BACKGROUND AND AIMS: In a pilot study, we demonstrated that current guidelines for duodenoscope and linear echoendoscope (DLE) reprocessing using a single cycle of high-level disinfection (HLD) in an automated reprocessor may be inadequate. In August 2015, the U.S. Food and Drug Administration offered double HLD as a possible response to address this concern. As a result, Providence Health and Services adopted double HLD as standard procedure for DLEs, but no rigorous clinical studies supported this practice. We undertook a quality improvement study to compare single HLD versus double HLD at 4 of our 34 hospitals.

METHODS: HLD of DLE was randomized, separately in each facility, to either single HLD or double HLD on weekdays, with standard double HLD on weekends or holidays. There was 99.7% compliance with the randomization scheme. Daily qualitative surveillance cultures of dried, post-HLD DLEs were collected for 6 months (1 swab sample from the elevator mechanism and 1 combined brush sample from the suction and working channels for each encounter), and each sample was incubated for 48 hours. Positivity rates of any microbial growth and growth of high-concern pathogens (potentially pathogenic enteric flora) were compared between the 2 study arms.

RESULTS: Altogether, 5850 surveillance culture specimens were obtained during 2925 encounters from the 45 DLEs in clinical use in the participating hospitals. Of these, 3052 (52.2%) were from endoscopes cleaned by double HLD. Double HLD demonstrated no benefit over single HLD because similar positivity rates were observed (all P > .05). The elevator mechanism was more frequently colonized than the biopsy channel (5.2% vs 2.9%, P < .001). Among the 224 encounters with positive growth, 140 (62.5%) recovered microbes from only the elevator mechanism specimens, 73 (32.6%) recovered microbes from only the channel specimens, and 11 (4.9%) recovered microbes from both types of specimens. Double HLD failed to improve contamination rates for either sample site at any of the 4 endoscopy facilities, although there were significant overall differences in contamination rates among the facilities (P < .001), as observed in our previous study. Only 8 high-concern pathogens were recovered from 5 DLEs, all from the elevator mechanism. Persistent growth was observed on 2 duodenoscopes. One grew Enterococcus spp (not vancomycin-resistant enterococci) on 3 occasions, and Escherichia coli was present on 2 of these occasions, 1 of which was a multidrug-resistant organism. The other grew different enteric flora on 2 specimens.

CONCLUSIONS: Our prospectively randomized study, involving 4 separate endoscopy facilities and standard automated endoscope reprocessing, showed that double HLD did not reduce culture positivity rates compared with single HLD in facilities with an already low positive culture rate. Alternative risk mitigation strategies will be assessed in an ongoing effort to reduce endoscope contamination.