A prospective study of the efficacy of routine decontamination for gastrointestinal endoscopes

A prospective study of the efficacyof routine decontamination for gastrointestinal endoscopes andthe risk factors for failureLinda Bisset,MSmed,a Yvonne E.Cossart,MBBS,DCP,FRCP,a Warwick Selby,MBBS,FRACP,MD,b Richard West, MBBS,FRCS,b Denise Catterson,RN,b Kate O’Hara,BMedSci(Hons),a and Karen Vickery,BVSc,MVSc,PhD a Sydney,New South Wales,AustraliaBackground:Patient-ready endoscopes were monitored over an80-week period to determine the efficacy of decontamination procedures in a busy endoscopy center.Decontamination failure was related to patient and procedural parameters.Methods:Samples from patient-ready endoscopes were cultured aerobically and anaerobically and subjected to polymerase chain reaction(PCR)to detect hepatitis B virus(HBV),hepatitis C virus(HCV),and HIV.PCR to detect coliforms from109culture negative washes was used as a surrogate marker for biofilm in endoscopes.PCR was used to detect the presence of Helicobactor pylori in endoscopes used on infected patients.Procedural information such as biopsy retrieval,endoscope number,diagnosis,attending personnel,and decontamination system procedures was collected.Results:Gastroscopes(n51376)and colonoscopes(n5987)were equally contaminated(1.8%vs1.9%,respectively)with low numbers of organisms commonly isolated from the nasopharynx and/or feces.Only1wash contained viral nucleic acid(HCV).There was a significant correlation(P,.001)between the number of times a patient-ready endoscope was contaminated and its frequency of use.Colonoscopes used on patients with gastrointestinal disease were significantly more likely to remain contam-inated through the decontamination process(P,.05).All other patient,staff,and decontamination system parameters remained not statistically significant.Coliform DNA was detected in40%of culture-negative washes collected from patient-ready endo-scopes,suggesting the presence of biofilm.No H pylori DNA was detected.Conclusion:Recommended decontamination procedures do not entirely eliminate persistence of low numbers of organisms on a few endoscopes,but this is unlikely to cause serious consequences in patients.Bacterial biofilm is difficult to remove and may explain this low-level persistence.(Am J Infect Control2006;34:274-80.)The risk of infection in patients undergoing surgical procedures has been progressively reduced by imple-mentation of increasingly stringent infection control policies.Endoscopy presents special problems because endoscopes cannot be steam sterilized,and surgical manipulation such as biopsy may provide tissue access for gutflora or bloodborne viruses.Endoscopy is an important tool for monitoring extremely ill patients,many of whom are immunocompromised and thus at increased risk from introduced pathogens.These pa-tients are also likely to harbor high loads of bacteria and viruses,which may lead to high-level contamina-tion on instruments.There are many case reports of bacterial infection after endoscopy,some of which report transmission of pathogenic organisms from patient to patient.1-3 Spach et al3reviewed scientific articles published between1966and1992,finding281episodes of nos-ocomial transmission of pathogens attributable to endoscopy.Since then,most published endoscopy-related nosocomial transmissions have been related to a lapse in endoscope-reprocessing protocols or gen-eral infection control practices such as reuse of syringes and have resulted in hepatitis C virus(HCV) being transmitted.4An increase in risk of pathogen transmission has been related to unacceptable clean-ing and disinfection,5,6failure to sterilize accessory equipment,7incorrect germicide use,8improper dry-ing,9or defective equipment.10Current infection control guidelines for endoscope reprocessing published by various organizations11-13 have been developed from general principles.ThereFrom the Department of Infectious Diseases and Immunology and The Australian Centre for Hepatitis Virology,The University of Sydney,a and Page5Endoscopy Suite,Royal Prince Alfred Hospital,b Sydney,NSW, Australia.Reprint requests:Karen Vickery,BVSc,MVSc,PhD,Department of Infectious Diseases and Immunology,Blackburn Building D06,Univer-sity of Sydney,Sydney,NSW,Australia2006.E-mail:kvickery@infdis. .au.Supported by NHMRC grant number9937934and a NHMRC Industry Fellowship(to K.V.).0196-6553/$32.00Copyrightª2006by the Association for Professionals in Infection Control and Epidemiology,Inc.doi:10.1016/j.ajic.2005.08.007274have been few direct attempts to assess the efficacy of these regimes,to validate recommendations about lab-oratory monitoring of decontamination procedures,or to pinpoint the factors leading to failure.The current study evaluated the survival of bacteria on endoscopes after routine cleaning and decontami-nation.This was related to patient and procedural parameters and to changes in unit staffing or pro-cessing.The reservoir of bloodborne infection in the patient cohort was assessed,and recovery of viral nu-cleic acid from washes of patient-ready endoscopes used on patients shown to be infected with a blood-borne virus was attempted.METHODSPatient recruitmentThis study had ethical approval from the University of Sydney and Royal Prince Alfred Hospital(RPAH)eth-ics committees.Each patient attending the endoscopy unit,during the80-week study period,was invited to participate,unless they were too ill to give consent or were unable to communicate with the study team in English.Of the2120patients enrolled,100(4.7%) were a HCV positive,45(2.2%)were positive for he-patitis B surface antigen(HBsAg),and17(0.8%)were positive for HIV.Only62were highly infectious for hepatitis B virus(HBV)or HCV as determined by being PCR positive.14No patient was known to be suffering from infectious gastroenteritis.Patient demographics,procedural information(in-cluding type of endoscopy,biopsy retrieval,individual instrument used,attending medical and nursing per-sonnel,the cleaner of the endoscope),and diagnoses were collected.Thefirst500patients were telephoned 1month after their procedure and asked about any complications related to the procedure. Endoscope decontaminationEndoscopes were washed and decontaminated in accordance with guidelines produced by the Gastro-enterological Society of Australia.13Decontamination involved a3-step process consisting of precleaning while the endoscope was still attached to the light source,the exterior of the insertion tube was wiped, and detergent was aspirated through the suction chan-nel;manual cleaning of the endoscope by immersion in an enzymatic detergent(generally Medizyme,Whiteley Industries,Sydney)and the exterior and interior sur-faces cleaned using appropriate brushes;followed by automated disinfection in a Medivator(Egan,Minne-sota)using2%gluteraldehyde for20minutes,followed byflushing with sterilefiltered water and then isopro-panol prior to forced air drying.Wash sample collection and evaluation Wash samples were obtained immediately,from the internal surface of the endoscope,after comple-tion of the decontamination cycle byflushing the endoscope biopsy channels with20-mL sterile diethyl pyrocarbonate(DEPC)-treated,phosphate-buffered saline(PBS),which was collected into a sterile con-tainer.These samples were stored at4°C for up to4 hours before culture then archived at220°C.The risk of bacterial contamination in these2377samples was related to patient and staff procedural parameters. MicrobiologyCulture.Wash samples(1-mL aliquots)were cul-tured aerobically in nutrient broth(Oxoid,Melbourne, Australia)and anaerobically in cooked meat broth (Oxoid,Melbourne,Australia).Cultures were incubated at37°C for48hours and then checked for bacterial growth.Positive cultures were sent to the RPAH Micro-biology Department for typing to species level by routine methods.Nucleic acid detection.Washes(10.5mL)collected from endoscopes used on14and48HBV and HCV viremic patients,respectively,were ultracentrifuged at130,000g at4°C for17hours,and the pellet was re-suspended in200m L of TE(10mmol/LTris,0.1mmol/L EDT A).The viral nucleic acid was extracted by in-house guanidinium methods,based on that published by Casas et al15for HBV DNA and that by Chomczynski and Sacchis16for extraction of HCV RNA.Both methods were adapted by the addition of a phenol,chloroform, isoamyl alcohol step.Extracted wash samples were tested for viral nucleic acids by PCR of the surface gene17for HBV and reverse transcription(RT)-PCR of the5’UTR for HCV.18Similarly,washes from randomly selected washes negative for microbial growth from56gastroscopes and53colonoscopes were centrifuged at12,000rpm at4°C for1hour,and the pellet was resuspended in 200m L TE.The Gentra Puregene DNA extraction kit (GeneWorks,Adelaide,Australia)was used to extract samples for bacterial DNA and tested for the presence of coliform bacteria by PCR amplification of the LacZ gene19and for Escherichia coli and Shigella species by amplification of the UidA gene.19,20Wash samples from gastroscopes used on58pa-tients with confirmed Helicobacter pylori infections were centrifuged as above,and the bacterial DNA was extracted using DNAzol(Invitrogen,Carlsbad,CA)ac-cording to manufacturer’s recommendations.Samples were subjected to nested PCR amplification using Li et al’s primers EHC-U and EHC-L21in thefirst round, followed by a second round of PCR using ourprimers Bisset et al June20062755’-GAG ACT TTC CT A GAA GCG GTG TT and5’-AAA AGA ACC GAA CGC AAC AG,which augmented an186-bp se-quence of the urease C gene.The reaction mixture con-sisted of2nmol/L of each primer;1.25U T aq;2mmol/L MgCl2;and0.2mmol/L(each)dA TP,dCTP,dTTP,and dGTP in a1X reaction mix.Cycling conditions for both rounds of PCR were95°C for5minutes,55°C for1minute,and72°C for1.5minutes,followed by 39cycles of95°C for30seconds,55°C for1minute, and72°C for1.5minutes but with afinal extension time of10minutes.PCR products were analyzed by electrophoresis on a2%agarose gel.This PCR had a sensitivity of60organisms.StatisticsThe x2analysis or Fisher exact test was used as ap-propriate to test for differences between groups using the SigmaStat(Jandel Scientific,San Rafael,CA)statis-tical program.Linear regression was used to evaluate the correlation between contamination rates of patient-ready endoscopes and the number of times an endoscope was used and the number of procedures performed by endoscopy staff.In all statistical methods employed,statistical significance was assumed for a P value of,.05.For all tests,the confidence level was at95%,and the a value for test sensitivity was set at.05.RESULTSOverall,79%of endoscopy patients attending the unit during the study period participated in the study;8.5%of patients were not asked to join the study be-cause of clinical contraindications,poor communica-tion,or unavailability of a member of the study team. Reasons given by the12.5%of patients who refused to join the study included needle phobia,sick of tests, too old or sick,lives too far away,and concern about impending procedure,but over one third of patients gave no reason.Approximately equal numbers of men(1077)and women(1043)took part in the study.The median age of patients was52.3years(range,16-92years).Patient-ready endoscope wash resultsBacterial culture.Wash samples were obtained im-mediately following decontamination from1376upper procedures and987lower procedures.Wash samples from patient-ready gastroscopes and colonoscopes were equally likely to grow bacteria,with a1.9%and 1.8%contamination rate,respectively(see T able1). Generally,the numbers of bacteria cultured were low (,10organisms/mL).More than1organism was iso-lated from some endoscopes.However,in all samples that grew Bacillus species,only Bacillus species were found.Because glutaraldehyde is not expected to kill all spore-forming organisms,removing these isolates reduced the rate of patient bacteria persisting through the decontamination process to1.45%and1.5%for pa-tient-ready gastroscopes and colonoscopes,respectively.Organisms isolated from upper endoscopes(T able2) were Staphylococcus epidermidis,Staphylococcus sapro-phyticus,methicillin-resistant Staphylococcus aureus (MRSA),Pseudomonas aeruginosa,Bacillus species, Escherichia coli,Enterococcus species,diptheroids, Enterobacter aerogenes,and Serratia marcescens.The organisms isolated from lower endoscopes were Acinetobacter lwoffii,E coli,Enterococcus faecalis, Enterobacter cloacae,Staphylococcus epidermidis,he-molytic Streptococci,Bacillus species,Pseudomonas aeruginosa,Pseudomonasfluorescens,and coliforms.Nucleic acid testing.Bacterial DNA was detected by PCR in44of109(40%)of the culture negative washes collected from patient-ready endoscopes.However, H pylori DNA was not detected on any of the58wash samples obtained from gastroscopes used on con-firmed H pylori patients.There were14and48wash samples obtained from patient-ready endoscopes that had previously been used on HBV and HCV viremic patients,respectively. None of the endoscopes was positive for HBV DNA, and only1sample was positive for HCV RNA by PCR. This sample was from a gastroscope that had been used on a HCV-positive male patient who was not biop-sied and had no evidence of gastrointestinal disease. None of the500patients telephoned after their endos-copy had any complications that could be related to the procedure.Investigation of possible causes of decontamination breakdownPatient parameters.Endoscope decontamination failure following upper procedures was unrelated to patient gastrointestinal disease(P5.960),bloodborne viral status(P5.575),or sex(P5.814)or whether or not they were biopsied(P5.867)(see T able1).Similarly,decontamination failure of colonoscopes was unrelated to patient bloodborne viral status(P5 .613)or sex(P5.252)or whether or not the patient was biopsied or not(P5.594).In contrast,endoscopes used on those patients with evidence of lower gastro-intestinal disease were significantly more likely to remain contaminated through the decontamination process than colonoscopes used on patients with no evidence of gastrointestinal disease(P#.05)(see T able1).Instrument parameters.The overall contamination rate of patient-ready gastroscopes was1.9%,whereas276Vol.34No.5Bisset etalcontamination rates for individual gastroscopes varied between0%and3%.Linear regression analysis pro-duced a straight-line graph,indicating(R50.853, P,.001,F550.653)that the number of times an individual endoscope was contaminated was directly proportional to the number of occasions that the in-strument was used(see Fig1).Similarly,for colono-scopes,as the frequency of use increased,so did the number of times the individual endoscope was con-taminated(R50.945,P,.001,F5133.080). Endoscopy personnelThe number of procedures performed by individual endoscopists during the study period ranged from 1to763(T able3).The overall contamination rate of patient-ready endoscopes was1.86%,whereas con-tamination rates of patient-ready endoscopes follow-ing use by an endoscopist varied between0%and 8%.These differences were not statistically significant.Nurse assistants can have a major effect on decon-tamination efficacy if they fail to follow standard proto-cols.The study team evaluated cleaning protocols and found them to be followed rigidly.All the nurse assistants were fully trained,although some only worked on a casual basis and hence were not responsi-ble for cleaning many endoscopes.The number of en-doscopes cleaned by individual nurse assistants during the study period ranged from1to475.The differences in the percentage(0%-3.3%)of patient-ready endo-scopes remaining contaminated between nurses were not statistically significant(T able3). Decontamination procedureThe same decontamination protocol was followed throughout the study period,except for a1-week pe-riod during which,on the advice from the endoscope manufacturer,the endoscope buttons were disinfected, attached to the scope.During this period,there was a sharp but statistically nonsignificant increase in the number of contaminated patient-ready endoscopes (see T able4).Decontamination systemWaterfilters were changed when a loss of pressure occurred,on average,every15.2days or every thirdT able1.Contamination rates and percentage of patient-ready endoscopes related to previous patient use and demographic and procedural informationContaminationrateBiohazardous statusAbnormalreportNormalreport Carrier Noncarrier Male Female Biopsy No biopsyGastroscopes,n26/1376(1.9%)1/1220.8%25/12512.0%11/6931.6%15/6832.2%20/10501.9%6/3261.8%19/9732.0%7/4031.7% Colonoscopes,n18/987(1.8%)0/540%18/9331.9%6/4891.2%12/4982.4%9/4062.2%9/5811.5%14/5032.8%4/4840.8%T able2.Bacterial species isolated from patient-ready endoscopes*Number of isolations Bacteria isolated Gastroscopes Colonoscopes Staphlococcus epidermidis94 Staphlococcus saprophyticus1Staphlococcus aureus2yPseudomonas aeruginosa31 Pseudomonasfluorescens1 Bacillus species63 Escherichia coli23 Enterococcus species23 Enterobacter aerogenes1Enterobacter cloacae2 Micrococcus luteus1Serratia marcescens1Diptheroids1Acinetobacter lwoffii1 Hemolytic Streptococci2*Some endoscopes were contaminated with more than1organism(see text).y OneMRSA.Fig1.Linear regression analysis for gastroscopes, showing99%and95%confidence levels and prediction of contamination incidence numbers against total washes collected pergastroscope. Bisset et al June2006277working week.T o prevent dilution and hence loss of efficacy,the glutaraldehyde was replaced every 40cycles,which occurred,on average,every 12.1days or every second working week.There was no relation-ship between contamination of patient-ready endo-scopes and the life cycle of either the water filters or glutaraldehyde.DISCUSSIONThere has been considerable discussion regarding the clinical value of routine bacteriologic monitoring of endoscopes.Current Australian practice is for duo-denoscopes to be sampled monthly,colonoscopes and gastroscopes every 4months,and monthly sam-pling of the automated decontamination system.22In our study,we sampled immediately following each de-contamination cycle and inoculated 1mL wash sample into fluid cultures,and the methods were substantially more sensitive than the commonly recommendedtesting of small volumes of wash fluid on agar plates.17Even so,we found that,under conditions of careful adherence to working guidelines,there were very few instances in which we detected viable organisms in washes from endoscopes at the point of ing a contamination level of 2%and the recommended level of monitoring,it would take 1.8years to detect 1posi-tive wash sample in an endoscopy center possessing 10endoscopes.Even extrapolated to 20%contamination,there would still only be 1or 2positive samples occur-ring in a 4-month sampling period.These findings sug-gest that plate cultures performed on an infrequent schedule are unlikely to contribute to improved infec-tion control in endoscope units.Our samples were obtained from the endoscope’s internal surface immediately following decontamina-tion.Contamination of the samples by personnel is very unlikely because it is impossible to touch the area of the endoscope we sampled.All water to the automatic reprocessors was filter sterilized,thus pre-venting contamination of the clean endoscope with waterborne bacteria.The organisms isolated from both gastroscopes and colonoscopes are commonly isolated from the nasopharynx and/or the feces.In most instances,only a single organism was grown rather than a mixture,as expected from gastrointestinal contents.However,1gastroscope,which had been used on a patient who was currently infected with MRSA,the same strain of Staphylococcus aureus ,as well as Pseudo-monas aeruginosa and enterococci were recovered,and,in another wash,both Enterobacter and enterococci were isolated.Three colonoscopes also yielded 2or-ganisms (Enterobacter cloacae /Enterococcus species,Enterobacter cloacae/Pseudomonas aeruginosa ,andT able 3.Contamination rate of patient-ready endoscopes related to the endoscopist who performed the procedure and the nurse assistant who washed the instrument *Endoscopist Contaminationrate (%)T otal number of proceduresNurse assistantContamination rate(%)T otal number of washes18(1.7%)48310107203320931(4.8%)2138(3%)26540784013657(3.4%)2075032602866(2.2%)27978(4.9%)16475(2.5%)20380981(1%)10193(1.9%)16199(1.9%)4751012(1.6%)763106(1.9%)323112(1.4%)142119(3.3%)2701209612015131(1.3%)78130571407514062152(8%)251528*Endoscopists and nurse assistants numbers 1-14participated in procedures on a regular basis,whereas endoscopist or nurse assistant number 15represents a group of nonregular participants with only 1or 2procedures each.T able 4.Contamination rates of patient-readyendoscopes during a period of change in reprocessing protocol comparing disinfection of the endoscope buttons attached to or separated from the endoscopeWeek ContaminationNumber of scopesButton protocol 43179Separate 44064Separate 45582Attached to endoscope 46078Separate 47082Separate 4860Separate278Vol.34No.5Bisset etalE coli/hemolytic Streptococcus).The spectrum of orga-nisms recovered probably reflects both their ability to survive the decontamination procedure and selection by our culture methods.An important limitation of the latter was the time the washes were refrigerated under aerobic conditions before being cultured,which would have eliminated strict anaerobes.Nine of the 44positive washes grew Bacillus species,which is acceptable when instruments are disinfected and not sterilized.Removal of these washes reduced the rate of patient bacteria persisting through the decontami-nation process to1.5%or less.Biofilm has been shown to be present on channels of endoscopes sent for servicing,23and its practical effect on decontamination efficacy is unknown.Direct study of biofilm on endoscope channels was not prac-ticable in this clinical setting,so PCR detection of coli-form bacteria DNA was chosen as a surrogate marker for biofilm presence in culture-negative washes.A random selection of109culture-negative washes was centrifuged,extracted,and PCR amplified to observe for the presence of coliform bacteria.Coliform bacteria were chosen because they are likely to be derived from patients rather than the instrument room environment. Forty percent(22/56gastroscopes and21/53colono-scopes)were positive.PCR detection of nucleic acids does not necessarily indicate the presence of viable organisms,but organisms from biofilms may be very difficult to revive in planktonic culture,suggesting that further investigation of this issue is warranted. Only sterile water was used in the Medivator,with the employed waterfiltration system preventing orga-nisms from water pipe biofilms reaching the washing area and the automated reprocessor;therefore,bacte-rial DNA found in culture negative washes is likely to originate from the endoscope or the tubing of the auto-mated reprocessor.Overall,ourfindings suggest that carryover of orga-nisms on endoscopes after routine decontamination can be demonstrated at very low frequency and at low concentrations.There are unlikely to be serious consequences for individual patients from this carry-over of organisms from the normal gastrointestinal flora.Indeed,the500patients interviewed1month af-ter their procedures failed to recognize any complica-tions related to their procedures.We have also shown that many culture-negative washes contain bacterial nucleic acid,which we interpret as putative evidence of bacterial biofilm buildup on the surface of the endo-scope channels.The indications for endoscopy and the actual condi-tions diagnosed varied very widely in the study group. There was a slight increase in the likelihood of failure of the decontamination process for colonoscopes used on patients with an abnormal report.In contrast to epidemiologic evidence suggesting that biopsy during endoscopy was associated with an increased risk for HCV infection,24biopsy did not increase the rate of either viral nucleic acid or viable bacteria persisting through the decontamination cycle.Carryover of viral nucleic acid was found on only1of the62endoscopes that had been used to examine HBV and HCV viremic patients.This negative evidence is reassuring,and it should be noted that the positive PCR may well have been derived from inactivated rather than viable virus.25 Contamination of patient-ready endoscopes may also be attributed to inadequate maintenance of the de-contamination system.Adequate maintenance of the waterfiltration unit is necessary to prevent this.Reuse of glutaraldehyde results in its dilution and hence its activity with each cycle.The maintenance of thefilter system and changing of glutaraldehyde solutions were satisfactory throughout the study period.A change in procedure to decontamination of endoscope buttons attached to the endoscope resulted in a cluster of cul-ture-positive results.No clinically untoward conse-quences were observed,but thefinding shows the desirability of culture monitoring during changes in protocols to confirm satisfactory in use performance.Individual staff members carry an onerous responsi-bility for the maintenance of infection control in hos-pital practice.Endoscopy is particularly vulnerable because of the complexity of protocols for decon-tamination of expensive and fragile,heat-sensitive in-struments,which must be performed under time pressures in a ward rather than in centralized facilities. The study showed excellent adherence to protocols and a correspondingly high level of microbiologic safety in the unit.The authors thank the staff and patients of the Page5Endoscopy Suite,Royal Prince Alfred Hospital,Sydney,because,without their help and collaboration,this study could not have taken place and the Microbiology Department of the Royal Prince Alfred Hospital,Sydney,for identifying the bacterial species isolated. References1.Moayyedi P,Lynch D,Axon A.Pseudomonas and endoscopy.Endos-copy1994;26:554-8.2.Struelens MJ,Rost F,Deplano A,Maas A,Schwam V,Serruys E,et al.Pseudomonas aeruginosa and Enterobacteriaceae bacteremia after biliary endoscopy:an outbreak investigation using DNA macrorestriction analysis.Am J Med1993;95:489-98.3.Spach DH,Silverstein FE,Stamm WE.T ransmission of infection bygastrointestinal endoscopy and bronchoscopy.Ann Intern Med1993;118:117-28.4.Nelson DB,Eisen G.Glutaraldehyde-based formulations.GastrointestEndosc2000;51:378.5.Bronowicki JP,Venard V,Botte´C,Monhoven N,Gastin I,Chone´L,et al.Patient-to-patient transmission of hepatitis C virus during colon-oscopy.N Engl J Med1997;337:237-40.6.Cryan EMJ,Falkiner FR,Mulvihill TE,Keane CT,Keeling PWN.Pseudo-monas aeruginosa cross-infection following endoscopic retrograde cholangiopancreatography.J Hosp Infect1984;5:371-6.Bisset et al June20062797.Lo Passo C,Pernice I,Celeste A,Perdichizzi G,T odaro-Luck F .T rans-mission of Trichosporon asahii esophagitis by a contaminated endo-scope.Mycoses 2001;44:13-21.8.Weber DJ,Rutala WA,DiMarino AJ Jr.The prevention of infection fol-lowing gastrointestinal endoscopy:the importance of prophylaxis and reprocessing.In:DiMarino AJ Jr,Benjamin SB,editors.Gastrointestinal diseases:an endoscopic approach.Thorofare,NJ:Slack Inc;2002.p.87-106.9.Noy MF ,Harrison L,Holmes GKT ,Cockel R.The significance of bacte-rial contamination of fiberoptic endoscopes.J Hosp Infect 1980;1:53-61.10.Coney S.Patients recalled after endoscopic ncet1999;354:578.11.Alvarado CJ,Reichelderfer M.APIC guidelines for infection preventionand control in flexible endoscopy.Am J Infect Control 2000;28:138-55.12.SGNA.Standards of infection control in reprocessing of flexiblegastrointestinal endoscopes.Gastroenterol Nurs 2000;23:172-87.13.Gastroenterological Society and Gastroenterological Nurses Societyof Australia.Guidelines:infection and endoscopy,3rd edition; .14.Vickery K,Bisset L,Selby W ,West R,Catterson D,Cossart YE.Bloodborne virus transmission during endoscopy–viral prevalence or decon-tamination breakdown?In:Jilbert AR,Grgacic EVL,Vickery K,Burrell C,Cossart YE,editors.Proceedings 11th International Symposium on Viral Hepatitis and Liver Disease.Melbourne:The Australian Centre for Hepatitis Virology;2004.p.344-6.15.Casas I,Powell PE,Klapper PE,Cleator GM.New method for the ex-traction of viral RNA and DNA from cerebrospinal fluid for use in the polymerase chain reaction assay.J Virol Methods 1995;53:25-36.16.Chromczynski P ,Sacchi N.Single-step method of RNA isolation byacid guanidinium thiocyanate-phenol-chloroform extraction.Anal Bio-chem 1987;162:156-9.17.Deva AK,Vickery K,Zou J,West RH,Selby W ,Benn RAV ,et al.Detection of persistent vegetative bacteria and amplified viral nucleic acid from in-use testing of gastrointestinal endoscopes.J Hosp Infect 1998;39:149-57.18.McGuinness P ,Bishop GA,Liem A,Wiley B,Parsons C,McCaughanGW .Detection of serum hepatitis C virus RNA in HCV antibody/seropositive volunteer blood donors.Hepatology 1993;18:485-90.19.Bej AK,DiCesare JL,Haff L,Atlas RM.Detection of Escherichia coli andShigella spp.in water by using the polymerase chain reaction and gene probes for uid .Appl Environ Microbiol 1991;57:1013-7.20.Fricker EJ,Fricker CR.Application of the polymerase chain reaction tothe identification of Escherichia coli and coliforms in water.Lett App Microbiol 1994;19:44-6.21.Li C,Ha T ,Ferguson D,Chi DS,Zhao R,Patel NR,et al.A newlydeveloped PCR assay of H.pylori in gastric biopsy,saliva,and feces:evidence of high prevalence of H.pylori in saliva supports oral trans-mission.Dig Dis Sci 1996;41:2142-9.22.Anonymous.Cleaning and disinfection of equipment for gastrointesti-nal endoscopy:report of a working party of the British Society of Gastroenterology.Endoscopy committee.Gut 1998;42:585-93.23.Pajkos A,Vickery K,Cossart YE.Is biofilm accumulation on endo-scope tubing a contributer to failure of cleaning and decontamination?J Hosp Infect 2004;58:224-9.24.Andrieu J,Barny S,Colardelle P ,Maisonneuve P ,Giraud V ,Robin E,et al.Prevalance and risk factors for hepatitis C infection in a hospital-ised population in gastroenterology:role of perendoscopic biopsies.Gastroenterol Clin Biol 1995;19:340-5.25.Deva AK,Vickery K,Zou J,West RH,Harris JP ,Cossart YE.Establish-ment of an in-use testing method for evaluating disinfection of surgical instruments utilising the duck hepatitis B model.J Hosp Infect 1996;33:119-30.280Vol.34No.5Bisset etal。