Koen van der Ploeg, Margreet C Vos, Saskia H J Renkens, Bibi C G C Mason-Slingerland, Corné H W Klaassen, Hardik Rughwani, D Nageshwar Reddy, Sadhana Yelamanchili Veturi, Ravikanth Vishnubhotla, Sana Fathima Memon, Marco J Bruno, Juliëtte A Severin
Rectal screening is unreliable for detecting strain-specific duodenal MDROs and overestimates true carriage due to low PPV values. However, it provides excellent rule-out value in low-prevalence settings. Consequently, its utility for guiding infection prevention strategies for duodenoscope contamination is greatest in low-MDRO-prevalence regions.
BACKGROUND: Duodenoscopes have been implicated in patient-to-patient transmission of multidrug-resistant organisms (MDROs). Current gastrointestinal MDRO surveillance relies on rectal screening, yet the duodenum is the primary site of duodenoscope exposure. This study evaluated the predictive value of rectal and throat-nose screening for detecting duodenal MDROs as a marker of duodenoscope contamination risk.
METHODS: Adult patients undergoing endoscopic retrograde cholangiopancreatography (ERCP) at tertiary care centers in the Netherlands and India were included. Rectal swabs, throat-nose swabs, and duodenal aspirates were analyzed for MDROs. The detected MDROs were compared using species identification, antibiotic susceptibility patterns, and whole-genome sequencing.
RESULTS: Among 512 participants (Netherlands: 339; India: 173), rates of duodenal and rectal MDRO carriage were higher in India (56.6% (98/173); 79.8% (138/173), respectively) than in the Netherlands (5.6% (19/339); 10.3% (35/339)). Given the low prevalence of throat-nose carriage, only the predictive value of rectal screening was assessed. Rectal screening sensitivity for genetically related strains in duodenal fluid was 32.7% in India and 68.4% in the Netherlands . For detecting any duodenal MDRO, sensitivity reached 91.8% in India and 84.2% in the Netherlands, with specificities of 36.0% and 94.1%. Positive predictive value (PPV) was low (India: 65.2%; Netherlands: 45.7%), while negative predictive value was 77.1% and 99%, respectively.
CONCLUSIONS: Rectal screening is unreliable for detecting strain-specific duodenal MDROs and overestimates true carriage due to low PPV values. However, it provides excellent rule-out value in low-prevalence settings. Consequently, its utility for guiding infection prevention strategies for duodenoscope contamination is greatest in low-MDRO-prevalence regions.