Assessment of postmanual cleaning adenosine triphosphate tests to prevent the use of contaminated duodenoscopes and linear echoendoscopes: the DETECT study

Open AccessPublished:March 24, 2022DOI:https://doi.org/10.1016/j.gie.2022.03.022

      Background and Aims

      We investigated whether the use of postmanual cleaning adenosine triphosphate (ATP) tests lowers the number of duodenoscopes and linear echoendoscopes (DLEs) contaminated with gut flora.

      Methods

      In this single-center before-and-after study, DLEs were ATP tested after cleaning. During the control period, participants were blinded to ATP results: ATP-positive DLEs were not recleaned. During the intervention period, ATP-positive DLEs were recleaned. DLEs underwent microbiologic sampling after high-level disinfection (HLD) with participants blinded to culture results.

      Results

      Using 15 endoscopes of 5 different DLE types, we included 909 procedures (52% duodenoscopes, 48% linear echoendoscopes). During the intervention period, the absolute rate of contamination with gut flora was higher (16% vs 21%). The main analysis showed that contamination was less likely to occur in the intervention period (odds ratio, .32; 95% credible interval [CI], .12-.85). A secondary analysis showed that this effect was based on 1 particular duodenoscope type (estimated probability, 39% [95% CI, 18%-64%] vs 9% [95% CI, 2%-21%]), whereas no effect was seen in the other 4 DLE types. In detail, of the 4 duodenoscopes of this type, 2 had lower contamination rates (69% vs 39% and 36% vs 10%). During the control period, both these duodenoscopes had multiple episodes with ongoing contamination with the same microorganism that ended weeks before the start of the intervention period (ie, they were not terminated by ATP testing).

      Conclusions

      Postmanual cleaning ATP tests do not reduce post-HLD gut flora contamination rates of DLEs. Hence, postcleaning ATP tests are not suited as a means for quality control of endoscope reprocessing.

      Abbreviations:

      AM20 (any microorganism with ≥20 colony-forming units per 20 mL), ATP (adenosine triphosphate), CFU (colony-forming unit), DLE (duodenoscope and linear echoendoscope), HLD (high-level disinfection), RLU (relative light unit)
      In recent years, multiple endoscope-associated infectious outbreaks have been reported, in particular related to duodenoscopes used for ERCP.

      U.S. Senate. Preventable tragedies: superbugs and how ineffective monitoring of medical device safety fails patients [U.S. Senate–Minority Staff Report]. Washington, DC, 2016 [updated 13-01-2016]. Available at: https://www.help.senate.gov/imo/media/doc/Duodenoscope%20Investigation%20FINAL%20Report.pdf. Accessed January 30, 2021.

      • Rubin Z.A.
      • Kim S.
      • Thaker A.M.
      • et al.
      Safely reprocessing duodenoscopes: current evidence and future directions.
      • Balan G.G.
      • Sfarti C.V.
      • Chiriac S.A.
      • et al.
      Duodenoscope-associated infections: a review.
      Many of these were outbreaks with multidrug-resistant organisms. In several of these outbreaks, it was shown that duodenoscopes were already contaminated for months before transmission was discovered with attack rates (ie, number of infected or colonized cases/number of exposed persons) up to 40%.
      • Carbonne A.
      • Thiolet J.M.
      • Fournier S.
      • et al.
      Control of a multi-hospital outbreak of KPC-producing Klebsiella pneumoniae type 2 in France, September to October 2009.
      • Valderrama S.
      • Miranda C.J.L.
      • Rubio A.P.G.
      • et al.
      Successful control of an endoscopic retrograde cholangiopancreatography–associated nosocomial outbreak caused by Klebsiella pneumoniae carbapenemase producing Klebsiella pneumoniae in a University Hospital in Bogota, Colombia.
      • Verfaillie C.J.
      • Bruno M.J.
      • Voor in 't Holt A.F.
      • et al.
      Withdrawal of a novel-design duodenoscope ends outbreak of a VIM-2-producing Pseudomonas aeruginosa.
      • Kola A.
      • Piening B.
      • Pape U.F.
      • et al.
      An outbreak of carbapenem-resistant OXA-48-producing Klebsiella pneumonia associated to duodenoscopy.
      • Alrabaa S.F.
      • Nguyen P.
      • Sanderson R.
      • et al.
      Early identification and control of carbapenemase-producing Klebsiella pneumoniae, originating from contaminated endoscopic equipment.
      Because outbreaks are shown to be under-reported,

      U.S. Senate. Preventable tragedies: superbugs and how ineffective monitoring of medical device safety fails patients [U.S. Senate–Minority Staff Report]. Washington, DC, 2016 [updated 13-01-2016]. Available at: https://www.help.senate.gov/imo/media/doc/Duodenoscope%20Investigation%20FINAL%20Report.pdf. Accessed January 30, 2021.

      ,
      • Rubin Z.A.
      • Kim S.
      • Thaker A.M.
      • et al.
      Safely reprocessing duodenoscopes: current evidence and future directions.
      ,
      • García-Cano J.
      • Perez-Miranda M.
      • Enric B.
      • et al.
      Duodenoscope-transmitted infection: results from a national Spanish survey [abstract].
      the published number of 490 patient infections and 32 deaths because of contaminated duodenoscopes between 2008 and 2018 is considered an underestimation.

      U.S. Senate. Preventable tragedies: superbugs and how ineffective monitoring of medical device safety fails patients [U.S. Senate–Minority Staff Report]. Washington, DC, 2016 [updated 13-01-2016]. Available at: https://www.help.senate.gov/imo/media/doc/Duodenoscope%20Investigation%20FINAL%20Report.pdf. Accessed January 30, 2021.

      ,
      • Balan G.G.
      • Sfarti C.V.
      • Chiriac S.A.
      • et al.
      Duodenoscope-associated infections: a review.
      After each procedure, endoscopes are reprocessed. This process, which includes flushing, manual cleaning, high-level disinfection (HLD), and drying, has little margin for error.
      • Rutala W.A.
      • Weber D.J.
      Outbreaks of carbapenem-resistant Enterobacteriaceae infections associated with duodenoscopes: What can we do to prevent infections?.
      • Roberts C.G.
      Studies on the bioburden on medical devices and the importance of cleaning. Disinfection, sterilization and antisepsis: principles and practices in healthcare facilities.
      • Alfa M.J.
      • Degagne P.
      • Olson N.
      Worst-case soiling levels for patient-used flexible endoscopes before and after cleaning.
      • Rutala W.A.
      • Weber D.J.
      FDA labeling requirements for disinfection of endoscopes: a counterpoint.
      Reprocessing nonetheless has been shown to be error prone.
      • Ofstead C.L.
      • Wetzler H.P.
      • Snyder A.K.
      • et al.
      Endoscope reprocessing methods: a prospective study on the impact of human factors and automation.
      • Dirlam Langlay A.M.
      • Ofstead C.L.
      • Mueller N.J.
      • et al.
      Reported gastrointestinal endoscope reprocessing lapses: the tip of the iceberg.
      • Schaefer M.K.
      • Jhung M.
      • Dahl M.
      • et al.
      Infection control assessment of ambulatory surgical centers.
      • Armellino D.
      • Cifu K.
      • Wallace M.
      • et al.
      Implementation of remote video auditing with feedback and compliance for manual-cleaning protocols of endoscopic retrograde cholangiopancreatography endoscopes.
      Duodenoscopes and linear echoendoscopes (DLEs) are more difficult to clean than other endoscopes because of their complex design, which includes a side-viewing tip, forceps elevator, and elevator wire channel. Even when manufacturers’ instructions for use were strictly followed, large outbreaks have occurred.
      • Epstein L.
      • Hunter J.C.
      • Arwady M.A.
      • et al.
      New Delhi metallo-beta-lactamase-producing carbapenem-resistant Escherichia coli associated with exposure to duodenoscopes.
      ,
      • Wendorf K.A.
      • Kay M.
      • Baliga C.
      • et al.
      Endoscopic retrograde cholangiopancreatography-associated AmpC Escherichia coli outbreak.
      Because current reprocessing techniques do not achieve reliable decontamination, patients undergoing ERCP are regularly exposed to contaminated duodenoscopes. Duodenoscope manufacturers’ postmarket studies show a 5.4% contamination rate with indicator microorganisms associated with disease transmission (ie, Escherichia coli, Pseudomonas aeruginosa),

      U.S. Food and Drug Administration. The FDA continues to remind facilities of the importance of following duodenoscope reprocessing instructions: FDA Safety Communication 2019 [updated 12-04-2019]. Available at: https://www.fda.gov/medical-devices/safety-communications/fda-continues-remind-facilities-importance-following-duodenoscope-reprocessing-instructions-fda. Accessed January 30, 2021.

      which is in line with the 8% contamination prevalence of these microorganisms found in Dutch prevalence studies.
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      Patients undergoing EUS may also be exposed to contaminated equipment, as indicator microorganism contamination rates of linear echoendoscopes range between 1.1% and 8%.
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      • Brandabur J.J.
      • Leggett J.E.
      • Wang L.
      • et al.
      Surveillance of guideline practices for duodenoscope and linear echoendoscope reprocessing in a large healthcare system.
      • Chapman C.G.
      • Siddiqui U.D.
      • Manzano M.
      • et al.
      Risk of infection transmission in curvilinear array echoendoscopes: results of a prospective reprocessing and culture registry.
      A systematic review showed an overall duodenoscope contamination rate of 15%.
      • Larsen S.
      • Russell R.V.
      • Ockert L.K.
      • et al.
      Rate and impact of duodenoscope contamination: a systematic review and meta-analysis.
      In Dutch hospitals the DLE contamination prevalence with microorganisms from GI and oral origin was 15% for 2 consecutive years.
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      ,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      High prevalence rate of digestive tract bacteria in duodenoscopes: a nationwide study.
      Gastroenterology, microbiology, and regulatory agencies have stressed the need for easy to implement and effective control measures to check endoscope decontamination.
      • Petersen B.T.
      • Cohen J.
      • Hambrick R.D. 3rd
      • et al.
      Reprocessing Guideline Task Force
      Multisociety guideline on reprocessing flexible GI endoscopes: 2016 update.
      ,

      Centers for Disease Control and Prevention. Duodenoscope surveillance sampling & culturing: reducing the risks of infection 2018. Available at: https://www.fda.gov/downloads/MedicalDevices/ProductsandMedicalProcedures/ReprocessingofReusableMedicalDevices/UCM597949.pdf. Accessed January 30, 2021.

      Current guidelines advise microbial surveillance to prevent transmission of microorganisms by contaminated endoscopes.

      Centers for Disease Control and Prevention. Duodenoscope surveillance sampling & culturing: reducing the risks of infection 2018. Available at: https://www.fda.gov/downloads/MedicalDevices/ProductsandMedicalProcedures/ReprocessingofReusableMedicalDevices/UCM597949.pdf. Accessed January 30, 2021.

      ,
      • Beilenhoff U.
      • Biering H.
      • Blum R.
      • et al.
      Prevention of multidrug-resistant infections from contaminated duodenoscopes: position statement of the European Society of Gastrointestinal Endoscopy (ESGE) and European Society of Gastroenterology Nurses and Associates (ESGENA).
      However, culturing is labor intensive and expensive and gives delayed feedback because of the 48- to 72-hour incubation period. Real-time audits of “bacterial load” during reprocessing could drastically improve endoscope safety. The most often proposed alternative is the use of the adenosine triphosphate (ATP) assay, a test originating from the food industry and recently introduced into hospitals to monitor cleanliness of the environment. The ATP test is a bioluminescence assay using luciferase-catalyzed oxidation of luciferin causing ATP-dependent emission of light, measured in relative light units (RLUs). Bacteria and human cells and secretions contain ATP, and the presence of ATP after cleaning may indicate organic residue requiring recleaning.
      • Day L.W.
      • Kwok K.
      • Visrodia K.
      • et al.
      American Society for Gastrointestinal Endoscopy Infection Control Summit: updates, challenges, and the future of infection control in GI endoscopy.
      It is unclear whether the use of ATP tests to guide reprocessing prevents the use of contaminated endoscopes. ATP was not found to be useful for testing after HLD because of the poor correlation with terminal reprocessing cultures.
      • Batailler P.
      • Saviuc P.
      • Picot-Gueraud R.
      • et al.
      Usefulness of adenosine triphosphate bioluminescence assay (ATPmetry) for monitoring the reprocessing of endoscopes.
      • Hansen D.
      • Benner D.
      • Hilgenhöner M.
      • et al.
      ATP measurement as method to monitor the quality of reprocessing flexible endoscopes.
      • Olafsdottir L.B.
      • Wright S.B.
      • Smithey A.
      • et al.
      Adenosine triphosphate quantification correlates poorly with microbial contamination of duodenoscopes.
      • Olafsdottir L.B.
      • Whelan J.
      • Snyder G.M.
      A systematic review of adenosine triphosphate as a surrogate for bacterial contamination of duodenoscopes used for endoscopic retrograde cholangiopancreatography.
      The correlation with bacterial growth after cleaning is also questionable.
      • Washburn R.E.
      • Pietsch J.J.
      Assessment of test methods for evaluating effectiveness of cleaning flexible endoscopes.
      ,
      • Visrodia K.
      • Hanada Y.
      • Pennington K.M.
      • et al.
      Duodenoscope reprocessing surveillance with adenosine triphosphate testing and terminal cultures: a clinical pilot study.
      However, with a validated cutoff of 200 RLUs to differentiate for acceptable levels of protein and bacteria after cleaning,
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      Validation of adenosine triphosphate to audit manual cleaning of flexible endoscope channels.
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      The adenosine triphosphate test is a rapid and reliable audit tool to assess manual cleaning adequacy of flexible endoscope channels.
      • Alfa M.J.
      • Olson N.
      • Murray B.-L.
      Comparison of clinically relevant benchmarks and channel sampling methods used to assess manual cleaning compliance for flexible gastrointestinal endoscopes.
      it has been suggested as a quality control indicator for manual cleaning.
      • Visrodia K.
      • Hanada Y.
      • Pennington K.M.
      • et al.
      Duodenoscope reprocessing surveillance with adenosine triphosphate testing and terminal cultures: a clinical pilot study.
      ,
      • Alfa M.J.
      Quality systems approach for endoscope reprocessing: you don't know what you don't know!.
      Because a growing number of endoscopy centers use ATP tests,
      • Kenters N.
      • Tartari E.
      • Hopman J.
      • et al.
      Worldwide practices on flexible endoscope reprocessing.
      it is essential to scientifically establish the usefulness and clinical value of ATP. Therefore, the aim of this study was to assess whether postcleaning ATP tests, with recleaning in the event of a positive ATP test, lower the number of clinically relevant positive cultures of DLEs.

      Methods

      From July 2017 to October 2018, we conducted a prospective single-center intervention study in a before-and-after design to assess the utility of ATP testing after manual cleaning. The study was conducted at the 1320-bed tertiary Erasmus MC University Medical Centre in Rotterdam, the Netherlands, which performs approximately 900 ERCP and 750 EUS procedures yearly.

      Endoscopes and reprocessing

      Multiple DLEs were used throughout the study. Four ED34-i10T duodenoscopes (Pentax Medical, Dodewaard, The Netherlands) were in use until they were replaced by Pentax ED34-i10T2 duodenoscopes with a disposable elevator cap. One ED34-i10T2 was introduced after 7 weeks and 4 after week 40. During the entire study, 2 TJF-160VR duodenoscopes (Olympus, Zoeterwoude, The Netherlands), 3 Pentax EG-3870UTK linear echoendoscopes, and 2 Pentax EG-3270UK linear echoendoscopes were in use. Servicing, maintenance, and loan endoscopes were provided by Pentax and Olympus.
      Reprocessing was performed according to manufacturers’ instructions for use and the handbook of the Dutch Steering Group for Flexible Endoscope Cleaning and Disinfection

      Advisory Board Cleaning and Disinfection Flexible Endoscopes (SFERD). Professional standard handbook version 3.0. Flexible endoscopes—cleaning and disinfection. 2013 [updated 18-09-2013]. Available at: https://www.infectiepreventieopleidingen.nl/downloads/SFERDHandbook3_1.pdf. Accessed January 30, 2021.

      and consisted a bedside flush with water, manual cleaning using Neodisher MediClean Forte cleaning agent (dr. Weigert, Assen, The Netherlands), and HLD using Neodisher Septo peracetic acid in 8 Wassenburg WD440-PT automated endoscope reprocessors (Wassenburg, Dodewaard, The Netherlands). In April 2018, the endoscopy department relocated, and 10 new Wassenburg WD440-PT automated endoscope reprocessors were put in use. In December 2017, the original staff of 4 disinfection assistants was merged with the sterilization unit of 15 members. The existing clinical monthly surveillance cultures were continued during the study; if contaminated with gut flora, endoscopes were quarantined until effectively decontaminated as shown by follow-up cultures.

      Study design

      This before-and-after study consisted of 2 parts: a baseline control period followed by an intervention period (Supplementary Fig. 1, available online at www.giejournal.org). In both periods DLEs were subjected to postcleaning ATP tests and cultured after HLD by dedicated sampling staff. In the control period, participants were blinded to ATP results: ATP-positive endoscopes were not recleaned. ATP test results were shown to the disinfection assistant only in the intervention period; if positive, the endoscope was cleaned and ATP tested again, with a maximum of 2 repeated cycles. DLEs underwent HLD regardless of the result of a possible positive third ATP test. Sampling staff, disinfection assistants, and researchers were blinded to the study culture results.
      The primary endpoint was the reduction of the percentage of DLEs contaminated with gut flora between the control period and the intervention period. We considered the effect of the ATP test as clinically relevant if the gut flora contamination rate was reduced by 50%. Based on Dutch prevalence studies, we estimated an 8% gut flora contamination rate for DLEs,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      ,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      High prevalence rate of digestive tract bacteria in duodenoscopes: a nationwide study.
      without differences between the different DLE brands or types.

      U.S. Senate. Preventable tragedies: superbugs and how ineffective monitoring of medical device safety fails patients [U.S. Senate–Minority Staff Report]. Washington, DC, 2016 [updated 13-01-2016]. Available at: https://www.help.senate.gov/imo/media/doc/Duodenoscope%20Investigation%20FINAL%20Report.pdf. Accessed January 30, 2021.

      ,
      • Rubin Z.A.
      • Kim S.
      • Thaker A.M.
      • et al.
      Safely reprocessing duodenoscopes: current evidence and future directions.
      ,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      • Brandabur J.J.
      • Leggett J.E.
      • Wang L.
      • et al.
      Surveillance of guideline practices for duodenoscope and linear echoendoscope reprocessing in a large healthcare system.
      • Chapman C.G.
      • Siddiqui U.D.
      • Manzano M.
      • et al.
      Risk of infection transmission in curvilinear array echoendoscopes: results of a prospective reprocessing and culture registry.
      ,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      High prevalence rate of digestive tract bacteria in duodenoscopes: a nationwide study.
      This required a sample size (power, 80%; α = .05) of 870 procedures. The study was approved by the Erasmus MC Medical Ethical Research Committee (MEC-2017-291); no patients were subject to any study procedures, and no patient data were collected.

      ATP tests

      Depending on DLE type, 3 to 4 sites were sampled. First, the distal 10 cm of the endoscope, then the forceps elevator, and finally the protection cap (if unsealed and reusable) were swabbed with Clean-Trace ATP surface tests (3M Company, Maplewood, Minn, USA). Finally, 40 mL of sterile water flushed through the biopsy/suction channel was sampled with the Clean-Trace ATP water test. Endoscopes were considered ATP-positive if ≥1 sample tested ≥201 RLUs. This cutoff has been validated for the distal tip and flush
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      Validation of adenosine triphosphate to audit manual cleaning of flexible endoscope channels.
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      The adenosine triphosphate test is a rapid and reliable audit tool to assess manual cleaning adequacy of flexible endoscope channels.
      • Alfa M.J.
      • Olson N.
      • Murray B.-L.
      Comparison of clinically relevant benchmarks and channel sampling methods used to assess manual cleaning compliance for flexible gastrointestinal endoscopes.
      and also used for the elevator and cap for which this was not the case.

      Microbiologic cultures

      Depending on DLE type, 3 to 4 sites were sampled. First, the forceps elevator and then the protection cap (if unsealed and reusable) were swabbed with e-Swabs (Copan Italia SpA, Brescia, Italy). Next, the biopsy/suction channels were flushed with sterile physiologic saline solution of which 20 mL was collected in a sterile container at the distal tip. Finally, the biopsy/suction channels were brushed with either Olympus BW-412T or Pentax CS6021T single-use cleaning brushes. Brush tips and e-Swabs were vortexed in e-Swab 1-mL Amies medium of which .75 mL was poured on blood agar. Channel flushes were filtrated over a .22-μm filter, of which the filtrate was forced on R2A agar. Cultures were incubated at 35ºC at 5% CO2 and examined for growth after day 4. Culture results were presented in colony forming units (CFUs)/20 mL per microorganism.
      We used 2 contamination definitions: presence of microbial growth (≥1 CFUs/20 mL) of GI microorganisms (gut flora) and microbial growth with ≥20 CFUs/20 mL of any type of microorganism (AM20) as used by the European and Dutch guidelines.

      Advisory Board Cleaning and Disinfection Flexible Endoscopes (SFERD). Professional standard handbook version 3.0. Flexible endoscopes—cleaning and disinfection. 2013 [updated 18-09-2013]. Available at: https://www.infectiepreventieopleidingen.nl/downloads/SFERDHandbook3_1.pdf. Accessed January 30, 2021.

      ,
      • Beilenhoff U.
      • Neumann C.S.
      • Rey J.F.
      • et al.
      ESGE-ESGENA guideline for quality assurance in reprocessing: microbiological surveillance testing in endoscopy.
      Endoscopes could be contaminated according to either or both definitions.

      Statistical analysis

      To investigate the effect of the intervention on the probability of contamination measured at any of the sampling sites, we fitted Bayesian logistic mixed models taking into account a correlation between repeated measurements of the same endoscope using endoscope-specific (random) intercepts. The models were fitted using the R package JointAI, which performs simultaneous analysis and imputation of incomplete covariates.
      • Erler N.S.
      • Rizopoulos D.
      • Lesaffre E.M.E.H.
      JointAI: joint analysis and imputation of incomplete data in R.
      We used 2 models. Model 1 was the main analysis, in which we included all covariates potentially influencing reprocessing: study period, study time (per 10 weeks), study procedures (per 50 procedures), transfer to a new endoscopy building, and introduction of new disinfection assistants. Model 2 was used for the secondary analysis, in which we assessed differences between endoscope types by adding the endoscopy type and its interaction with the study period to the model.
      Results are presented as posterior means and 95% credible intervals (CIs) of the odds ratios of the covariates. For clinical interpretation of the results of model 2, estimated probabilities were calculated for both study periods, whereas other values were fixed to reasonable reference values. Analyses were performed in SPSS version 23.0 (SPSS Institute, Chicago, Ill, USA) and R version 4.0.3 (Vienna, Austria).

      Results

      Of 909 procedures, 431 were performed during the control period of 20 weeks and 478 during the intervention period of 45 weeks (Table 1). Of these procedures, 473 were performed with duodenoscopes and 433 with linear echoendoscopes. Supplementary Figure 2 (available online at www.giejournal.org) shows the inclusion and contamination timelines per endoscope type. Eight procedures in which the endoscope used was not known and 6 procedures with unknown ATP test results were excluded from the analyses. Three unknown values for the number of study procedures and 7 missing values pertaining to which disinfection assistant was involved were imputed during the analysis.
      Table 1Contamination prevalence of duodenoscopes and linear echoendoscopes
      No. of casesGut floraAM20
      Duodenoscopes and linear echoendoscopes
       Control period43116 (67)34 (145)
       Intervention period47821 (102)39 (185)
      1 time manual cleaning31528 (88)40 (126)
      ≥2 times manual cleaning1629 (14)36 (59)
      Duodenoscopes47330 (143)59 (280)
       Control period22427 (61)60 (136)
       Intervention period24933 (82)58 (144)
      Linear echoendoscopes4336 (26)12 (50)
       Control period2053 (6)4 (9)
       Intervention period2289 (20)18 (41)
      Values are % (n).
      AM20, Microbial growth with ≥20 colony-forming units/20 mL of any type of microorganism.

      Gut flora contamination

      The main analysis, assessing covariates potentially influencing reprocessing, showed that during the intervention period DLEs were less likely to be contaminated with gut flora (odds ratio, .32; 95% CI, .12-.85) (Fig. 1). However, as shown in Table 1, the DLE gut flora contamination rate was higher in the intervention period (21%; n = 102) than in the control period (16%; n = 67). This difference follows from the performance of 2 distinct duodenoscopes, explained in the following 3 steps.
      Figure thumbnail gr1
      Figure 1Odds ratios for reprocessing covariates. AM20, Microbial growth with ≥20 colony-forming units/20 mL of any type of microorganism; ATP, adenosine triphosphate; OR, odds ratio; CI, credible interval.
      First, a secondary analysis (Table 2, Fig. 2) assessing the intervention period effect per endoscope type showed that the effect was based on only ED34-i10T duodenoscopes, whereas no effect was seen in the other endoscope types. ED34-i10T duodenoscopes had during the intervention period a lower estimated probability of contamination with gut flora compared with the control period (control vs intervention period: mean, 39.0% [95% CI, 17.6%-63.6%] vs 9.1% [95% CI, 2%-21%]).
      Table 2Distribution of the estimated probability of gut flora contamination
      Gut flora
      ControlInterventionDifference
      Duodenoscopes
       Olympus TJF-160VR15.4 (3.5-37.2)8.8 (2.0-22.0)–6.6 (–27.7 to 8.7)
       Pentax ED34-i10T2 disposable elevator cap15.6 (.2-60.0)43.4 (18.4-72.1)27.8 (–18.6 to 63.5)
       Pentax ED34-i10T39.0 (17.6-63.6)9.1 (2.4-20.9)–29.9 (-52.3 to –11.0)
      Linear echoendoscopes
       Pentax EG-3270UK4.7 (.1-23.1)10.2 (1.4-33.4)5.5 (–10.4 to 26.5)
       Pentax EG-3870UTK4.8 (1.2-12.3)4.2 (1.2-10.2)–.06 (–7.7 to 5.3)
      Estimated probabilities (in %) of gut flora contamination during the control and intervention periods with corresponding 95% credible intervals, under the assumption of reference values (old building, core team, study week 23, 39 procedures) for the other covariates. Pentax ED34-i10T includes duodenoscopes A, B, C, and D as shown in Figure 3.
      Figure thumbnail gr2
      Figure 2Distribution of the estimated probability of contamination by control period and intervention period as well as endoscope type. Other covariates were old endoscopy unit, core disinfection assistant team, study week 23, and 39 procedures. Pentax ED34-i10T includes duodenoscopes A, B, C, and D as shown in . AM20, Microbial growth with ≥20 colony-forming units/20 mL of any type of microorganism; ATP, adenosine triphosphate.
      Second, close observation of the 4 ED34-i10T duodenoscopes showed that duodenoscopes A and B had lower gut flora contamination rates in the intervention period: A, 69% (n = 29) versus 39% (n = 16), and B, 36% (n = 13) versus 10% (n = 2). The other 2 duodenoscopes had very few observations (C, 13% [n = 6] vs 14% [n = 1], and D, 20% [n = 4] vs 25% [n = 2]) and could thus contribute little to the results of the secondary analysis.
      Third, it was observed that the control period gut flora rates of duodenoscopes A and B were predominantly caused by episodes of ongoing contamination, which ended before the start of the intervention period (Fig. 3). Both duodenoscopes had an episode with Enterobacter cloacae complex and Klebsiella pneumoniae, and duodenoscope A had another episode with Candida parapsilosis. The episodes were only detected by regular surveillance cultures after multiple positive blind study cultures, but the C parapsilosis episode was not detected at all. These episodes ended in November, and further study and surveillance cultures remained negative for gut flora in the control period. Because no gut flora was present at the start of the intervention period, this indicates that ATP testing did not terminate these episodes. Also in the intervention period, new episodes with gut flora emerged among 1 ED34-i10T duodenoscope and 3 ED34-i10T2 duodenoscopes. Duodenoscope A had a third episode with Stenotrophomonas maltophilia in its final month in the intervention period (Fig. 3), and 3 ED34-i10T2 duodenoscopes had up to 2-month-long episodes with Enterobacter aerogenes, E cloacae complex, S maltophilia, C parapsilosis, and yeasts. These new episodes during the intervention period indicate that ATP testing did not prevent ongoing gut flora contamination.
      Figure thumbnail gr3
      Figure 3Timeline of Pentax ED34-i10T duodenoscopes. Four Pentax ED34-i10T duodenoscopes (A, B, C, and D) were in use until week 40. Microbiologic surveillance was performed monthly; endoscopes contaminated with gut flora were quarantined until they were effectively decontaminated as shown by surveillance cultures. Results of surveillance cultures are delayed because of the 72-hour incubation period. Columns represents study cultures. Both researchers and clinicians were blinded to the result of study cultures. Circles represent surveillance cultures, which are only shown if positive for gut flora. Gray columns are negative for gut flora, whereas colored columns or circles are positive for gut flora. Pentax inspected endoscopes after ongoing positive surveillance cultures. This is shown as an inspection if no defects were detected and as repair if damage and/or wear required repairs and replacement of parts.
      Sample sites that were ATP-positive after cleaning harbored gut flora after HLD in a selected number of cases. Of all endoscopes, the elevator and cap were the sites that were most often ATP-positive during the control (both sites ≥66%) and intervention (both sites ≥27%) periods (Supplementary Table 1, available online at www.giejournal.org). However, these sites had the lowest gut flora contamination rates (elevator, 2% [n = 16]; cap, <1% [n = 1]) (Supplementary Table 2, available online at www.giejournal.org). A similar pattern was seen in the channels of linear echoendoscopes. Although 38% tested ATP-positive in the control period and 16% in the intervention period, the gut flora contamination rates detected by channel flush and brush were lower (control period, 3%; intervention period, 7%). On the other hand, sample sites that did harbor gut flora were ATP-positive in a low number of cases. Of all sample sites, duodenoscope channels harbored the most gut flora (flush and brush combined: control period, 26%; intervention period, 32%), but the channels tested ATP-positive the least often in the control period (15%) and intervention period (5%). This was in particular the case for ED34-i10T2 duodenoscopes: 52% (n = 54) were contaminated with gut flora during the intervention period, but 3% (n = 3) tested ATP-positive. This contrast indicates a low correlation between a positive postcleaning ATP test and presence of gut flora in terminal cultures.

      AM20 contamination

      The main analysis also showed that the intervention period did not reduce AM20 contamination rates (odds ratio, 2.03; 95% CI, .77-5.85) (Fig. 1), which is consistent with the AM20 contamination rates of all DLEs (>34% in both periods) (Table 1). AM20 contamination did become more likely per each 50 procedures performed during the time that the study progressed (odds ratio, 1.01; 95% CI, 1.00-1.02). In particular, ED34-i10T (86%, n = 197) and ED34-i10T2 (49%, n = 60) duodenoscopes had consistently high AM20 rates (Supplementary Table 3, available online at www.giejournal.org). The secondary analysis (Table 2, Fig. 2), assessing the intervention period effect per endoscope type, showed slightly higher estimated probabilities for all endoscope types in the intervention period but not large enough to conclude a negative impact with certainty. Contamination rates in relation to ERCP and EUS specifics are presented in Supplementary Table 4 (available online at www.giejournal.org).

      Recleaning

      In the intervention period, DLEs that underwent extra cleaning because of a positive ATP test had lower gut flora contamination rates (Table 1). This was both the case for duodenoscopes (gut flora: 37% [n = 74] vs 15% [n = 8]) and linear echoendoscopes (gut flora: 12% [n = 14] vs 6% [n = 6]).

      Discussion

      In our prospective before-and-after study, introduction of postmanual cleaning ATP tests did not reduce the post-HLD gut flora contamination rate of DLEs and thus did not stop or prevent the use of endoscopes contaminated with gut flora. We found a lower odds on gut flora contamination during the intervention period (21% vs 16%). However, this was based on 2 duodenoscopes with episodes of ongoing contamination during the control period, which were ended by quarantining and repairs. Until detected by routine clinical surveillance cultures, these episodes could go unnoticed; both researchers and clinicians were blinded to results of study cultures. The results of this study suggest that the currently used ATP test (ie, 4 sample sites with a 200-RLU cutoff) is unsuitable to detect inadequately cleaned endoscopes and therefore should not be used as a cleaning quality control indicator. Furthermore, monthly surveillance cultures are inadequate to prevent the use of contaminated equipment. Improvement of the reliability of quality checks and microbiologic surveillance regimens is required to prevent the risk of microbial transmission by contaminated endoscopes. Eventually, this risk must be eliminated by a radical redesign of DLEs and reprocessing methods.
      The current study confirms the results of a small clinical pilot and a simulated-use study, which both showed that postmanual cleaning ATP tests are not effective in preventing contamination of pathogenic bacteria in duodenoscopes.
      • Visrodia K.
      • Hanada Y.
      • Pennington K.M.
      • et al.
      Duodenoscope reprocessing surveillance with adenosine triphosphate testing and terminal cultures: a clinical pilot study.
      ,
      • Singh H.
      • Duerksen D.R.
      • Schultz G.
      • et al.
      Impact of cleaning monitoring combined with channel purge storage on elimination of Escherichia coli and environmental bacteria from duodenoscopes.
      In another small pilot study, endoscope cultures remained negative.
      • Sethi S.
      • Huang R.J.
      • Barakat M.T.
      • et al.
      Adenosine triphosphate bioluminescence for bacteriological surveillance and reprocessing strategies for minimizing risk of infection transmission by duodenoscopes.
      However, all previous studies were limited in size and lacked a control group
      • Visrodia K.
      • Hanada Y.
      • Pennington K.M.
      • et al.
      Duodenoscope reprocessing surveillance with adenosine triphosphate testing and terminal cultures: a clinical pilot study.
      ,
      • Singh H.
      • Duerksen D.R.
      • Schultz G.
      • et al.
      Impact of cleaning monitoring combined with channel purge storage on elimination of Escherichia coli and environmental bacteria from duodenoscopes.
      ,
      • Sethi S.
      • Huang R.J.
      • Barakat M.T.
      • et al.
      Adenosine triphosphate bioluminescence for bacteriological surveillance and reprocessing strategies for minimizing risk of infection transmission by duodenoscopes.
      or microbiologic cultures,
      • Quan E.
      • Mahmood R.
      • Naik A.
      • et al.
      Use of adenosine triphosphate to audit reprocessing of flexible endoscopes with an elevator mechanism.
      ,
      • Chang W.K.
      • Liu T.C.
      • Liu T.L.
      • et al.
      Enhanced manual cleaning efficacy of duodenoscope in endoscopy units: results of a multicenter comprehensive quality control program.
      which prevented drawing a final conclusion about the clinical merits of the ATP test. The current study, with a controlled and adequately powered design, showed that postcleaning ATP tests do not have a clinically relevant effect.
      We found a discrepancy between an ATP-positive test result and the presence of gut flora. The elevator and cap of all DLEs as well as the channels of echoendoscopes had high ATP-positive rates, but only a small number harbored gut flora. On the contrary, ED34-i10T2 duodenoscopes were ATP-positive in only 3 cases. Although some studies did find a correlation between ATP and microbial load after manual cleaning,
      • McCafferty C.E.
      • Abi-Hanna D.
      • Aghajani M.J.
      • et al.
      The validity of adenosine triphosphate measurement in detecting endoscope contamination.
      our results are in line with earlier pilot studies that also found no correlation between postcleaning ATP tests and post-HLD cultures.
      • Washburn R.E.
      • Pietsch J.J.
      Assessment of test methods for evaluating effectiveness of cleaning flexible endoscopes.
      ,
      • Visrodia K.
      • Hanada Y.
      • Pennington K.M.
      • et al.
      Duodenoscope reprocessing surveillance with adenosine triphosphate testing and terminal cultures: a clinical pilot study.
      The discrepancy can be explained by the fact that positive ATP results combined with cultures negative for gut flora can be the result of organic residue containing ATP, such as human secretions or cells, a high microbial load of non–gut flora, or noncultivable microorganisms. The 200-RLU cutoff is validated for the distal tip and flush,
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      Validation of adenosine triphosphate to audit manual cleaning of flexible endoscope channels.
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      The adenosine triphosphate test is a rapid and reliable audit tool to assess manual cleaning adequacy of flexible endoscope channels.
      • Alfa M.J.
      • Olson N.
      • Murray B.-L.
      Comparison of clinically relevant benchmarks and channel sampling methods used to assess manual cleaning compliance for flexible gastrointestinal endoscopes.
      but for the experimental cap and elevator sites no validation data are known. ATP-negative results in combination with cultures positive for gut flora can be the result of a gut flora microbial load too low to raise the number of RLUs.
      • Alfa M.J.
      • Fatima I.
      • Olson N.
      Validation of adenosine triphosphate to audit manual cleaning of flexible endoscope channels.
      Another explanation can be that the ATP water test is not sensitive enough because the material is diluted by the flush water (40 mL), whereas cultures detect all viable CFUs in a channel flush as its filter is cultured. Detection is further increased by also sampling channels with a brush.
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      These results imply that the ATP test as used in this study cannot adequately detect gut flora. The lower ATP test positivity rate in the intervention period can be the result of revised reprocessing routines (ie, improved cleaning and endoscope handling) and the introduction of disposable caps for ED34-i10T duodenoscopes.
      Recleaned DLEs had lower gut flora contamination rates, suggesting extra cleaning reduces organic debris. Because manual cleaning is error-prone,
      • Ofstead C.L.
      • Wetzler H.P.
      • Snyder A.K.
      • et al.
      Endoscope reprocessing methods: a prospective study on the impact of human factors and automation.
      • Dirlam Langlay A.M.
      • Ofstead C.L.
      • Mueller N.J.
      • et al.
      Reported gastrointestinal endoscope reprocessing lapses: the tip of the iceberg.
      • Schaefer M.K.
      • Jhung M.
      • Dahl M.
      • et al.
      Infection control assessment of ambulatory surgical centers.
      • Armellino D.
      • Cifu K.
      • Wallace M.
      • et al.
      Implementation of remote video auditing with feedback and compliance for manual-cleaning protocols of endoscopic retrograde cholangiopancreatography endoscopes.
      ,

      Departement Welzijn VeG-Z. Beleidsrapport Desinfectie van flexibele endoscopen met lumen. Depotnummer D/2017/3241/232 2017 [in Dutch]. Available at: https://www.departementwvg.be/sites/default/files/media/documenten/Beleidsrapport_desinfectie_flexibele_endoscopen_met_lumen_juli_2017.pdf. Accessed January 30, 2021.

      a quality control indicator that appropriately detects inadequately cleaned endoscopes would be a benefit. Current evidence does not support the ATP test for this use: Improvement of ATP test rates while gut flora rates remain high is a false-negative test result providing an incorrect and false sense of security.
      Contamination rates of linear echoendoscopes in this study were in line with previous studies,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
      ,
      • Chapman C.G.
      • Siddiqui U.D.
      • Manzano M.
      • et al.
      Risk of infection transmission in curvilinear array echoendoscopes: results of a prospective reprocessing and culture registry.
      ,
      • Rauwers A.W.
      • Voor In 't Holt A.F.
      • Buijs J.G.
      • et al.
      High prevalence rate of digestive tract bacteria in duodenoscopes: a nationwide study.
      whereas duodenoscope rates were remarkably higher. These high rates can be the result of the study design, endoscope-related issues, and/or sampling and culture methods. Large-scale studies with far lower rates conducted daily or postprocedure surveillance in an open study design
      • Brandabur J.J.
      • Leggett J.E.
      • Wang L.
      • et al.
      Surveillance of guideline practices for duodenoscope and linear echoendoscope reprocessing in a large healthcare system.
      ,
      • Ross A.S.
      • Baliga C.
      • Verma P.
      • et al.
      A quarantine process for the resolution of duodenoscope-associated transmission of multidrug-resistant Escherichia coli.
      ,
      • Higa J.T.
      • Choe J.
      • Tombs D.
      • et al.
      Optimizing duodenoscope reprocessing: rigorous assessment of a culture and quarantine protocol.
      allowed immediate quarantine of contaminated endoscopes, which prevented further use. For most hospitals daily surveillance is too labor intense and expensive.
      High contamination rates in studies blinded to study culture outcomes are perhaps more representative of their clinical practice. High rates were also found in cross-sectional studies,
      • Ribeiro M.M.
      • de Oliveira A.C.
      Analysis of the air/water channels of gastrointestinal endoscopies as a risk factor for the transmission of microorganisms among patients.
      ,
      • Ofstead C.L.
      • Wetzler H.P.
      • Johnson E.A.
      • et al.
      Simethicone residue remains inside gastrointestinal endoscopes despite reprocessing.
      and several outbreak reports showed that contamination can persist for months.
      • Verfaillie C.J.
      • Bruno M.J.
      • Voor in 't Holt A.F.
      • et al.
      Withdrawal of a novel-design duodenoscope ends outbreak of a VIM-2-producing Pseudomonas aeruginosa.
      ,
      • Alrabaa S.F.
      • Nguyen P.
      • Sanderson R.
      • et al.
      Early identification and control of carbapenemase-producing Klebsiella pneumoniae, originating from contaminated endoscopic equipment.
      ,
      • Epstein L.
      • Hunter J.C.
      • Arwady M.A.
      • et al.
      New Delhi metallo-beta-lactamase-producing carbapenem-resistant Escherichia coli associated with exposure to duodenoscopes.
      ,
      • Wendorf K.A.
      • Kay M.
      • Baliga C.
      • et al.
      Endoscopic retrograde cholangiopancreatography-associated AmpC Escherichia coli outbreak.
      ,
      • Rauwers A.W.
      • Troelstra A.
      • Fluit A.C.
      • et al.
      Independent root-cause analysis of contributing factors, including dismantling of 2 duodenoscopes, to investigate an outbreak of multidrug-resistant Klebsiella pneumoniae.
      • Kim S.
      • Russell D.
      • Mohamadnejad M.
      • et al.
      Risk factors associated with the transmission of carbapenem-resistant Enterobacteriaceae via contaminated duodenoscopes.
      • Aumeran C.
      • Poincloux L.
      • Souweine B.
      • et al.
      Multidrug-resistant Klebsiella pneumoniae outbreak after endoscopic retrograde cholangiopancreatography.
      The high rates in this study were mainly based on contamination episodes of a select number of Pentax-type duodenoscopes. Recurrent episodes of gut flora and ongoing contamination with waterborne and/or skin flora in both types suggest the presence of biofilms in endoscope channels, including new duodenoscopes. Once present, a biofilm can be the cause of failure of reprocessing.
      • Akinbobola A.B.
      • Amaeze N.J.
      • Mackay W.G.
      • et al.
      "Secondary biofilms" could cause failure of peracetic acid high-level disinfection of endoscopes.
      Duodenoscope damage and wear can contribute to contamination.
      • Wendorf K.A.
      • Kay M.
      • Baliga C.
      • et al.
      Endoscopic retrograde cholangiopancreatography-associated AmpC Escherichia coli outbreak.
      ,
      • Ross A.S.
      • Baliga C.
      • Verma P.
      • et al.
      A quarantine process for the resolution of duodenoscope-associated transmission of multidrug-resistant Escherichia coli.
      ,
      • Rauwers A.W.
      • Troelstra A.
      • Fluit A.C.
      • et al.
      Independent root-cause analysis of contributing factors, including dismantling of 2 duodenoscopes, to investigate an outbreak of multidrug-resistant Klebsiella pneumoniae.
      ,
      • Lee D.H.
      • Kim D.B.
      • Kim H.Y.
      • et al.
      Increasing potential risks of contamination from repetitive use of endoscope.
      In our study in duodenoscope A, the first episode ended only after the third servicing, whereas in duodenoscope B the episode continued despite servicing. Close cooperation and taking joint responsibility by both end users and endoscope manufacturers for reprocessing, strict surveillance, and servicing are important in guarding endoscope safety.
      This study showed that monthly surveillance was inadequate for the timely detection of contamination. Until detected by surveillance cultures, episodes that lasts multiple weeks continue to expose patients to contaminated equipment. One episode of C parapsilosis was “missed” (ie, not present in any surveillance culture). This is in line with the fact that during contamination episodes not all consecutive study cultures are positive for the same bacteria species. Therefore, high-frequency daily or weekly surveillance should be considered to detect ongoing contamination patterns.
      To the best of our knowledge, this is the first controlled study with blinded study cultures required to investigate postcleaning ATP tests. The current study also has some limitations. Generalizability of the results is limited by multiple factors including the single-center design and susceptibility of distinct endoscopes to contamination. The intervention period had a longer duration than the control period because of a lower inclusion rate. This was because depending on the number of daily procedures and available endoscopes, not all endoscopes could undergo the longer intervention period ATP test cycle. Major reprocessing-affecting events that occurred during the study were accounted for by including these factors in the statistical analysis model.
      To conclude, this before-and-after intervention study shows that postmanual cleaning ATP tests as used in this study do not reduce the number of contaminated DLEs after HLD. To prevent the use of contaminated equipment, reliable control measures are required to assess whether reprocessing of reusable DLEs was adequate. Until the risk of contamination is eliminated by sterilization of DLEs or single-use endoscopes, strict and frequent microbiologic surveillance is indicated.

      Acknowledgment

      We thank the reprocessing staff, sampling staff, medical devices experts, infection control professionals, medical microbiologists, and gastroenterologists at the Erasmus MC for their participation and efforts in this study.

      Appendix

      Figure thumbnail fx1
      Supplementary Figure 1Flowchart. Adenosine triphosphate (ATP) negative refers to test results of all sites that were ≤200 relative light units (RLUs), whereas ATP positive refers to ≥1 test site result that was ≥201 RLUs.
      Figure thumbnail fx2
      Supplementary Figure 2Timeline of inclusion and outcome rates per endoscope type. Each line represents 1 endoscope, and each dot marks a measurement. Vertical dotted lines represent the divide between the control and intervention periods. Blue dots represent a negative outcome. Red dots in the first column represent endoscopes contaminated with gut flora, red dots in the second column represent endoscopes contaminated with AM20, and red dots in the third column represent endoscopes with a positive first ATP test. AM20, Microbial growth with ≥20 colony-forming units/20 mL of any type of microorganism; ATP, adenosine triphosphate.
      Supplementary Table 1Positive ATP test rate per sample site and endoscope type per study period


      No. of casesPositive ATP test
      ≥1 siteDistal tipChannel flushElevatorCap
      All duodenoscopes and linear echoendoscopes
       Control423344 (81)226 (53)111 (26)280 (66)139/204 (68)
       Intervention first ATP test472162 (34)78 (17)49 (10)127 (27)32/114 (28)
       Intervention second ATP test15956 (35)27 (17)8 (5)38 (24)10/43 (23)
       Intervention third ATP test5326 (49)10 (19)4 (8)22 (42)5/13 (38)
      Duodenoscopes
       Pentax ED34-i10T
      Control150125 (83)62 (41)23 (15)87 (58)106 (71)
      Intervention first ATP test7727 (35)7 (9)3 (4)22 (29)13/51 (25)
      Twenty-seven weeks after the start of the intervention period, the endoscopy department changed to single-use protection caps for the Pentax ED34-i10T. Therefore, the number is lower than the total number of endoscopes.
      Intervention second ATP test278 (30)2 ()06 (22)3/21 (14)
      Twenty-seven weeks after the start of the intervention period, the endoscopy department changed to single-use protection caps for the Pentax ED34-i10T. Therefore, the number is lower than the total number of endoscopes.
      Intervention third ATP test73 (43)1 (7)02 (29)1/5 (20)
      Twenty-seven weeks after the start of the intervention period, the endoscopy department changed to single-use protection caps for the Pentax ED34-i10T. Therefore, the number is lower than the total number of endoscopes.
       Pentax ED34-i10T2 with disposable elevator cap
      Control177 (41)3 (18)2 (12)6 (35)
      Intervention first ATP test1063 (3)1 (1)2 (2)2 (2)
      Intervention second ATP test31 (33)2 (66)02 (66)
      Intervention third ATP test10000
       Olympus TJF-160VR
      Control5444 (81)29 (54)10 (19)38 (70)32 (59)
      Intervention first ATP test6424 (38)11 (17)8 (13)20 (31)19/61 (31)
      Intervention second ATP test248 (33)3 (13)2 (8)6 (25)7/22 (32)
      Intervention third ATP test85 (63)2 (25)04 (50)4/8 (50)
      Linear echoendoscopes
       Pentax EG-3870UTK
      Control145125 (86)106 (73)56 (39)111 (77)
      Intervention first ATP test18790 (48)47 (25)27 (14)69 (37)
      Intervention second ATP test8835 (40)18 (20)5 (6)23 (26)
      Intervention third ATP test3315 (45)5 (15)4 (12)14 (42)
       Pentax EG-3270UK
      Control5743 (75)26 (46)20 (35)38 (67)
      Intervention first ATP test3818 (47)12 (32)9 (24)14 (37)
      Intervention second ATP test184 (22)3 (17)1 (6)2 (11)
      Intervention third ATP test43 (75)2 (50)02 (50)
      Values are n or n/N (%) unless otherwise defined.
      ATP, Adenosine triphosphate; —, not applicable.
      Twenty-seven weeks after the start of the intervention period, the endoscopy department changed to single-use protection caps for the Pentax ED34-i10T. Therefore, the number is lower than the total number of endoscopes.
      Supplementary Table 2Gut flora contamination rate per sample site and endoscope type per study period


      No. of casesGut flora contamination
      ≥1 siteBrushChannel flushElevatorCap
      All duodenoscopes and linear echoendoscopes
       Total909169 (19)96 (11)109 (12)16 (2)1
       Control43067 (16)42 (10)42 (10)3 (1)1/209 (1)
      Not all duodenoscope and linear echoendoscope types have (reusable) protection caps.
       Intervention first cleaning31586 (27)48 (15)57 (18)11 (3)0/71
      Not all duodenoscope and linear echoendoscope types have (reusable) protection caps.
       Intervention second cleaning1069 (8)5 (5)6 (6)2 (2)0/30
      Not all duodenoscope and linear echoendoscope types have (reusable) protection caps.
       Intervention third cleaning565 (9)1 (2)4 (7)00/14
      Not all duodenoscope and linear echoendoscope types have (reusable) protection caps.
      Duodenoscopes
       Pentax ED34-i10T
      Control15153 (35)34 (23)28 (19)2 (1)1 (1)
      Intervention first cleaning5017 (34)12 (24)11 (22)3 (6)0/30
      Twenty-seven after the start of the intervention period, single-use protection caps were introduced, resulting in a lower number lower of cap samples.
      Intervention second cleaning192 (11)2 (11)2 (11)00/15
      Twenty-seven after the start of the intervention period, single-use protection caps were introduced, resulting in a lower number lower of cap samples.
      Intervention third cleaning82 (25)1 (13)1 (13)00/6
      Twenty-seven after the start of the intervention period, single-use protection caps were introduced, resulting in a lower number lower of cap samples.
       Pentax ED34-i10T2 with disposable elevator cap
      Control171 (6)1 (6)00
      Intervention first cleaning10352 (50)29 (28)39 (38)1 (1)
      Intervention second cleaning21 (50)01 (50)0
      Intervention third cleaning11111
       Olympus TJF-160VR
      Control567 (13)2 (4)5 (9)00
      Intervention first cleaning404 (10)1 (3)1 (3)2 (5)0
      Intervention second cleaning161 (6)01 (6)00
      Intervention third cleaning82 (25)02 (25)00
      Linear echoendoscopes
       Pentax EG-3870UTK
      Control1475 (3)4 (3)1 (1)1 (1)
      Intervention first cleaning9712 (12)5 (5)4 (4)5 (5)
      Intervention second cleaning552 (4)1 (2)2 (4)0
      Intervention third cleaning351 (3)01 (3)0
       Pentax EG-3270UK
      Control581 (2)1 (2)00
      Intervention first cleaning201 (5)01 (5)0
      Intervention second cleaning143 (21)2 (14)02 (14)
      Intervention third cleaning40000
      Values in parentheses are percents.
      —, Not applicable.
      Not all duodenoscope and linear echoendoscope types have (reusable) protection caps.
      Twenty-seven after the start of the intervention period, single-use protection caps were introduced, resulting in a lower number lower of cap samples.
      Supplementary Table 3Contamination prevalence per endoscope type
      No. of casesGut floraAM20
      Duodenoscope types
       Olympus TJF-160VR12012 (14)18 (21)
       Pentax ED34-i10T22832 (74)86 (197)
       Pentax ED34-i10T2 with disposable elevator cap12344 (54)49 (60)
      Linear echoendoscope types
       Pentax EG-3870UTK3346 (20)12 (40)
       Pentax EG-3270UK965 (5)10 (10)
      Values are % (n). AM20, Microbial growth with ≥20 colony-forming units/20 mL of any type of microorganism.
      Supplementary Table 4Contamination prevalence for ERCP and EUS characteristics
      No. of casesGut flora contaminationAM20 contamination
      ERCP characteristics
       Invasive procedure
      ERCP invasive procedures include sphincterotomy or use of accessories.
      38432% (121)62% (238)
       Noninvasive procedure7923% (18)46% (36)
       Use of ERCP accessories
      ERCP accessories are biopsy needles, brushes, balloons, baskets, and lithotripsy.
      22829% (65)56% (128)
       No accessories23729% (69)62% (147)
       Stent placement24933% (83)65% (161)
       No placement21526% (56)53% (114)
       Stent removal14834% (51)64% (94)
       No removal31928% (89)57% (181)
       Efflux
      Efflux includes sludge, pus, and/or active bleeding.
      7135% (25)69% (49)
       No efflux39029% (113)57% (222)
      EUS characteristics
       Invasive procedure
      EUS invasive procedures include biopsy sampling or FNA.
      1525% (8)9% (13)
       Noninvasive procedure2686% (16)13% (35)
       Efflux
      Efflux includes sludge, pus, and/or active bleeding.
      15(0)13% (2)
       No efflux4096% (25)7% (29)
      Values are % (n).
      AM20, Microbial growth with ≥20 colony-forming units/20 mL of any type of microorganism.
      ERCP invasive procedures include sphincterotomy or use of accessories.
      ERCP accessories are biopsy needles, brushes, balloons, baskets, and lithotripsy.
      Efflux includes sludge, pus, and/or active bleeding.
      § EUS invasive procedures include biopsy sampling or FNA.

      References

      1. U.S. Senate. Preventable tragedies: superbugs and how ineffective monitoring of medical device safety fails patients [U.S. Senate–Minority Staff Report]. Washington, DC, 2016 [updated 13-01-2016]. Available at: https://www.help.senate.gov/imo/media/doc/Duodenoscope%20Investigation%20FINAL%20Report.pdf. Accessed January 30, 2021.

        • Rubin Z.A.
        • Kim S.
        • Thaker A.M.
        • et al.
        Safely reprocessing duodenoscopes: current evidence and future directions.
        Lancet Gastroenterol Hepatol. 2018; 3: 499-508
        • Balan G.G.
        • Sfarti C.V.
        • Chiriac S.A.
        • et al.
        Duodenoscope-associated infections: a review.
        Eur J Clin Microbiol Infect Dis. 2019; 38: 2205-2213
        • Carbonne A.
        • Thiolet J.M.
        • Fournier S.
        • et al.
        Control of a multi-hospital outbreak of KPC-producing Klebsiella pneumoniae type 2 in France, September to October 2009.
        Euro Surveill. 2010; 15: 19734
        • Valderrama S.
        • Miranda C.J.L.
        • Rubio A.P.G.
        • et al.
        Successful control of an endoscopic retrograde cholangiopancreatography–associated nosocomial outbreak caused by Klebsiella pneumoniae carbapenemase producing Klebsiella pneumoniae in a University Hospital in Bogota, Colombia.
        Open Forum Infect Dis. 2016; 3: 1398
        • Verfaillie C.J.
        • Bruno M.J.
        • Voor in 't Holt A.F.
        • et al.
        Withdrawal of a novel-design duodenoscope ends outbreak of a VIM-2-producing Pseudomonas aeruginosa.
        Endoscopy. 2015; 47: 493-502
        • Kola A.
        • Piening B.
        • Pape U.F.
        • et al.
        An outbreak of carbapenem-resistant OXA-48-producing Klebsiella pneumonia associated to duodenoscopy.
        Antimicrob Resist Infect Control. 2015; 4: 8
        • Alrabaa S.F.
        • Nguyen P.
        • Sanderson R.
        • et al.
        Early identification and control of carbapenemase-producing Klebsiella pneumoniae, originating from contaminated endoscopic equipment.
        Am J Infect Control. 2013; 41: 562-564
        • García-Cano J.
        • Perez-Miranda M.
        • Enric B.
        • et al.
        Duodenoscope-transmitted infection: results from a national Spanish survey [abstract].
        Gastrointest Endosc. 2016; 83: AB255
        • Rutala W.A.
        • Weber D.J.
        Outbreaks of carbapenem-resistant Enterobacteriaceae infections associated with duodenoscopes: What can we do to prevent infections?.
        Am J Infect Control. 2016; 44: e47-e51
        • Roberts C.G.
        Studies on the bioburden on medical devices and the importance of cleaning. Disinfection, sterilization and antisepsis: principles and practices in healthcare facilities.
        Association for Professionals in Infection Control and Epidemiology, Washington, DC2001: 63-69
        • Alfa M.J.
        • Degagne P.
        • Olson N.
        Worst-case soiling levels for patient-used flexible endoscopes before and after cleaning.
        Am J Infect Control. 1999; 27: 392-401
        • Rutala W.A.
        • Weber D.J.
        FDA labeling requirements for disinfection of endoscopes: a counterpoint.
        Infect Control Hosp Epidemiol. 1995; 16: 231-235
        • Ofstead C.L.
        • Wetzler H.P.
        • Snyder A.K.
        • et al.
        Endoscope reprocessing methods: a prospective study on the impact of human factors and automation.
        Gastroenterol Nurs. 2010; 33: 304-311
        • Dirlam Langlay A.M.
        • Ofstead C.L.
        • Mueller N.J.
        • et al.
        Reported gastrointestinal endoscope reprocessing lapses: the tip of the iceberg.
        Am J Infect Control. 2013; 41: 1188-1194
        • Schaefer M.K.
        • Jhung M.
        • Dahl M.
        • et al.
        Infection control assessment of ambulatory surgical centers.
        JAMA. 2010; 303: 2273-2279
        • Armellino D.
        • Cifu K.
        • Wallace M.
        • et al.
        Implementation of remote video auditing with feedback and compliance for manual-cleaning protocols of endoscopic retrograde cholangiopancreatography endoscopes.
        Am J Infect Control. 2018; 46: 594-596
        • Epstein L.
        • Hunter J.C.
        • Arwady M.A.
        • et al.
        New Delhi metallo-beta-lactamase-producing carbapenem-resistant Escherichia coli associated with exposure to duodenoscopes.
        JAMA. 2014; 312: 1447-1455
        • Wendorf K.A.
        • Kay M.
        • Baliga C.
        • et al.
        Endoscopic retrograde cholangiopancreatography-associated AmpC Escherichia coli outbreak.
        Infect Control Hosp Epidemiol. 2015; 36: 634-642
      2. U.S. Food and Drug Administration. The FDA continues to remind facilities of the importance of following duodenoscope reprocessing instructions: FDA Safety Communication 2019 [updated 12-04-2019]. Available at: https://www.fda.gov/medical-devices/safety-communications/fda-continues-remind-facilities-importance-following-duodenoscope-reprocessing-instructions-fda. Accessed January 30, 2021.

        • Rauwers A.W.
        • Voor In 't Holt A.F.
        • Buijs J.G.
        • et al.
        Nationwide risk analysis of duodenoscope and linear echoendoscope contamination.
        Gastrointest Endosc. 2020; 92: 681-691
        • Brandabur J.J.
        • Leggett J.E.
        • Wang L.
        • et al.
        Surveillance of guideline practices for duodenoscope and linear echoendoscope reprocessing in a large healthcare system.
        Gastrointest Endosc. 2016; 84: 392-399
        • Chapman C.G.
        • Siddiqui U.D.
        • Manzano M.
        • et al.
        Risk of infection transmission in curvilinear array echoendoscopes: results of a prospective reprocessing and culture registry.
        Gastrointest Endosc. 2017; 85: 390-397
        • Larsen S.
        • Russell R.V.
        • Ockert L.K.
        • et al.
        Rate and impact of duodenoscope contamination: a systematic review and meta-analysis.
        EClin Med. 2020; 25: 100451
        • Rauwers A.W.
        • Voor In 't Holt A.F.
        • Buijs J.G.
        • et al.
        High prevalence rate of digestive tract bacteria in duodenoscopes: a nationwide study.
        Gut. 2018; 67: 1637-1645
        • Petersen B.T.
        • Cohen J.
        • Hambrick R.D. 3rd
        • et al.
        • Reprocessing Guideline Task Force
        Multisociety guideline on reprocessing flexible GI endoscopes: 2016 update.
        Gastrointest Endosc. 2017; 85: 282-294
      3. Centers for Disease Control and Prevention. Duodenoscope surveillance sampling & culturing: reducing the risks of infection 2018. Available at: https://www.fda.gov/downloads/MedicalDevices/ProductsandMedicalProcedures/ReprocessingofReusableMedicalDevices/UCM597949.pdf. Accessed January 30, 2021.

        • Beilenhoff U.
        • Biering H.
        • Blum R.
        • et al.
        Prevention of multidrug-resistant infections from contaminated duodenoscopes: position statement of the European Society of Gastrointestinal Endoscopy (ESGE) and European Society of Gastroenterology Nurses and Associates (ESGENA).
        Endoscopy. 2017; 49: 1098-1106
        • Day L.W.
        • Kwok K.
        • Visrodia K.
        • et al.
        American Society for Gastrointestinal Endoscopy Infection Control Summit: updates, challenges, and the future of infection control in GI endoscopy.
        Gastrointest Endosc. 2020; 93: 1-10
        • Batailler P.
        • Saviuc P.
        • Picot-Gueraud R.
        • et al.
        Usefulness of adenosine triphosphate bioluminescence assay (ATPmetry) for monitoring the reprocessing of endoscopes.
        Infect Control Hosp Epidemiol. 2015; 36: 1437-1443
        • Hansen D.
        • Benner D.
        • Hilgenhöner M.
        • et al.
        ATP measurement as method to monitor the quality of reprocessing flexible endoscopes.
        Ger Med Sci. 2004; 2: 1-5
        • Olafsdottir L.B.
        • Wright S.B.
        • Smithey A.
        • et al.
        Adenosine triphosphate quantification correlates poorly with microbial contamination of duodenoscopes.
        Infect Control Hosp Epidemiol. 2017; 38: 678-684
        • Olafsdottir L.B.
        • Whelan J.
        • Snyder G.M.
        A systematic review of adenosine triphosphate as a surrogate for bacterial contamination of duodenoscopes used for endoscopic retrograde cholangiopancreatography.
        Am J Infect Control. 2018; 46: 697-705
        • Washburn R.E.
        • Pietsch J.J.
        Assessment of test methods for evaluating effectiveness of cleaning flexible endoscopes.
        Am J Infect Control. 2018; 46: 685-688
        • Visrodia K.
        • Hanada Y.
        • Pennington K.M.
        • et al.
        Duodenoscope reprocessing surveillance with adenosine triphosphate testing and terminal cultures: a clinical pilot study.
        Gastrointest Endosc. 2017; 86: 180-186
        • Alfa M.J.
        • Fatima I.
        • Olson N.
        Validation of adenosine triphosphate to audit manual cleaning of flexible endoscope channels.
        Am J Infect Control. 2013; 41: 245-248
        • Alfa M.J.
        • Fatima I.
        • Olson N.
        The adenosine triphosphate test is a rapid and reliable audit tool to assess manual cleaning adequacy of flexible endoscope channels.
        Am J Infect Control. 2013; 41: 249-253
        • Alfa M.J.
        • Olson N.
        • Murray B.-L.
        Comparison of clinically relevant benchmarks and channel sampling methods used to assess manual cleaning compliance for flexible gastrointestinal endoscopes.
        Am J Infect Control. 2014; 42: e1-e5
        • Alfa M.J.
        Quality systems approach for endoscope reprocessing: you don't know what you don't know!.
        Gastrointest Endosc Clin North Am. 2020; 30: 693-709
        • Kenters N.
        • Tartari E.
        • Hopman J.
        • et al.
        Worldwide practices on flexible endoscope reprocessing.
        Antimicrob Resist Infect Control. 2018; 7: 153
      4. Advisory Board Cleaning and Disinfection Flexible Endoscopes (SFERD). Professional standard handbook version 3.0. Flexible endoscopes—cleaning and disinfection. 2013 [updated 18-09-2013]. Available at: https://www.infectiepreventieopleidingen.nl/downloads/SFERDHandbook3_1.pdf. Accessed January 30, 2021.

        • Beilenhoff U.
        • Neumann C.S.
        • Rey J.F.
        • et al.
        ESGE-ESGENA guideline for quality assurance in reprocessing: microbiological surveillance testing in endoscopy.
        Endoscopy. 2007; 39: 175-181
        • Erler N.S.
        • Rizopoulos D.
        • Lesaffre E.M.E.H.
        JointAI: joint analysis and imputation of incomplete data in R.
        arXiv preprint arXiv. 2019; (Available at: https://doi.org/10.48550/arXiv.1907.10867. Accessed January 30, 2021)190710867.
        • Singh H.
        • Duerksen D.R.
        • Schultz G.
        • et al.
        Impact of cleaning monitoring combined with channel purge storage on elimination of Escherichia coli and environmental bacteria from duodenoscopes.
        Gastrointest Endosc. 2018; 88: 292-302
        • Sethi S.
        • Huang R.J.
        • Barakat M.T.
        • et al.
        Adenosine triphosphate bioluminescence for bacteriological surveillance and reprocessing strategies for minimizing risk of infection transmission by duodenoscopes.
        Gastrointest Endosc. 2016; 85: 1180-1187e1
        • Quan E.
        • Mahmood R.
        • Naik A.
        • et al.
        Use of adenosine triphosphate to audit reprocessing of flexible endoscopes with an elevator mechanism.
        Am J Infect Control. 2018; 46: 1272-1277
        • Chang W.K.
        • Liu T.C.
        • Liu T.L.
        • et al.
        Enhanced manual cleaning efficacy of duodenoscope in endoscopy units: results of a multicenter comprehensive quality control program.
        Am J Infect Control. 2019; 47: 1233-1239
        • McCafferty C.E.
        • Abi-Hanna D.
        • Aghajani M.J.
        • et al.
        The validity of adenosine triphosphate measurement in detecting endoscope contamination.
        J Hosp Infect. 2018; 100: e142-e145
      5. Departement Welzijn VeG-Z. Beleidsrapport Desinfectie van flexibele endoscopen met lumen. Depotnummer D/2017/3241/232 2017 [in Dutch]. Available at: https://www.departementwvg.be/sites/default/files/media/documenten/Beleidsrapport_desinfectie_flexibele_endoscopen_met_lumen_juli_2017.pdf. Accessed January 30, 2021.

        • Ross A.S.
        • Baliga C.
        • Verma P.
        • et al.
        A quarantine process for the resolution of duodenoscope-associated transmission of multidrug-resistant Escherichia coli.
        Gastrointest Endosc. 2015; 82: 477-483
        • Higa J.T.
        • Choe J.
        • Tombs D.
        • et al.
        Optimizing duodenoscope reprocessing: rigorous assessment of a culture and quarantine protocol.
        Gastrointest Endosc. 2018; 88: 223-229
        • Ribeiro M.M.
        • de Oliveira A.C.
        Analysis of the air/water channels of gastrointestinal endoscopies as a risk factor for the transmission of microorganisms among patients.
        Am J Infect Control. 2012; 40: 913-916
        • Ofstead C.L.
        • Wetzler H.P.
        • Johnson E.A.
        • et al.
        Simethicone residue remains inside gastrointestinal endoscopes despite reprocessing.
        Am J Infect Control. 2016; 44: 1237-1240
        • Rauwers A.W.
        • Troelstra A.
        • Fluit A.C.
        • et al.
        Independent root-cause analysis of contributing factors, including dismantling of 2 duodenoscopes, to investigate an outbreak of multidrug-resistant Klebsiella pneumoniae.
        Gastrointest Endosc. 2019; 90: 793-804
        • Kim S.
        • Russell D.
        • Mohamadnejad M.
        • et al.
        Risk factors associated with the transmission of carbapenem-resistant Enterobacteriaceae via contaminated duodenoscopes.
        Gastrointest Endosc. 2016; 83: 1121-1129
        • Aumeran C.
        • Poincloux L.
        • Souweine B.
        • et al.
        Multidrug-resistant Klebsiella pneumoniae outbreak after endoscopic retrograde cholangiopancreatography.
        Endoscopy. 2010; 42: 895-899
        • Akinbobola A.B.
        • Amaeze N.J.
        • Mackay W.G.
        • et al.
        "Secondary biofilms" could cause failure of peracetic acid high-level disinfection of endoscopes.
        J Hosp Infect. 2020; 107: 67-75
        • Lee D.H.
        • Kim D.B.
        • Kim H.Y.
        • et al.
        Increasing potential risks of contamination from repetitive use of endoscope.
        Am J Infect Control. 2015; 43: e13-e17