Endoscopic retrograde cholangiopancreatography (ERCP) is a widely used procedure. Bacteremia or infection are a known complication of ERCP. Reference Anderson, Shimpi and McDonald1 The rate of ERCP-related infections is 15.2%. Reference Larsen, Russell and Ockert2 Duodenoscopes contain many small functional parts, so proper reprocessing and disinfection is essential to prevent endoscope-related infections. Reference Beilenhoff, Biering and Blum3,Reference Guideline Task Force, Petersen and Cohen4 However, several outbreaks of multidrug-resistant organisms (MDROs) have been reported after ERCP even though the manufacturers’ reprocessing instructions were strictly followed. Reference Epstein, Hunter and Arwady5,Reference Humphries, Yang and Kim6
We describe 2 ERCP-related outbreaks of carbapenem-resistant New Delhi metallo-β-lactamase (NDM)–producing Escherichia coli that occurred 7 months apart at Rambam Health Care Campus (RHCC), a 1,000-bed, university-affiliated hospital in Haifa, northern Israel. Approximately 500 patients undergo ERCP at RHCC each year. At the time of the outbreaks, 2 types of duodenoscopes were in use: TJF-Q190V (with disposable end cap) and TJF-Q180V (with fixed end cap). Carbapenem-resistant Enterobacteriaceae screening guidelines at RHCC are listed in Supplementary Table 1 (online).
First outbreak
On day 1 and day 3, E. coli NDM was detected in 2 patients hospitalized at the surgical department during routine surveillance. Remarkably, E. coli NDM was extremely rare in our hospital, so the infection prevention and control (IPAC) staff initiated an epidemiological investigation, which revealed that both patients had undergone ERCP procedures in our hospital 1 week earlier.
On day 18, a third patient was admitted to the hospital with septic shock secondary to E. coli NDM 3 days after ERCP. The 3 cases raised immediate concern that the infections were related to recent ERCP procedures.
Second outbreak
On day 214, 7 months later, a rectal screening of a 2-year-old patient performed after transfer from the pediatric intensive care unit yielded a positive result for E. coli-NDM. A rapid investigation revealed that he had undergone an ERCP procedure 3 days earlier (day 211). Also, 8 days earlier (day 203), a known carrier of E. coli-NDM (index case 2) had been treated with the same endoscope (endoscope 2-TJF-Q180V, serial no. 2507365) (Table 1).
Note. NDM, New Delhi metallo; ERCP, endoscopic retrograde cholangiopancreatography; N/A, not available.
Methods
In both investigations, the medical records of all patients who had undergone an ERCP procedure up to 45 days before the mentioned events were reviewed. The IPAC team observed the endoscope’s reprocessing techniques. All involved endoscopes were sampled and returned to the manufacturer for evaluation.
Results
First investigation
The 3 patients underwent procedures with the same endoscopic device (endoscope 1-TJF-Q180V, serial no. 2507371). Another endoscope was used to re-examine one of the patients (endoscope 2-TJF-Q180V, serial no. 2507365). Both endoscopes were immediately taken out of service, and ERCP procedures at the hospital were discontinued.
An investigation revealed that a known E. coli NDM carrier with E. coli NDM detected in his gallbladder aspirate underwent ERCP with the first endoscope 42 days before day 1. This patient was presumably the index case 1 (Table 1).
All patients who had undergone an ERCP procedure with endoscope 1 (16 patients) after the index case or with endoscope 2 (7 patients), which was used to re-examine one of the patients, were invited for rectal screening. All 23 patients were screened at least once; 9 were screened twice at least 1 month apart. All results were negative.
Deficiencies in cleaning and storage of endoscopes were noted. Reprocessing of the duodenoscopes was performed according to manufacturer’s instructions and included immediate manual wiping of the outside of the duodenoscope and leak testing, then soaking with an enzymatic solution and manual reprocessing within an hour of completion of the procedure, followed by automated reprocessing with HLD, and finally vertical hanging for drying in an unventilated cabinet. The process of cleaning and storing clean endoscopes occurred in the same room without separation between clean and contaminated zones.
After the outbreak, cleaning and disinfection processes of duodenoscopes were moved to the main gastroenterology cleaning unit that work according to infection prevention and control standards. Clean endoscopes were stored in standardized lockers in a separate room adjacent to the ERCP unit.
Affected endoscopes were sampled and E. coli-NDM grew in cultures from 3 channels of the first endoscope. Cultures from the second endoscope and the brushes were sterile. Molecular typing revealed that all organisms had carbapenemase gene bla NDM-19. All isolates were E. coli ST167 with the same resistance profile and antibiotic resistance genes. According to the phylogenetic tree, the 5 isolates were very closely related (Supplementary Fig. 1 online). In addition, all samples appeared to contain the same plasmid, IncX3. After laboratory confirmation, all 5 endoscopes used in the ERCP unit were immediately sterilized with ethylene oxide gas.
The endoscopes were revised by the manufacturer and showed numerous cracks and corrosion damage to the insertion tube and protector, the control unit connector, and the distal end cap. The damage was repaired before they were reused.
For devices used in patients known to be MDRO carriers, a cleaning protocol with double reprocessing cycles was implemented. Small brushes that fit on the tip of the elevator portion of the endoscope were procured to clean inaccessible parts of the endoscopes.
Second investigation
E. coli-NDM was detected in cultures from the endoscope 2. However, 5 patients who had undergone ERCP with this endoscope after the index case were screened for CPE and the culture results were negative. Positive samples from the index case, the child, and endoscope 2, were sent to the referral microbiology laboratory for molecular identification.
Molecular typing revealed that all samples had the carbapenemase gene bla NDM-1. All isolates were E. coli ST115 with the same resistance profile, which was different from the isolates of the first outbreak (Supplementary Table 2 online). According to the phylogenetic tree, the 3 isolates were very closely related (Supplementary Fig. 1 online). Similarly, all samples appeared to contain the plasmid ColpVC.
Inspection by the manufacturer revealed numerous cracks and deformations in the distal end cap. The TJF-Q180V endoscopes were not returned to service due to nonusability, and all endoscopes were replaced with PENTAX duodenoscopes. The ED34-i10T2 video duodenoscope from PENTAX Medical (Montvale, NJ) has a sterile disposable elevator cap that is easily replaced and improves the cleanability of the duodenoscope. No additional cases have been reported since the second outbreak (28 months).
Discussion
We report here 2 genetically unrelated outbreaks of E. coli-NDM transmission via ERCP endoscopes. The occurrence of 2 unrelated events within a few months despite improvement in HLD compliance supports our hypothesis of contamination from inaccessible parts of the elevator channels. Contamination rates can be as high as 5.7% despite double HLD or ethylene gas sterilization, Reference Larsen, Russell and Ockert2 underscoring the difficulty of cleaning and supporting our decision to use endoscopes with disposable elevator end caps. This action is supported by the recent US Food and Drug Administration recommendation to switch to disposable end caps, which allows more effective reprocessing of endoscopes. 7
Awareness of violations in the cleaning process needs to be heightened, and audits should be conducted regularly. Reference Cristina, Sartini and Schinca8 Manual bedside cleaning and leak testing should be performed immediately after the procedure. Proper manual cleaning of endoscopes with special brushes compatible with the internal channels and detachable parts of the endoscope is critical to remove debris and to prevent biofilm formation. Reference Valeriani, Agodi and Casini9
Small cracks in the internal canals of duodenoscopes could serve as escape points for organic materials and bacteria and subsequent biofilm formation. Thus, annual compatibility testing by the manufacturer may be warranted.
In addition, microbiologic surveillance of duodenoscopes should be performed at specific intervals, as recommended by the Center for Disease Control and Prevention. 10 In the case of repeated positive cultures, evaluation by the manufacturer is warranted. Reference Cristina, Sartini and Schinca8,10 The investigations of these outbreaks were conducted after the detection of a very rare MDRO infection. Our findings may suggest that other events caused by more common pathogens may go unnoticed.
In conclusion, regular review of duodenoscope reprocessing procedures by IPAC teams is essential, even in the absence of visible deficiencies. However, strict adherence to manufacturers’ reprocessing guidelines is probably not sufficient to prevent all ERCP-related infections. Endoscopes should be inspected regularly for the presence of internal cracks or damage that could promote biofilm formation and that duodenoscopes be used with disposable end caps to ensure optimal cleaning. 10
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2023.21
Acknowledgments
Financial support
No financial support was provided relevant to this article.
Conflicts of interest
All authors report no conflicts of interest relevant to this article.