if they have a compromised immune system.

• Recent coronary artery stenting. Antibiotic prophylaxis has been recommended by some authorities in the first 3–4 months following stenting until epithelialization has occurred.
  
• Orthopedic prostheses. There are isolated case reports of orthopedic prosthetic infection associated with endoscopic procedures. However, the risk is extremely low. A recent survey of program directors of infectious disease training programs found that more than 50% of respondents felt that antibiotic prophylaxis was not indicated for any endoscopic procedures in patients with artificial joints [45].
  
• However, there were wide variations in recommendations and there appeared to be little scientific basis for some views. The risk is certainly highest immediately after joint replacement and many would recommend antibiotic prophylaxis for the first few months after joint replacement, particularly if the patient has any impairment of immune competence.
  

What antibiotic regimen?   

While there is little agreement on details of prophylactic regimens, the general principles are accepted. It is important to ensure adequate antibiotic concentrations in the serum during and after the procedure. To reduce the likelihood of microbiological resistance it is important that prophylactic antibiotics are given during the operative period. Commonly used antibiotic regimes are detailed in Fig. 2.

 Antibiotic prophylaxis for ERCP   

The value of antibiotic prophylaxis for ERCP is also controversial [46]. Some of this confusion has arisen because of the inappropriate use of the term 'prophylactic'. There can be little argument that patients with clinical cholangitis or other evidence of biliary or pancreatic sepsis should be on appropriate antibiotics. There is also general consensus that patients who have undergone traumatic procedures with major tissue manipulation, incomplete drainage of obstruction, or widely dilated duct systems should continue to receive appropriate antibiotics.

The area of controversy is whether patients with minimal or no bile duct dilatation undergoing simple procedures such as endoscopic sphincterotomy and stone removal require antibiotics commencing before the procedure. A recent meta-analysis of studies examining antibiotic prophylaxis prior to ERCP concluded that while it may reduce the instance of bacteremia, it did not substantially reduce the incidence of clinical sepsis/cholangitis. One of the difficulties in deciding for or against antibiotic prophylaxis commencing before the procedure is that the complexity and outcome of the procedure cannot always be accurately predicted.

Optimum benefit of antibiotics will only be obtained if therapeutic levels are present in the bile and tissues at the time of examination. Patients should commence antibiotic prophylaxis intravenously at least 1–2 hours before the procedure. The common pathogenic organisms encountered in the biliary tree are Pseudomonas aeruginosa, Klebsiella spp., E. coli, Bacteroides spp., and Enterococci.

Prophylactic antibiotic regimens for ERCP   

Ciprofloxin oral 750 mg—2 hours before procedure

IV 200 mg—2 hours before procedure

Piperacillin—before procedure

OR

Piperacillin + tazobactam

The reason for giving antibiotics needs to be clearly borne in mind. Is the risk simply of cholangitis or is there also a significant risk of endocarditis because of valvular damage or other abnormalities? Cephalosporins and ureidopenicillins (e.g. piperacillin) have very poor activity against enterococci and are generally considered inappropriate for endocarditis prophylaxis.

 Principles of effective decontamination protocols   

Cleaning is essential   

The most important step in the process of endoscope decontamination is scrupulous manual cleaning prior to disinfection.

Manual cleaning refers to the physical task of removing secretions and contaminants from the endoscope with appropriate brushes, cloths, detergents, and water. This is a two-stage process, beginning immediately upon removal of the endoscope from the patient in the procedure room, and continuing once the instrument has been taken to the specific reprocessing area.

Automated industrial processes used for cleaning have yet to be shown to be effective in endoscope reprocessing. All instruments currently need to be manually processed prior to chemical disinfection. It is desirable that in future automated systems be developed which physically lift the soils present and then remove them by fluid flow.

In order for manual cleaning to be effective it must:

• Be performed by a person conversant with the structure of the particular endoscope and trained and certified in cleaning techniques.
  
• Be undertaken immediately after the endoscope is used so that secretions do not dry and harden.
  
• Follow a protocol which, using appropriate detergents and cleaning equipment, allows all surfaces of the endoscope, internal and external, to be cleaned. Recommended protocols produced by respected authors are detailed in references [47–51].
  
• Be followed by thorough rinsing to ensure that all debris and detergents are removed prior to disinfection.
  

Effectiveness of recommended protocols   

A standard for testing of cleaning efficacy in endoscope reprocessing procedures has not yet been developed. Several studies have examined methods such as ATP bioluminescence in an endeavor to provide an accurate marker of cleanliness.

Recommended reprocessing protocols remove microbiological contamination. However, even minor deviations from cleaning protocols have resulted in persistent microbiological contamination after disinfection. This emphasizes that present reprocessing techniques are less than ideal and have a lower margin of safety than is desirable. It reinforces the need for all steps in reprocessing protocols to be carried out meticulously.

Endoscope structure   

There are at least 50 different models of flexible endoscopes available, with each manufacturer constantly expanding their range. The manufacturer supplies an instruction book with each endoscope.

It is essential that every person responsible for endoscope decontamination reads these instruction books and is familiar with the particular characteristics of each model of endoscope required to be cleaned.

Common features   

External features   

All flexible fiberscopes have a light guide plug, an umbilical cable (cord), a control head, and an insertion tube.

• The light guide plug. The light guide plug fits into the light source. The air/water and suction channels have ports in the light guide plug. The light guide plug of a video endoscope is heavier than that of a fiberscope and needs to be handled with care. Most light guide plugs have electrical connections that need to be sealed by a protective cap prior to immersion in liquid.
  
• The umbilical cable/universal cord. The umbilical cable connects the light guide plug to the body of the endoscope. The external surface may be contaminated by splashes or hand contact during endoscopic procedures.
  
• The control head. The control head contains the control handles, which allow the operator to flex the instrument and to access the suction and air/water functions by use of valves. Fiberoptic endoscopes have an eyepiece on the control head. Video endoscopes are similar in construction to fiberoptic endoscopes, except that they do not have an eyepiece—the image is seen on a video screen. The control head is contaminated during endoscopic procedures by the operator's hands. The control handles have grooved surfaces, which must be carefully brushed during cleaning. The hollow structure of some control handles should be noted and care taken to ensure that the undersurface is thoroughly rinsed and emptied of fluids. The seats which house the suction and air/water valves (buttons) must be thoroughly cleaned. The biopsy channel port is located at the base of the control handle near its junction with the insertion tube. This port must be brushed carefully during the cleaning process.
  
• The insertion tube. The insertion tube enters the patient's body and is grossly contaminated during the procedure. The distal tip of the insertion tube houses the microchip (in video endoscopes), the openings for the suction and air/water channels, and the lens covering the flexible fiberoptic light guides. The section of the insertion tube adjacent to the distal tip is known as the bending section. Its covering is made from soft flexible material and is particularly vulnerable to damage.
  

Common internal features   

The suction and air/water channels and the fiberoptic light guide extend from the light guide plug to the distal tip. In non-video models an additional fiberoptic bundle, the image guide, extends from the control head to the distal tip. Wire cables, which allow the tip to be flexed, run through the insertion tube. Any damage to either the umbilical cable or the insertion tube can potentially damage any of the internal structures. Care must be taken during cleaning procedures to ensure that the umbilical cable and insertion tube do not become kinked or acutely bent. Kinks in biopsy channels trap debris and lead to failure of the cleaning process. Suspected damage should be referred to the supplier for assessment and repair. A negative leakage test does NOT preclude damage to internal endoscope structures.

Special internal features   

Most duodenoscopes have an additional channel—the forceps elevator (raiser) which is extremely fine (capacity 1–2 ml) and requires scrupulous attention during the cleaning process. Cleaning adaptors for this channel are provided with each duodenoscope.

Some colonoscopes have a carbon dioxide (CO₂) channel that may be connected to the air channel. Cleaning protocols must include flushing of this channel.

Jet (washing) channels are found in some endoscopes. These are contaminated during procedures and must be independently flushed during cleaning, whether or not they have been used.

Cleaning equipment   

All endoscopes are supplied with appropriate cleaning adaptors. It is vital that persons cleaning endoscopes are conversant with these adaptors and use them correctly. Rubber 'O' rings on the adaptors must be inspected regularly for defects or looseness and should be replaced as required.

Cleaning brushes for both channels and valve ports are also supplied. These have a limited life. They should be inspected regularly and replaced when worn or kinked. Metal wear from abrasion by cleaning brushes and other endoscope accessories may occur on the edge of the biopsy valve port or the suction button port.

Soft toothbrushes are useful to clean grooved control handles and to brush the distal tip and biopsy ports. Cotton buds may be used to clean biopsy valve caps but should not be used in the air/water port as threads can become caught and cause blocked channels.

Adequate supplies of disposable cloths or swabs should be available.

Cleaning fluids   

The desirable properties of detergents include loosening of particulate and other soil and microorganisms so that they are removed by the flushing process. Enzyme products promote protein lysis and enhance the efficacy of brushing and flushing and are the preferred detergent. Where an enzyme product is not immediately available, a neutral instrument detergent can be used. Household detergent is NOT suitable.

Manufacturers of enzymatic solutions report optimum efficacy when used in warm water. However, enzymes will continue to be active in water that has cooled to room temperature (20°C). The use of hot water (> 60°C) denatures proteins, inactivating the enzyme, and may fix both the enzyme from the detergent and any protein soil onto the instrument. Heavy contamination may exceed the enzyme capacity.

Rinsing   

Rinsing should take place under running water so that all traces of detergent and disinfectant are flushed away. Failure to adequately rinse glutaraldehyde from endoscopes has been reported to cause severe post colonoscopy colitis and may be responsible for some cases of post-ERCP pancreatitis. Ortho-phthaldehyde [OPA] can cause marked skin staining in both patients and staff if not thoroughly rinsed off equipment. In more severe cases blistering and skin necroses have occurred. Static rinsing, i.e. rinsing in bowls of water, is not recommended.

The amount of water required to thoroughly rinse an endoscope after disinfection will vary according to the design and length of the instrument and the chemical used for disinfection. Manufacturers' instructions for volume of rinse water should be followed.

See also 'Problem areas—rinsing water'.

Disinfectants   

Disinfectants for endoscope reprocessing need to have wide bacteriocidal properties, together with the ability to kill relevant viruses including HIV, HBV, and HCV. Testing should have been conducted under clinical operating conditions as well as under laboratory conditions. Many disinfectants have either a restricted spectrum of activity or have not been adequately tested.

Worldwide, glutaraldehyde is the most frequently used chemical disinfectant for use in unsealed systems. The most common formulation used is 2% activated alkaline. Other products used include peracetic acid, hydrogen peroxide, chlorine dioxide, and ortho-phthalaldehyde.

Soaking time   

Effective manual cleaning of the item to be soaked is critical in determining the effectiveness of chemical disinfection.

Endoscopes which are not adequately cleaned will not be adequately disinfected even with prolonged soaking times.

Chemical manufacturers are legally required to indicate disinfectant contact times on the product label. Many professional organizations have published guidelines, which recommend a shorter soaking time. Those recommendations are based on evidence that significantly less time is needed if the instrument has first been manually cleaned. Soaking times of 5–45 min for glutaraldehyde are common.

Other issues which effect soaking time include temperature and concentration of the disinfectant.

Staff required to chemically disinfect endoscopes must be provided with education in the safe use of hazardous chemicals, and with personal protective clothing which includes impervious gowns (or gowns and plastic aprons), gloves which have been approved for use with the specific chemical, and face shields.

General maintenance   

Leak testing of endoscopes should be performed after use as per manufacturers' instructions. Failure to detect a leak prior to thorough cleaning and disinfection may result in major damage to the instrument.

Examination of the instrument lens and outer sheath should be performed following each session to detect any signs of cracking or damage. The function of angulation cables should be checked.

Inspection of 'O' rings on valves for signs of wear should be performed at the end of each session. 'O' rings should be changed when signs of wear are detected. Biopsy caps should be checked for signs of wear and replaced as required.

Lubrication   

Lubrication is used to ensure optimal functioning of both endoscopes and accessories. The 'O' rings on suction and air/water control buttons require lubrication to prevent the buttons sticking in the depressed position. Traditionally silicone oil supplied with the endoscope has been used. Silicone oils can be either petroleum based or in a water soluble base. There is evidence that both preparations may impair reprocessing. Biological fluid can be entrapped within oil globules and protected from disinfectant action. The choice is therefore to either take particular pains to ensure complete removal of silicone based lubricants or to use surgical instrument lubricant.

Recommendations   

• Accessory items processed in ultrasonic cleaners should be lubricated with an instrument lubricant following completion of the ultrasonic cleaning. They should then be wiped with a clean, lint-free cloth and allowed to air dry prior to packaging for steam sterilization.
  
• Where silicone oil lubricants are used for suction and air/water control buttons, they should be applied immediately before use (after chemical disinfection). It is essential to remove lubricant residue to allow germicide contact. Ultrasonic cleaning will remove any small remaining amounts of lubricant.
  

Work areas   

Endoscopy should not be performed in centers where adequate facilities for cleaning and disinfection are not available.

Chemical disinfection must take place in an area with adequate physical controls such as forced air extraction. Soaking bowls must have close fitting occlusive lids. Forced air extraction should extend to the rinsing sink.

Work areas should be planned carefully. The areas should be well ventilated and the reprocessing area should include the following:

• At least one sink designated for the cleaning of instruments, referred to as the 'dirty' sink. This should be made of materials which are impervious to chemicals. Suitable materials include stainless steel, porcelain, or a plastic bonded material. The sink must be of sufficient dimensions to adequately hold a coiled full-length colonoscope without causing the instrument damage. The sink should be supplied with hot and cold running water.
  
• An area adjacent to this sink where the components of the instrument are removed for cleaning. The 'dirty' bench is then suitable for holding instruments awaiting chemical disinfection.
  
• An area for disinfection of instruments. In the case of automated disinfectors the dimensions and requirements are determined by the make and model of the machine to be installed. For manual disinfection, a chemical container of sufficient dimensions to hold an instrument without damage to the instrument is required. It is preferable that this container be a fixed sink placed under an appropriate fume extraction system. Otherwise a container especially designed for chemical disinfection of instruments should be available. This must be placed in a fume extraction cupboard.
  
• Where an automated disinfector is used, rinsing is performed within the machine. Where manual rinsing occurs, a sink designated for rinsing only clean instruments must be available and located within the fume extraction cover.
  

 Reprocessing regimens   

Disinfect before and after procedures   

It is known that stored endoscopes may become colonized with vegetative bacteria during storage, especially if the drying process is not adequate. Unfortunately the complex structure and fine channels of endoscopes preclude absolute certainty that drying processes are always effective. Therefore endoscopes must have a full disinfection process performed prior to use on the day.

At the end of a list, using 70% isopropyl alcohol to enhance the drying process, the endoscope must be thoroughly forced air dried prior to storage. Methylated spirits is NOT suitable for this process.

Manual cleaning   

The following steps should be performed immediately following a procedure.

• IMMEDIATELY after each procedure, with the endoscope still attached to the light source, grasp the control head. Using a disposable cloth soaked in detergent solution, wipe the insertion tube from the control head to the distal tip. Discard cloth.
  
• Place distal tip in detergent solution. Aspirate through suction channel—depress and release suction button rapidly to promote debris dislodgement.
  
• Depress and release air/water button several times to flush water channel. Occlude air button to force air through the air channel.
  
• Some types of endoscope have an air/water channel cleaning button that should be placed in the air/water valve seat at this point as per manufacturer's instructions to flush both the air and water channels. If there is no such adaptor, remove the water bottle connector from the endoscope, taking care not to contaminate its end. Drain the water channel by occluding the water inlet on the endoscope light guide plug. The endoscope should be removed from the light source, the waterproof cap should be placed over the electrical pins on the light guide plug, and the endoscope taken to the cleaning area. (If, due to local circumstances, there is a delay prior to thorough cleaning, perform the leak test then place the endoscope in a bowl of enzyme solution and soak.) It is essential that the endoscope is not allowed to dry prior to cleaning as this will allow organic material to dry, making removal from channels difficult or impossible.
  
• Remove all valves and control buttons. Leak test the instrument as per manufacturer's instructions. WARNING: It is essential that manufacturer's instructions be followed.
  
• Brush and clean valves and buttons, paying particular attention to internal surfaces. Place buttons in an ultrasonic cleaner.
  
• Place endoscope in enzymatic/detergent solution and using appropriate brushes provided by the manufacturer brush the suction/biopsy channel and the air/water channels if design permits. Particular attention needs to be given to ensuring that all sections of the channels have been accessed. If the brush contains obvious debris it should be cleaned before being withdrawn. Each channel should be brushed until all visible debris is removed. Wash all outer surfaces.
  
• Using a soft toothbrush, gently clean the distal tip of the endoscope.
  
• Brush control handles and biopsy port. Brush around valve seats.
  
• Clean valve seats thoroughly—check that all visible debris has been removed. Use special brushes if provided by manufacturer.
  
• Fit cleaning adaptors. Thoroughly flush all channels with enzymatic solution, ensuring all air is removed from channels. Allow solution to remain in contact for time required for the specific product.
  
• Rinse outer surfaces. Flush all channels thoroughly with fresh water. It is essential that all detergent be removed prior to disinfection.
  
• Purge channels with air to remove rinsing water.
  
• Disinfect as follows or machine reprocess as per specific machine instructions.
  

Manual disinfection   

• After manual cleaning immerse endoscope in disinfectant so that the entire endoscope is submerged. Fill all channels with disinfectant so that all air bubbles are expelled. All channel entrances must be under the surface of the disinfectant during this procedure to ensure that no air enters the channel. Remove the buttons from the ultrasound, rinse, dry, and then immerse buttons and valves in disinfectant or autoclave if applicable. It is preferable to have extra supplies of buttons and valves to ensure that adequate cleaning is performed prior to immersion in disinfectant.
  
• Soak instrument for required time in disinfectant of choice. A timer with an alarm is essential to ensure that accurate soak times are achieved. Digital timers are more accurate. A fluid thermometer with digital readout is recommended to constantly monitor temperature of disinfectant solution.
  
• Purge disinfectant from all channels with air and remove endoscope, valves, and buttons from disinfectant, taking care to avoid drips and splashes.
  
• Rinse exterior of endoscope thoroughly and flush channels with fresh water to remove traces of chemical. Rinse all valves and buttons thoroughly.
  
• Purge all rinsing water from channels.
  
• If the instrument is being prepared for reuse, remove the cleaning adaptors. Dry exterior surfaces with a soft cloth, lubricate 'O' rings on buttons, and reassemble endoscope.
  

If the instrument is to be stored do not remove cleaning adaptors and refer to point 3.

At the end of the list   

• Flush all channels with 70% isopropyl alcohol (approximately 2 ml for elevator channels, approximately 20 ml for each other channel). If using a multichannel cleaning adaptor the quantities of alcohol may need to be increased [51].
  
• Force air dry all channels. Ensure that the air source has a flow regulator and use lower pressure on fine channels. Use bayonet fittings rather than luer lock to attach the air tubing to the cleaning adaptors and fit securely but not tightly—if safe pressure is exceeded the bayonet fitting will give way. Use of excessive air pressure may cause damage to the instrument [51].
  
• Ensure that all outer surfaces are dry.
  
• Check the instrument for any sheath or lens damage. Polish the lens with the cleaner provided by the manufacturer. DO NOT REASSEMBLE ENDOSCOPE.
  
• Store endoscope (disassembled) in a well-ventilated storage cupboard, which permits full length hanging on appropriate support structures. Endoscopes should not be stored in transport cases as these may themselves become contaminated.
  
• Store buttons and valves separately (not attached to endoscopes). 'O' rings on buttons and valves should only be lubricated following disinfection prior to use.
  

Endoscopic accessory equipment   

The cleaning and disinfection of reusable endoscopic accessories is equally as important as that of the endoscope. Endoscopic accessories have been implicated in the transmission of infection. As with endoscopes, the cleaning of accessories as a prerequisite to sterilization is mandatory.

Cleaning accessories   

• All equipment should be immersed in enzymatic detergent immediately following use until cleaning can be performed.
  
• The equipment should be dismantled as far as possible and all visible soiling removed.
  
• Any spiral coil, hinged, or complex structured accessories should be placed in an ultrasonic cleaner and processed according to manufacturers' recommendations. The ultrasound should be used with the lid in place to avoid dispersal of aerosols generated as these may contain biological material.
  
• Any fine bore cannula or tubing accessory items will require thorough flushing with enzymatic detergent. Other accessory items, depending on design, will require a combination of flushing and brushing to clean surfaces.
  
• Following cleaning by either of these methods, accessory items should be thoroughly rinsed and dried prior to disinfection, autoclaving, or low temperature sterilization.
  

Disinfection   

General accessory equipment used in gastroenterological procedures requires high-level disinfection. Accessories that enter sterile tissue or the vascular system must be sterile. This includes biopsy forceps, injection sclerotherapy needles, and all accessories used for ERCP. Where an alternative exists, all non-autoclavable reusable accessories should be phased out.

1. All autoclavable equipment must be cleaned thoroughly prior to sterilization process.

2. All non-autoclavable equipment should be immersed in disinfectant, ensuring all cavities are flushed with the fluid. The soaking time will depend on whether the accessory item will be required to enter sterile tissue (see section on Sterilization and Disinfection).

Special accessory items   

Sclerotherapy needles   

Sclerotherapy needles are difficult to clean and reprocess to a sterile state and may provide an occupational hazard. Therefore it is recommended that only single use sclerotherapy needles be used.

Water bottles and connectors   

These accessory items should be autoclaved at the beginning and end of each session as they have been implicated in the transmission of infection. All non-autoclavable bottles and connectors should be replaced with those that are fully autoclavable.

Dilators   

Dilators are likely to come in contact with tissue that has been abraded or otherwise damaged by the dilation process. They should ideally be sterilized. They must at least have undergone high-level disinfection immediately before the session. Note the operative field will not be sterile as the patient's own microbiological flora will contaminate the area.

 Problem areas in endoscope reprocessing   

Rinsing water   

Poor quality water   

The endoscope may well become colonized with significant pathogens after appropriate cleaning and high-level disinfection if the rinsing water itself is contaminated. Pseudomonas spp., atypical mycobacteria, Legionella spp., cryptosporidia, and a variety of other organisms are frequent contaminants of hospital tap water. The microbiological quality of water varies dramatically from country to country. Surprisingly it may vary widely even within the same city dependent upon a variety of local factors. Municipal water quality varies with the age and condition of water mains. Local hospital factors will include the age and extent of plumbing alterations. 'Dead runs' are particularly important. A temperature of at least 55°C at the point of use is necessary to minimize the growth of organisms in the hot water supply. Contaminants other than microbiological agents may also affect the quality of delivered water. High sediment levels may block filters quickly.

Infections from rinsing water   

Organisms introduced into the endoscope by the rinsing water may colonize the instrument and be transferred to patients subsequently examined with the endoscope. The greatest clinical risk has proved to be the transfer of pseudomonas to patients at the time of ERCP. However, many organisms found in the water supply may pose significant clinical hazards to immunocompromised patients. The problem of contamination of AFERs is considered in more detail in that section. Atypical mycobacterial contamination of bronchoscopes has caused epidemics of pseudo-infection with significant clinical consequences to patients [10].

Bacteria free water   

Bronchoscopes and duodenoscopes should therefore be rinsed in bacteria free water. This may be prohibitively expensive for gastroscopes and colonoscopes but substantially bacteria-free water should be used. Unfortunately this is more difficult to achieve than first appears [52–54]. Adequate water filtration is both technically challenging and expensive. A filter bank (usually four) of decreasing mesh size down to a final filter of 0.2 microns is required. There must be a mechanism whereby the filters can be treated to remove microbiological contamination. The way in which this can be achieved will depend on a variety of circumstances including the particular filters, the available water pressure, and the quality of the water both in terms of sediment and microbiological content. In general it is best to have a closed loop arrangement with shut off valves on the input and output side of the filters. Access ports on the filter side of these shut off valves allow circulation of sterilants through the filter bank (note some filters will be damaged by reverse circulation). Agents used to treat the filter bank have included hot water, chlorine releasing agents, and glutaraldehyde. It is important to ensure that the agent used is compatible with the particular filters. Filters should be changed regularly and sent for microbiological examination at the time of replacement. It is important not to assume that 0.2 micron filters will be failsafe. Excessive water pressure, excess sediment, gross bacteriological contamination, etc. may all result in bacteria translocating through filter defects. Hospitals with high quality water may not have any problems. Others, presumably with lesser water quality, have reported almost insurmountable difficulties requiring overnight sterilization of filters on a daily basis together with frequent filter replacement.

Water testing   

The bacteriological quality of the rinse water available should be tested. Microbiological monitoring should be undertaken in liaison with the local microbiology laboratory. The tap mouth should be flamed to eliminate surface gram negatives (although these in themselves may constitute a significant finding) and a minimum of a litre of tap water collected. Bacteriological quality of tap water should be monitored at least yearly and after any known plumbing alterations at the hospital or unit. More frequent sampling may be appropriate where the tap water has previously been found to be contaminated, where the institution has old plumbing, and in hotter climates.

Recommendations for rinsing water   

• Duodenoscopes and bronchoscopes must always use sterile or 0.2 micron filtered water for rinsing.
  
• The bacteriological quality of the unit tap water should be monitored.
  
• Bacteriological filtering with 0.2 micron final filters is recommended for rinsing when poor water quality exists.
  
• Filter banks must be serviced and bacteriologically monitored on a regular basis.
  
• ALL endoscopes should have a final alcohol rinse followed by forced air drying at the end of procedure lists.
  
• Particular problems of AFERs are considered separately.
  

 Variation in cleaning and disinfection regimens depending upon the supposed infective status of the patient   

A number of surveys have shown that the practice of varying the cleaning and disinfection regimen according to the supposed infective status of the patient is widespread, with hospitals changing their reprocessing techniques after use in patients with known HIV infection, tuberculosis, or hepatitis [55–58]. There is clear evidence to show that the cleaning and disinfection schedule recommended in this review is adequate to prevent the transmission of infectious disorders including HIV infection, hepatitis, and tuberculosis [9,20,21]. There is therefore NO JUSTIFICATION to alter the cleaning and disinfection regimen if patients are known to have these disorders. (The problems associated with prion disease are considered separately.)

 Compliance with cleaning and disinfection protocols   

Investigation of clinical infections related to endoscopic procedures has almost invariably shown that there has been a breach of recommended cleaning and disinfection protocol. In a few cases it has not been possible to determine the reason for the infection. In at least one incident (hepatitis C transmission) the suspicion has fallen on the anesthetic technique [30]. Practice surveys in the past have shown poor compliance with recommended protocols. Raymond [55] in 1990 found that 73% of all units surveyed in France had serious protocol deficiencies. A study in the United States in 1992 showed 40% of units surveyed had unsatisfactory aspects in their reprocessing protocols [56]. Even worse compliance has been reported from a variety of other countries. Fortunately there is evidence of a substantial improvement in recent years. Surveys in the United States in 1998 [57] and 1999 [58] showed major improvements. Concerns remain with a small percentage of respondents reprocessing biopsy forceps and other critical items by glutaraldehyde disinfection rather than sterilization. Other concerns include rinsing water quality, persisting variation in reprocessing regimens depending on the patient's perceived infective status, and reuse of disposable items.

 The investigation of possible endoscopy infection transmission incidents   

Investigation may be necessary because of internal recognition of an equipment or protocol failure. Alternatively, a complaint may be initiated by a patient or external agency.

Common causes   

Common causes requiring incident investigation include:

• Self-recognition of protocol errors, e.g. failure to recognize and clean jet channels.
  
• Pump or other mechanical failure on AFERs without adequate alarm systems.
  
• Bacterial colonization of AFERs.
  
• Improper use of chemical disinfectants.
  

Golden rules for investigating potential infection incidents   

The golden rules for any unit involved in investigation of a potential infection transmission incident are:

• Inform the appropriate authority immediately.
  
• Do not try to avoid investigation and do not undertake investigation of patient complaints yourself.
  
• Do not argue with or suggest alternative possible sources of infection transmission to complainants.
  
• Ensure that the information listed above is readily available to the investigating authority.
  

The appropriate regulatory authority should be notified of any incident. Authorities commonly allow in-house investigation of self-recognized low-risk protocol failures. Investigation of patient complaints, and serious and potentially serious self-recognized incidents, must be conducted at arms length by an appropriate regulatory authority.

The investigation process   

The exact process of investigation will depend upon the particular incident but in general the regulatory authority will:

• Inspect the premises to ensure that there is compliance with registration, licensing, and credentialing requirements.
  
• Ensure that medical and nursing staff are qualified and have acceptable continuing education processes.
  
• Ensure that endoscopic equipment, endoscope accessories, reprocessing, and safety equipment are appropriate.
  
• Ensure that internationally accepted endoscope and accessory reprocessing protocols are used [47–51] and protocol compliance is documented.
  
• Ensure that bacteriological surveillance programs exist.
  
• Identify involved and at risk patients, endoscopes, accessories, and reprocessing equipment.
  
• Examine anesthetic procedures.
  
• Contact affected and at risk patients for serological sampling (e.g. HIV, HCV, HBV) or other appropriate investigations (e.g. sputum samples if tuberculosis transmission is a possibility).
  

Transmission of viral disease   

There is no current evidence that serious viral diseases such as HIV, HBV, or HCV have been transmitted from patient to patient by endoscopy if all aspects of internationally accepted endoscope reprocessing protocols have been followed [47–51].

 Automatic flexible endoscope reprocessors (AFERs)   

AFERs are widely used in the western world. American surveys in 1998 and 1999 showed around 70% utilization of AFERs [57,58]. These machines certainly reduce unpopular, arduous, repetitive tasks and reduce occupational exposure to irritant chemicals. Unfortunately AFERs have also been responsible for many serious clinical infections including deaths and epidemics of pseudo-infection [8,10,59–62]. The enthusiastic and largely uncritical acceptance of AFERs may owe more to their convenience than to clinical safety. There are numerous AFER models of widely varying quality, durability, and effectiveness.

Perceived advantages of AFERs include:

• Standardization of endoscopic reprocessing.
  
• Reduced exposure of staff to chemicals.
  
• Reduction of staff time spent on disinfection.
  

None of the currently available machines negate the need for thorough manual cleaning. This is an essential prerequisite to disinfection. Any claim that manual precleaning is unnecessary should be carefully scrutinized and not accepted unless it has been published in respected peer reviewed journals. Working parties are currently developing standards for AFERs under the European Committee for Standardization and the International Standards Organization. When completed, parts 1 and 4 of these documents will provide a reasonable international standard for machines.

Machine design and principles   

Contamination   

AFERs will rarely show contamination when new. Unfortunately this is when most AFERs are trialed. Problems with bacterial contamination rarely become apparent in machines before 6 months of use and become progressively more likely with ongoing use of machines and endoscopes. Increasing wear and age reveal unsuspected defects. Biofilms, valve failure, surface irregularities, fissuring, filter failures, and chemical familiarity all offer colonization opportunities for ever vigilant bacteria.

Water supply   

Machines should be plumbed into the water supply rather than be filled manually. Pre-filters are necessary. Filter systems must be regularly serviced and monitored. It is all too easy for filters themselves to become a major source of contamination (see section on Rinsing Water).

Alarm function   

The principal function of AFERs is to pump liquids through endoscope channels. Alarm functions to detect pump failure on all channels are essential.

Self-sterilization   

AFERs should have an effective self-sterilization cycle. Most AFERs claim to have one but manydo not stand up to rigorous scrutiny.

Fume containment   

The extraction of disinfectant fumes from within the machine should occur prior to completion of the operating cycle and prior to opening the machine. If this is not possible the machine must be contained within a fume extraction hood.

Disinfectant supply   

Machines which use a concentrated solution and in-use dilution for a single cycle (e.g. STERIS system) avoid the problem of dilution of the disinfectant with rinsing water. Machines which contain a tank of disinfectant for reuse should be monitored for disinfection concentration to determine appropriate disinfectant change schedules. Machines which require the filling of a disinfectant reservoir must incorporate a pump mechanism to obviate the need for pouring chemical sterilants into the machine.

Reprocessing time   

AFERs add significantly to endoscope reprocessing time.

AFERs cannot guarantee to sterilize endoscopes   

AFERs cannot guarantee to sterilize endoscopes despite some manufacturers' claims to the contrary. Remember that, unless adequate cleaning has taken place, endoscopes will remain seriously contaminated and have caused serious clinical infections including death, despite being subjected to 'sterilizing processes' including ethylene oxide and peracetic acid exposure.

Cost   

AFERs may increase reprocessing costs significantly. Real cost analysis should include the cost of purchase of the machines, extra endoscopes, machine service cost, and filtration costs. With poor quality water supply filtration costs can be extremely high.

Plumbing pathway   

Plumbing pathway diagrams must be factual and not schematic. Careful evaluation of plumbing pathways is essential to determine if claimed self-sterilization cycles are to amount to anything more than wishful thinking.

Rinse and dry cycle   

AFERs must have a terminal bacteria-free water rinse followed by 70% alcohol and air-drying cycle for use at the end of each list. If such a cycle is not part of the AFER's features it must be carried out manually prior to storage of each endoscope.

Regular bacteriological surveillance   

Regular bacteriological surveillance of AFERs is essential. Specimens are usually best collected by pumping not less than 1 litre of water through filters placed in line at the point of endoscope connection and culturing the filters.

 Quality control in endoscope reprocessing   

Reliable quality control systems are an integral part of any manufacturing or service process. Clinical medicine has been slow to embrace adequate quality control systems and endoscope reprocessing protocols have been particularly deficient. Given the complexity of endoscope construction, the difficulty of ensuring adequate cleaning, and the relatively low safety margin of chemical disinfectants used, formal quality control measures are essential. Automated systems which reduce human error, together with automatic recording of essential parameters and alarm systems are highly desirable.

Unfortunately endoscope cleaning remains a manual process. Education and certification of those cleaning endoscopes currently remains the most practical quality assurance process.

Disinfection, however, lends itself to automation and process monitoring. Regrettably the number of serious clinical infections associated with AFERs shows that current systems remain far from ideal.

Quality control measures   

Quality control measures in endoscope reprocessing should include:

• Appropriate education, examination, and certification of staff reprocessing endoscopes.
  
• Proof of compliance with internationally accepted endoscope and accessory reprocessing protocols [47–51].
  
• A record system which documents measurable reprocessing parameters. Where manual disinfection is employed this will include disinfection concentration, temperature, immersion time, etc.; for AFERs it will mean documentation of adequate cycle completion. It is essential that the unit has a clear system which links the particular endoscope with the patient examined, the patient's position on the list, the endoscope cleaner, and the person and/or machine involved in disinfection. A clear and reliable linkage system is essential in investigating possible infection transmission incidents.
  
• Accessories which breach sterile surfaces and are difficult to reprocess by a clearly validated sterilizing system should be single use only, e.g. sclerotherapy needles. Accessories which breach sterile surfaces but are not labeled for single use only and can be reprocessed by a validated reliable method such as steam sterilization need not be individually traceable. Accessories which breach sterile surfaces and are labeled 'for single use only' require an institutionally validated reprocessing protocol and must be individually traceable. In some countries, particularly the United States of America, institutionally validated reprocessing protocols will be extremely onerous [63].
  
• A formal system for equipment servicing, maintenance, and replacement. This must include endoscopes and accessories, AFERs, ultrasonic cleaners, and water filtration systems.
  
• A regular bacteriological surveillance program of endoscopes, AFERs, and water quality.
  
• It is desirable that each unit has some outcome auditing process, which should include at least intermittent infection detection surveys.
  

 Microbiological surveillance in endoscopy   

Microbiological surveillance of endoscopes, hospital water supply, water filtration systems, and AFERs is strongly recommended. The area remains controversial because of sampling difficulties, lack of methodological validation, and difficulty interpreting culture results. These criticisms have a degree of validity and highlight the need for supporting methodological evidence. They do not, however, remove the need for bacteriological surveillance, which remains an important indirect validation of reprocessing protocols, provides evidence of internal endoscope damage which may be otherwise undetectable, and allows the early detection of colonization of filtration systems and AFERs.

Deva et al. [22] have shown that the absence of bacterial contamination is an accurate reflection of viral elimination during reprocessing. Thus viral cultures are not performed for surveillance purposes.

Duodenoscopes   

There can be little serious argument against bacteriological surveillance of duodenoscopes. Serious clinical infections, usually with Pseudomonas aeruginosa or related species, occurred after ERCP in significant numbers in the past and continue to be reported. Endoscopists have frequently failed to recognize the endoscopic causation of these infections. As a result, errors in reprocessing protocols, internal endoscope damage, and colonization of AFERs have all failed early detection.

Bronchoscopes   

Failure to detect AFER colonization with atypical mycobacteria has led to a number of epidemics of pseudo-infection in patients undergoing bronchoscopy. (See also section on Pseudomonas.)

Recommendations   

• Hospital water supply: Endoscopy Unit water supply should be examined on at least a yearly basis, more frequently where there is evidence or recent plumbing alterations.
  
• Water filters: Water filters should be sent for microbiological examination when there are problems of machine filter contamination or known high contamination levels of the hospital water supply.
  
• Duodenoscopes and bronchoscopes: These should be monitored at least monthly.
  
• AFERs: AFERs should be monitored on a monthly basis or more frequently if there have been previous colonization problems.
  
• Gastroscopes and colonoscopes: These should be monitored 4-monthly depending on instrument age, previous positive culture results, or if other bacteriological surveillance (e.g. AFERs, water filters, etc.) has shown evidence of contamination.
  

Testing procedures   

Assessment should focus on the acceptability of the total number of organisms detected. Detailed taxonomic identification is not indicated except where microbiological failure persists after a rigorous review of compliance with cleaning and disinfection protocols and the structural soundness of the endoscope involved. Cultures should be directed to the detection of common enteric or respiratory organisms and organisms which are known to be associated with AFERs and filter contamination. The most important pathogens include Pseudomonas spp., Klebsiella spp., Proteus spp., E. coli, Salmonella spp., and atypical Mycobacterium spp. Consultation with a clinical microbiologist familiar with the endemic hospital pathogens is essential. Samples should be collected in an aseptic manner. The exact technique may vary with individual microbiological laboratories. In general, the most important sample collection from endoscopes is to fill the biopsy channel with sterile water, and brush it vigorously with a sterile channel cleaning brush which should then be agitated in sterile water in a specimen container. After brush removal, flush some sterile water down the channel. Brushing is not possible for all channels of some endoscopes. It is essential to ensure that specimens reach the laboratory without delay following collection.

Interpretation of cultures   

Workplace discussions reveal that one of the common but unvoiced reasons for resistance to bacteriological surveillance is the anxiety engendered by positive culture results. It has to be stressed that the finding of a few bacteria on bacteriological surveillance of an endoscope is not an infection transmission event and does not require patient infection detection protocols to be activated. It is essential to discuss positive culture findings with a clinical microbiologist and to look at the pattern of results from all endoscopes. Some interpretation examples include:

• A light growth of staphylococcus from a single instrument on a single occasion: This is almost certainly an environmental contaminant occurring during sample collection.
  
• Any growth of Pseudomonas spp. from a duodenoscope: This is cause for immediate withdrawal of that instrument. Full investigation of the possible source of contamination and activation of patient infection detection protocols for recent patients undergoing ERCP with that instrument.
  
• Light mixed growth of fecal organisms from different instruments over a period of time: Almost certainly reflects inadequate reprocessing procedures within the unit which need to be traced to either single or multiple staff members and corrective action taken.
  
• A heavy growth of salmonella from a colonoscope: Either inadequate reprocessing or internal instrument damage. The instrument should be withdrawn and carefully inspected by the manufacturer.
  
• Moderate growth of atypical mycobacteria from bronchoscope: Likely to be AFER or instrument accessory colonization. Clinicians should be notified that recently bronchoscoped patients may have specimens falsely interpreted as showing atypical mycobacterial infection.
  
• Growth of Mycobacterium tuberculosis from a bronchoscope: Immediate withdrawal of instrument. Full investigation of cause of contamination and activation of infection detection protocols for patients recently bronchoscoped with that instrument.
  

Microbiological surveillance of AFERs   

The method of sample collection for automatic disinfectors will vary depending upon the design of the individual machine. It is therefore appropriate to seek advice from the manufacturer and/or consult with the hospital clinical microbiologist. Common sense would suggest that the most appropriate point for machine sampling is the attachment of the machine to the endoscope. For machines with a single point of attachment (e.g. Medivator) this is relatively simple. Where there are multiple endoscope connections the process becomes more complicated. It is essential to know the design of the machine to determine which is the optimum part of the cycle in which to collect samples. In most cases this will be in the rinsing cycle.

Early detection of machine contamination is best effected by a concentration process. A sterile sealed filter (e.g. Millipore filter) is connected to the outlet of the machine where it normally attaches to the endoscope and a minimum of 200 ml of fluid is cycled through the filter in the rinse cycle mode. The disc can then be removed and plated directly. Since the principal contaminants of automatic disinfectors are Pseudomonas and related species and various forms of atypical mycobacteria, cultures should be particularly directed towards these organisms.

 Outstanding issues and future trends   

Serious infections associated with endoscopy are almost invariably due to failure to follow recognized guidelines for endoscope reprocessing or prophylactic antibiotic administration. Even minor deviations from accepted cleaning protocols results in persistent microbiological contamination of endoscopes after attempted high-level disinfection or sterilization. Present reprocessing techniques therefore have a lower margin of safety than is desirable. In the short term the emphasis must be to ensure that all staff involved in endoscope reprocessing have been adequately trained and assessed. Quality control systems must be fully implemented, including microbiological surveillance of all endoscopes. Endoscopes and AFER manufacturers and distributors must develop more responsible attitudes. There is a growing concern that companies fail to disseminate information on real or potential product defects. Restricting product warnings to known purchasers of individual instruments is quite inadequate, not least because such devices may have been on-sold or transferred to other hospitals with health care networks.

Currently there is widespread lack of recognition of the difficulties in providing rinse water of adequate quality. More effective education about the problems of biofilms and filter difficulties is urgently required.

AFER design and function continue to rapidly improve. However, few if any current machines can claim to have adequate self-sterilization cycles which involve all necessary parts of the machine, individual flow alarms on all channels, and reliable sterile water for rinsing cycles. Manufacturers clearly have ample scope for further improvement. The AFER manufacturer who can develop a machine with the above qualities combined with an automated adequate and reliable cleaning system will have instant command of the market.

The inherent complexity of endoscope design together with the clinical requirements for flexibility are often inadequately understood. The aim must be to develop endoscopes and accessories whose sterility can be guaranteed. Despite claims to the contrary we are still a long way from this goal.

 References  

 1 Favero, MS & Bond, WW. (1991) Chemical disinfection of medical and surgical materials, In: Disinfection, Sterilization and Preservation, 4th edn, (ed. Block, S. S.) pp. 617–41. Lea & Febiger, Philadelphia.

 2 Favero, MS. Sterility assurance: concepts for patient safety, In: Disinfection, Sterilization and Antisepsis: Principles and Practices in Healthcare Facilities2001, (ed. W. A. Rutla). pp. 16–27. APIC, Washington.

 3 Rutala, WA. APIC guidelines for selection and use of disinfectants. Am J Infect Control 1996; 24: 313–42. PubMed

 4 Bond, WW. Overview infection control problems. Gastroenterol Clin North Am 2000; 10 (2): 199–213.

 5 Webb, SF & Vall-Spinosa, A. Outbreak of Serratia marcescens associated with the flexible fiberbronchoscope. Chest 1975; 68: 703–8.

 6 Nicholson, G, Hudson, RA, Chadwick, MV & Gaya, H. The efficacy of the disinfection of bronchoscopes contaminated in vitro with Mycobacterium tuberculosis and Mycobacterium avium-intracellulare in sputum: a comparison of Sactimed-I-Sinald and glutaraldehyde. J Hosp Infect 1995; 29: 257–64. PubMed

 7 Beecham, HJ, Cohen, ML & Parkin, WE. Salmonella typhimurium: transmission by fiberoptic upper gastrointestinal endoscopy. JAMA 1979; 241: 1013–15. PubMed

 8 Wenzor, RP & Edmond, MB. Tuberculosis infection after bronchoscopy. JAMA 1997; 278 (13): 1111. PubMed

 9 Hanson, PJV, Chadwick, MV, Gaya, H & Collins, JV. A study of glutaraldehyde disinfection of fibreoptic bronchoscopes experimentally contaminated with Mycobacterium tuberculosis. J Hosp Infect 1992; 22: 137–42. PubMed

10 Gubler, JG, Salfinger, M & von Graevenitz, A. Pseudoepidemic of nontuberculosis mycobacteria due to a contaminated bronchoscope machine. Chest 1992; 101 (5): 1245–9. PubMed

11 Langenberg, W, Rauws, EA, Oudbier, JH & Tytgat, GN. Patient-to-patient transmission of Campylobacter pylori infection by fibreoptic gastro-duodenoscopy and biopsy. J Infect Dis 1990; 161: 507–11. PubMed

12 Classen, DC, Jacobson, JA, Burke, JP, Jacobson, JT & Scott Evans, R. Serious pseudomonas infections associated with endoscopic retrograde cholangiopancreatography. Am J Med 1988; 84: 590–6. PubMed

13 Seigman-Igra, Y, Isakov, A, Inbar, G & Cahaner, J. Pseudomonas aeruginosa septicemia following endoscopic retrograde cholangiopancreatography with a contaminated endoscope. Scand J Infect Dis 1987; 19: 527–30. PubMed

14 Kovacs, BJ, Aprecio, R & Kettering, JD et al. Efficacy of various disinfectants in killing a resistant strain of Pseudomonas Aeruginosa by comparing zones of inhibition: implications for endoscopic equipment reprocessing. Am J Gastroenterol 1998; 93: 2057–9. PubMed

15 Srinivasan, A, Wolfenden, LL, Song, X, Mackie, K, Hartsell, T & Jones, H et al. An outbreak of Pseudomonas aeruginosa infections associated with flexible bronchoscopes. New Engl J Med 2003; 348: 221–7. PubMed

16 Kirschke, DJ, Jones, TJ, Craig, AS, Chu, PS, Mayernick, GG, Patel, JA & Schaffner, W. Pseudomonas aeruginosa and Serratia marcescens contamination associated with a manufacturing defect in bronchoscopes. New Engl J Med 2003; 348: 214–19. PubMed

17 Vision Sciences Inc.'Vision Sciences Addresses Recent Bronchoscope Recall', press release dated March 11, 2002http://www.visionsciences.com/pressreleases/031102.html

18 http://www.myendosite.com.

19 Hanson, PJV, Gor, D, Jeffries, DJ & Collins, JV. Chemical inactivation of HIV on surfaces. Br Med J 1989; 298: 862–4.

20 Hanson, PJV, Gor, D & Clarke, JR et al. Contamination of endoscopes used in AIDS patients. Lancet 1989; 2: 86–8. PubMed

21 Hanson, PJV, Gor, D, Jeffries, DJ & Collins, JV. Elimination of high titre HIV from fibreoptic endoscopes. Gut 1990; 31: 657–9. PubMed

22 Deva, AK, Vickery, K, Zou, J, West, RH, Harris, JP & Cossart, YE. Establishment of an in use testing method for evaluating disinfection of surgical instruments using duck Hepatitis B model. J Hosp Infect 1996; 33: 119–30. PubMed

23 Birnie, GG, Quigley, EM, Clements, GB, Follet, EAC & Watkinson, G. Endoscopic transmission of hepatitis B virus. Gut 1983; 24: 171–4. PubMed

24 Lok, ASF, Lai, C-L & Hui, W-M et al. Absence of transmission of hepatitis B by fibreoptic upper gastrointestinal endoscopy. J Gastroenterol Hepatol 1987; 2: 175–80.

25 Tennenbaum, R, Colardelle, P & Chochon, M et al. Hepatite C apres cholangiographic retrograde. Gastroenterol Clin Biol 1993; 17: 763–4. PubMed

26 Bronowicki, JP, Venard, V & Botte, C et al. Patient-to-patient transmission of Hepatitis C virus during colonoscopy. New Engl J Med 1997; 337 (4): 237–40. PubMed

27 Muscarella, LF. Recommendations for preventing Hepatitis C virus infection: analyses of a Brooklyn endoscopy clinic outbreak. Infect Control Hosp Epidemiol 2001; 22 (11): 669. PubMed

28 Frieden, TR. 2002 Department of health letter. http://home.nyc.gov/html/doh/pdf/chi/ltr2-2002.pdf

29 Krisiunas, E. Forum posting: warning issued on reuse of needles. ekrisiunas@aol.com.

30 Crenn, P, Gigou, M & Passeron, J et al. Patient to patient transmission of Hepatitis C virus during gastroscopy on neuroleptanalgesia. Gastroenterology 1988; 114: 4, A1229.

31 Brown, P, Preece, M & Brandel, JP et al. Iatrogenic Creutzfeldt-Jakob disease at the millennium. Neurology 2000; 55: 1075–81. PubMed

32 Brown, P, Gibbs, C & Rodgers-Johnson, P et al. Human spongiform encephalopathy: the National Institutes of Health series of 300 cases of experimentally transmitted disease. Ann Neurol 1994; 35: 513–29. PubMed

33 Rutala, WA & Weber, DJ. Creutzfeldt-Jakob disease: recommendations for disinfection and sterilization. Healthcare Epidemiol CID 2001; 32: 1348–56.

34 Bruce, ME, Will, RG & Ironside, JW et al. Transmission to mice indicates that 'new variant' CJD is caused by the BSE agent. Nature 1997; 389: 498–501. PubMed

35 Bruce, ME, McConnell, I, Will, RG & Ironside, JW. Detection of variant Creutzfeldt-Jakob disease infectivity in extraneural tissues. Lancet 2001; 358 (9277): 208–9. PubMed

36 Wadsworth, JD & Joiner, S et al. Tissue distribution of protease resistant prion protein in variant Creutzfeldt-Jakob disease using a highly sensitive immunoblotting assay. Lancet 2001; 358 (9277): 171–80. PubMed

37 Stephenson, PM, Dorrington, L & Harris Od Rao, A. Bacteremia following oesophageal dilation and oesophago-gastroscopy. Aust N Z J Med 1977; 7: 32–5. PubMed

38 Cohen, FL, Koerner, RS & Taub, SJ. Solitary brain abscess following endoscopic injection sclerosis of oesophageal varices. Gastrointest Endosc 1985; 31: 331–3. PubMed

39 Tseng, C-C, Green, RM & Burke, SK et al. Bacteremia after endoscopic band ligation of oesophageal varices. Gastrointest Endosc 1992; 38: 336. PubMed

40 Struelens, MJ, Rost, F & Deplano, A et al. Pseudomonas aeruginosa and Enterobacteriaceae bacteremia after biliary endoscopy: an outbreak investigation using DNA macrorestriction analysis. Am J Med 1993; 95: 489. PubMed

41 Bass, DH & Oliver, S. Bornman PC. Pseudomonas septicemia after endoscopic retrograde cholangiopancreatography: an unresolved problem. S Afr Med J 1990; 77: 509. PubMed

42 Gossner, L, Keymling, J, Hahn, EG & Ell, C. Antibiotic prophylaxis in percutaneous endoscopic gastrostomy (PEG): a prospective randomised clinical trial. Endoscopy 1999; 31 (2): 119–24. PubMed

43 Dajani, AS, Taubert, KA & Wilson, W et al. Prevention of bacterial endocarditis: recommendations by the American Heart Association. Clin Infect Dis 1997; 25 (6): 1448–58. PubMed

44 British Society of Gastrointestinal Endoscopy. Antibiotic prophylaxis in gastrointestinal endoscopy. A report by a Working Party for the BSG. (2001). BSG Guidelines in Gastroenterology, pp. 1–10.

45 Meyer, GW & Artis, AL. Antibiotic prophylaxis for orthopedic prostheses and GI procedures: report of a survey. Am J Gastroenterol 1997; 92: 989–91. PubMed

46 Harris, A, Chan, ACH & Torres-Viera, C et al. Meta-analysis of antibiotic prophylaxis in endoscopic retrograde cholangiopancreatography (ERCP). Endoscopy 1999; 31 (9): 718–24. PubMed

47 Cowen, AE (1999) Infection Control in Endoscopy. Gastroenterological Society of Australia, Sydney.

48 Martin, MA & Reichelderfer, M. APIC guideline for infection prevention and control in flexible endoscopy. Am J Infect Control 1994; 22: 19–38. PubMed

49 Society of Gastroenterology Nurses and Associates. (1997) SGNA Monograph Series: Standards for Infection Control and Reprocessing of Flexible Gastrointestinal Endoscopes. pp. 1–13. SGNA, Chicago.

50 American Society for Gastrointestinal Endoscopy. Reprocessing of flexible gastrointestinal endoscopes. Gastrointest Endosc 1996; 43: 540–6. PubMed

51 American Society for Testing and Materials (1994). Standard Practice for Cleaning and Disinfection of Flexible Fibreoptic and Video Endoscopes Used in the Examination of Hollow Viscera. Designation F, pp. 518–94. ASTM, West Conshohocken, PA.

52 Muscarella, L. Déjà vu ... all over again? The importance of instrument drying. Letter to Editor. Custom Ultrasonics Inc, Ivyland, Pennsylvania. Infect Control Hosp Epidemiol 2000, 21 (10): 628–9. PubMed

53 Cooke, RPD, Whymant-Morris, A, Umasankar, RS & Goddard, SV. Bacteria-free water for automatic washer-disinfectors: an impossible dream? J Hospital Infect 1998; 39 (1): 63–5.

54 Hospital Infection Society & Public Health Laboratory Service. Rinse water for heat labile endoscopy equipment Report 26.6.01. Working parties/Rinse water for heat labile endoscopy equipment http://www.his.org.uk/

55 Raymond, JM et al. Evaluation des procédures de décontamination utilisées dans les centres d'endoscopie digestive de Gironde. Gastroenterol Clin Biol 1990; 14: 134–9. PubMed

56 Reynolds, CD, Rhinehart, E, Dreyer, P & Goldman, DA. Variability in reprocessing policies and procedures for flexible fiberoptic endoscopes in Massachusetts hospitals. Am J Infect Control 1992; 20: 283–90. PubMed

57 Cheung, RJ, Oritz, D & DiMarino, AJ Jr GI endoscopic reprocessing practices in the United States. Gastrointest Endosc 1999; 50 (3): 362–8. PubMed

58 Muscarella, LF. Current instrument reprocessing practices: results of a national survey. Gastroenterol Nurs 2001; 24 (5): 253–60. PubMed

59 Fraser, VJ, Jones, M & Murray, PR et al. Contamination of flexible fiberoptic bronchoscopes with Mycobacterium chelonae linked to an automated bronchoscope disinfection machine. Am Rev Respir Dis 1992; 145: 853. PubMed

60 Nosocomial infection and pseudoinfection from contaminated endoscopes and bronchoscopes – Wisconsin and Missouri. JAMA 1991; 266: 2197. PubMed

61 Alvarado, CJ, Stolz, SM & Maki, DC. Nosocomial infections from contaminated endoscopes: a flawed automated endoscope washer. An investigation using molecular epidemiology. Am J Med 1991; 91 (Suppl. 3B): 272S–280S. PubMed

62 Reichert, M. Automatic washers/disinfectors for flexible endoscopes. Infect Control Hosp Epidemiol 1991; 12: 497–9. PubMed

63 Ulatowski, TA. (2001) Re-use of single use devices.In: Disinfection, Sterilization and Antisepsis: Principles and Practices in Healthcare Facilities. (ed. W A. Rutala). pp. 16–27 APIC, Washington.

Copyright © Blackwell Publishing, 2004