Department of Veterans Affairs Office of Public Health Public Health Investigation Summary: Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs July 28, 2015 Veterans Health Administration Office of Public Health Table of Contents Executive Summary . . . . . . . . 3 Introduction . . . . . . . . 4 . . . . . . . 4 . Methods and Definitions Findings . . . . . . . . . . 5 Limitations . . . . . . . . . 6 Conclusions . . . . . . . . . 8 References . . . . . . . . . . 9 Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 2 Executive Summary • 55,676 endoscopic retrograde cholangiopancreatography (ERCP) or endoscopic ultrasonography (EUS) procedures were performed in 40,329 Veterans between 2010 through February 2015 • During the same time frame, 4,914 Carbapenem Resistant Enterobacteriaceae (CRE) isolates from 2,383 unique Veteran patients were found from any anatomic source • Ninety seven (97) patients from 39 VA medical centers were identified as having had an ERCP/EUS (total 148 procedures) and at least one CRE bacterial isolate • Twenty two (22) of the 39 VA medical centers had more than one patient who had an ERCP/EUS procedure and CRE isolate. Among 8 of these VA medical centers, there were 42 patient pairs who had procedures within 6 months of each other at the same facility and had a positive culture for the same bacterial species • There were 6 pairs that could not be completely ruled out, but any transmission appears highly unlikely • Our findings do not provide evidence of CRE transmission related to duodenoscopy Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 3 Introduction Recent media and investigative reports have described transmission of carbapenem resistant Enterobacteriaceae (CRE) infections potentially linked to duodenoscopy procedures (i.e., endoscopic retrograde cholangiopancreatography (ERCP) or endoscopic ultrasonography (EUS)) performed at UCLA Medical Center, Cedars-Sinai Medical Center, Virginia Mason Medical Center, and others1-5. The U.S. Food and Drug Administration issued a safety communication in February 2015 about duodenoscopes6. As a result, the Veterans Health Administration (VHA) Office of Public Health (OPH) received a request from the National Director of Gastroenterology, Specialty Care Services in February 2015 to conduct an epidemiologic review to determine whether there was any evidence that Veterans may have become infected with CRE following a duodenoscopy procedure. This final report summarizes results of the review. Methods and Definitions VA data warehouses were queried for hospital station number, patient identifiers, duodenoscopy procedure date, Current Procedural Terminology (CPT) codes, and International Classification of Diseases, 9th Revision (ICD-9) codes for ERCP and EUS procedures performed in a VA medical center or non-VA (via fee-basis) from January 1, 2010 through February 28, 2015, and patients with positive cultures for CRE (any specimen type) collected at a VA medical center (during the same time period). CRE were further characterized by bacterial species and anatomic site. CRE positive cultures were identified using the U.S. Centers for Disease Control and Prevention (CDC) definition: positive for Enterobacteriaceae (Citrobacter, Cronobacter, Edwardsiella, Enterobacter, Escherichia, Klebsiella, Morganella, Pantoea, Proteus, Providencia, Raoutella, Salmonella, Serratia, and Shigella species were included) meeting the following criteria7: 1) Nonsusceptible to one of the following carbapenems: doripenem, meropenem, or imipenem AND 2) Resistant to all of the following third-generation cephalosporins that were tested: ceftriaxone, cefotaxime, and ceftazidime. CRE were isolated in VA using standard microbiologic methods according to local VA medical center practice. Results from the queries were merged into one dataset containing patients who had ERCP/EUS procedure codes and at least one positive culture for CRE. From this dataset, clusters (3 or more) or pairs of patients who had an ERCP/EUS procedure within a 6-month period of each other (time period parameter based on personal communication with CDC staff performing ongoing investigation at Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 4 non-VA facilities) at the same VA medical center or from the same non-VA facility and had positive cultures for the same CRE bacterial species were identified. Possible patient clusters or pairs underwent electronic health record review to obtain additional information including procedure note documentation of endoscope make, model, and serial number, if available. A cluster or pair of patients was considered linked if they had an ERCP/EUS procedure in the defined timeframe, had the same CRE bacterial species identified (the potential case patient must have had the positive culture after their procedure), and duodenoscope information indicated the same duodenoscope was used in all patients. If duodenoscope information (i.e., make, model, or serial numbers) did not match, clusters or pairs were ruled out for the possibility of transmission of CRE. If the information could not be found or was unavailable, the VA medical center was contacted for assistance to resolve issues or provide further information. After initial review of the CRE data, there were 20 VA medical centers where we were unable to identify CRE based on the definition used in this review (i.e., where carbapenem or third generation cephalosporin susceptibility data may not have been available). Patients with isolates that were only carbapenem non-susceptible or only third generation cephalosporin resistant and had an ERCP/EUS were evaluated for patient clusters or pairs in these 20 VA medical centers in the same manner described above. VA medical center staff was contacted for assistance to provide specific microbiology information, if available, that was not found in the medical record or VA data warehouses. Findings Our analysis found that 55,676 ERCP/EUS procedures were performed for 40,329 Veterans from January 1, 2010 through February 28, 2015. During the same time frame, 4,914 CRE isolates from 2,383 unique Veteran patients were found from any anatomic source. Klebsiella species were the most common organism identified in all geographic regions. CRE was isolated most frequently in the urine (2,915 isolates; 59%); respiratory tract (821 isolates; 17%), blood (412 isolates; 8%), Gastrointestinal/abdominal (76 isolates; 2%), and other anatomic sites (690 isolates; 14%). Ninety seven (97) patients from 39 VA medical centers were identified as having had an ERCP/EUS (total 148 procedures) and at least one CRE isolate. Of the 97 patients, 20 had a CRE isolate pre-procedure only, 62 post-procedure only, and 15 had positive CRE cultures pre- and post-procedure. Sixteen (16) of these patients had CRE bacteremia, and 17 patients had CRE isolated from biliary tract or gastrointestinal Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 5 sources post-procedure. The time between ERCP/EUS and CRE isolation ranged from 1 to 1742 days. Seventeen (17) VA medical centers had a single patient with an ERCP/EUS procedure and CRE isolated from cultures obtained either before or after the procedure; we considered these as likely isolated events. Twenty two (22) VA medical centers had more than one patient who had an ERCP/EUS procedure and CRE isolate. Among eight of these VA medical centers, there were 42 patient pairs who had procedures within 6 months of each other at the same facility and had a positive culture for the same bacterial species. For the other 14 VA medical centers, patients were ruled out due to one or more of the following: procedure dates separated by more than 6 months, procedures performed at different facilities, no post-procedure cultures positive for CRE, and/or different CRE organisms identified. Following electronic health record review of the 42 pairs, 36 were ruled out due to scope model/serial numbers not matching, one or both patients not actually having an ERCP/EUS (i.e., miscodes), or patients not having the procedures performed at the same location. There were 6 pairs that we could not completely rule out the possibility of transmission. The duodenoscope model and/or serial numbers were not available for 3 pairs, and we were unable to determine if patients had been exposed to the same duodenoscope. In the other 3three pairs, the same duodenoscope was used. The same duodenoscope was used in additional patients during the time interval between the pair without incident or evidence of CRE. Transmission in these five pairs appears highly unlikely. In the secondary analysis of the 20 VA medical centers where we were initially unable to identify CRE based on the definition used in this review, there were 19 pairs among six VA medical centers who had ERCP/EUS procedures. Seventeen (17) pairs were ruled out, and there were 2 pairs that we could not completely rule out the possibility of transmission. In one of the two pairs, the facility was able to verify the organisms were CRE (this pair was included in the numbers reported in the previous two paragraphs). In both pairs, the same duodenoscope was used. The same duodenoscope was used for additional patients during the time interval between the pair without incident or evidence of CRE. Thus, transmission of CRE within these pairs appears highly unlikely. Limitations A number of limitations are present in this review. Clinical and Laboratory Standards Institute (CLSI) breakpoints for determining susceptibility to carbapenem and cephalosporin antimicrobial agents for Enterobacteriaceae changed in 2010, and it is not clear when or if all VA medical center laboratories have adopted these changes. Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 6 These changes affect whether an organism is identified as CRE. Additionally, there are differences among VA medical centers as to which carbapenem and cephalosporin antimicrobials are tested and/or what and how susceptibilities are reported which may have affected identifying positive CRE isolates for this review. We were able in some cases to retrieve primary susceptibility data from microbiology laboratory computer sources, but in other cases, this data was no longer available. Electronic microbiology data was not available for one VA medical center because microbiology laboratory testing is performed at a neighboring DoD facility and thus, could not be retrieved and included in this analysis. Therefore, it is possible based on our definition, CRE isolates could be underestimated over this 5-year period. Our review was limited to patients with a positive CRE isolate from any VA medical center. It is possible that patients undergoing duodenoscopy at a VA or non-VA facility may have had CRE infection diagnosed at a non-VA facility which we would not be able to identify using the electronic laboratory methods described above. In addition, some clusters or pairs had procedures at non-VA facilities (fee basis) making it difficult to obtain specific information about ERCP/EUS scope model/serial numbers from the affiliated/contracted medical center. Isolates of CRE organisms are not routinely saved at the respective VA medical centers, so molecular fingerprinting could not be performed on similar isolates to determine whether they were in fact the same strain. Because this review relied on identification of CRE via clinical cultures, we could not identify individuals undergoing duodenoscopy who may have been colonized with CRE without displaying overt signs of infection to warrant testing via a clinical culture or individuals where transmission of CRE via use of common duodenoscopes may have occurred at a subclinical level. Documentation of a unique duodenoscope identifier (such as a scope number or serial number) along with the make/model and manufacturer of the scope utilized for the procedures under review was inconsistently available in the electronic health record or through manual review of instrument records. We did not review all ERCP/EUS procedures to determine whether they were appropriately coded. Finally, we did not review every patient with a CRE isolate to determine whether they had a non-coded or miscoded ERCP/EUS procedure or a non-coded ERCP/EUS procedure at a non-VA facility. For the 17 VA medical centers where a single patient with an ERCP/EUS procedure and CRE isolate was identified, we cannot definitively rule out the possibility that these patients developed their infection as a result of their procedure, however, as there were Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 7 no additional patients found at these sites, there was no clear linkage to a duodenoscopy procedure. Finally, this review did not include an assessment of reprocessing practices at VHA facilities, therefore we cannot comment on the type or adequacy of duodenoscope reprocessing at VHA medical centers, including those where pairs were identified. Conclusions We found no evidence of CRE transmission related to duodenoscopy in VA between January 1, 2010 and February 28, 2015. However, given the limitations of our review outlined above, we cannot completely exclude the possibility that transmission occurred. Submitted on June 24, 2015 by: Mark Holodniy, MD VHA Office of Public Health Public Health Surveillance & Research With acknowledgements to OPH staff: Gina Oda, MS, Patricia Schirmer, MD, Cynthia Lucero-Obusan, MD, Russell Ryono, PharmD, and Gayathri Shankar MS, who performed the epidemiologic investigation; and Kent Davis, Specialty Care Services, for assisting with duodenoscope procedure workload. Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 8 References 1. CDC Statement: Los Angeles County/UCLA investigation of CRE transmission and duodenoscopes. Last updated February 20, 2015. Available at: http://www.cdc.gov/hai/outbreaks/cdcstatement-LA-CRE.html (Accessed March 13, 2015). 2. Muscarella LF. Risk of transmission of carbapenem-resistant Enterobacteriaceae and related “superbugs” during gastrointestinal endoscopy. World J Gastrointest Endosc 2014; 6(10): 457-474. 3. Notes from the Field: New Delhi Metallo-β-Lactamase–Producing Escherichia coli Associated with Endoscopic Retrograde Cholangiopancreatography — Illinois, 2013. Morbidity and Mortality Weekly Report, January 3, 2014 / 62(51); 1051-1051 http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6251a4.htm?s_cid=mm6251a4_w (Accessed March 16, 2015) 4. Wendorf KA, Kay M, Baliga C, et al. Endoscopic Retrograde Cholangiopancreatography-Associated AmpC Escherichia coli Outbreak. Infect. Control Hosp. Epidemiol 2015; 36:634-642. 5. Epstein L, Hunter JC, Arwady MA, et al. New Delhi Metallo-β-Lactamase–Producing Carbapenem-Resistant Escherichia Coli Associated with Exposure to Duodenoscopes. JAMA. 2014 Oct 8; 312(14):1447-55. 6. Design of Endoscopic Retrograde Cholangiopancreatography (ERCP) Duodenoscopes May Impede Effective Cleaning: FDA Safety Communication. Last updated March 4, 2015. Available at: http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm434871.htm (Accessed March 13, 2015). 7. Centers for Disease Control and Prevention. Available at: http://www.cdc.gov/hai/organisms/cre/definition.html (Accessed March 16, 2015) Epidemiologic Review of Carbapenem Resistant Enterobacteriaceae (CRE) and Duodenoscopes in the Department of Veterans Affairs 9