To the Editor—In July 2015, patient A was admitted from an outside hospital to the University of Minnesota Medical Center with complaints of abdominal pain, nausea, and weakness. The patient’s medical history was significant for multiple gastric-bypass surgeries, abdominal fistulas, and abdominal operations. The patient also had a prior history of carbapenem-resistant and extended-spectrum β-lactamase–producing Klebsiella pneumoniae (ESBL-KP) from an abdominal surgical site and from endotracheal sputum. Repeated cultures from the sites remained positive for ESBL-KP. On day 6 of admission, the patient underwent an esophagogastroduodenoscopy (EGD) procedure to assess for possible bowel obstruction.

At the same time, the Infection Prevention Department at University of Minnesota Medical Center identified 5 patients with past or current carbapenem-resistant Enterobacteriaceae (CRE) infections who were currently inpatients at the hospital. An outbreak investigation was initiated and subsequently revealed that these patients were on a variety of hospital units and received a variety of services, but several of these patients had had multiple endoscopy or bronchoscopy procedures over the preceding weeks. As a part of the investigation, all gastroscopes, bronchoscopes, and duodenoscopes used on the CRE-positive patients were identified, sequestered, and cultured for microbiological growth. The scopes were sampled according to the Interim Duodenoscope Sampling Method ¹ from the Centers for Disease Control and Prevention, with slight adjustments to accommodate the scopes without an elevator channel. For rinsing, sterile water was used instead of phosphate-buffered saline (0.01M) with 0.02% Tween®-80 solution. The samples were processed immediately. The brushes were cultured in tryptic soy broth and the rinse water was cultured using membrane filtration on tryptic soy agar.

None of the cultures from the 3 sequestered bronchoscopes were positive. One duodenoscope was cultured and grew organisms that indicated oral flora contamination. A total of 3 gastroscopes were cultured: 1 had no growth and 1 grew 1 colony-forming unit of Burkholderia cepacia. The third gastroscope was used on patient A and cultured positive for ESBL-KP. Pulsed-field gel electrophoresis (PFGE) testing was conducted on K. pneumoniae isolates from the gastroscope, patient abdominal site, and patient sputum in coordination with the Minnesota Department of Health, Public Health Laboratory (Figure 1). PFGE was performed using the PulseNet standardized protocolReference Ribot, Fair, Gautom, Cameron, Hunter, Swaminathan and Barrett ² with slight modifications. The resulting K. pneumoniae isolate subtypes differed by 1 band and were therefore closely related according to the Tenover criteria.Reference Tenover, Arbeit, Goering, Mickelsen, Murray, Persing and Swaminathan ³ These results suggest contamination of the gastroscope with the ESBL-KP.

FIGURE 1

PFGE results for abdominal mesh isolate obtained January 12, 2015, and sputum and gastroscope isolates obtained July 26–30, 2015. These isolates have a 1-band difference.

This gastroscope had previously undergone high-level disinfection 12 times and had been used on 9 patients between the day it was used on patient A and the date it was cultured. Endoscope cleaning and high-level disinfection procedures in the endoscopy unit were reviewed regularly in the preceding year and at the time of this incident, and no gaps in following the manufacturer’s instructions for use and reprocessing were identified. Repeated cultures collected after high-level disinfection were negative for ESBL-KP. The 9 patients, along with their care teams, were notified of the exposure to a contaminated gastroscope. To date, no evidence of transmission of the ESBL-KP to any of the 9 patients has appeared, and they remain closely watched for signs and symptoms of transmission.

Examples of CRE outbreaks associated with exposure to duodenoscopes have been well documented.Reference Epstein, Hunter and Arwady ⁴ Despite meticulous adherence to manufacturer’s guidelines for endoscope reprocessing, there have also been reports of persistent biological contamination of colonoscopes and gastroscopes.Reference Ofstead, Wetzler and Doyle ⁵ In France, a well-documented outbreak of ESBL-producing Pseudomonas aeruginosa was associated with a contaminated gastroscope.Reference Bajolet, Ciocan and Vallet ⁶ However, subsequent observations of endoscope reprocessing identified deviations from guidelines. We believe this is the first reported evidence of cross contamination of ESBL-KP from a patient to a gastroscope, with subsequent persistent contamination despite reprocessing using the manufacturer’s instructions. This finding demonstrates a need for more effective methods of cleaning and disinfection and an improvement in the scope design that allows for better disinfection.